Kfcp phenology report 2 dec 2013 s16

Page 1

PHENOLOGY PAPER 2 Trends in fruiting and flowering phenology with relation to abiotic variables in Bornean peat-swamp forest tree species suitable for restoration activities M. E. Harrison, Simon J. Husson, Nicole Zweifel, Laura J. D’Arcy, Helen C. Morrogh-Bernard, Maria A. van Noordwijk and Carel P. van Schaik, as part of the OuTrop multi-disciplinary research project and University of Zurich. Kalimantan Forests and Climate Partnership (KFCP)



PHENOLOGY PAPER 2 Trends in fruiting and flowering phenology with relation to abiotic variables in Bornean peat-swamp forest tree species suitable for restoration activities.

M. E. Harrison, Simon J. Husson, Nicole Zweifel, Laura J. D’Arcy, Helen C. Morrogh-Bernard, Maria A. van Noordwijk and Carel P. van Schaik, as part of the OuTrop multi-disciplinary research project and University of Zurich. May 2013


This report was prepared in accordance with the guidelines at the time of writing, including the overview of the KFCP project below. This research was carried out in collaboration with the Governments of Australia and Indonesia, but the analysis and findings in this paper represent the views of the author/s and do not necessarily represent the views of those Governments.

Phenology Paper 2

Page i


EXECUTIVE SUMMARY In this report, we present the results of a six-year study of the fruiting phenology of Bornean peat-swamp forest trees at two sites: Sabangau and Tuanan. The objective of this report is “to determine the phenology of a broad range of tropical peat-swamp forest tree species found in Central Kalimantan”, in order to facilitate peat-swamp forest restoration activities in Indonesia. Particularly, our aim here was to provide information on the timing, duration and amount of fruit produced during fruiting events, and to identify potential abiotic triggers of fruiting and flowering for a selected sub-set of species identified by KFCP as being potentially best suited for restoration activities. In doing so, we provide the first analysis of potential triggers for flowering and fruiting in Bornean peat-swamp forest; provide insight into potential intra-specific variations in flowering and fruiting phenology, and phenological triggers, between sites; present novel methods for assessing abiotic triggers of flowering and fruiting; and present information useful for the establishment of seed collection protocols for forest restoration practices. Analyses are presented in the form of both summary tables and figures comparing information between sites, and species accounts, in which the different investigations are described in detail for each species. In addition to the data provided to assist with the development of species-specific seed collection protocols, the following key findings emerged from this analysis: 1. The timing of the onset of flowering and fruiting was related between sites for some species, but this was not true in the majority of cases. In most cases, this appears related to the frequency of flowering and fruiting events at the two sites. 2. At least some species appear to have different abiotic triggers that trigger flowering and fruiting events. 3. Comparison of the most likely potential abiotic triggers of flowering and fruiting indicates that different abiotic variables trigger flowering and fruiting in different species. 4. In apparent contrast to the fact that peat-swamp forest is not a masting habitat, some species in the sample do appear to respond to abiotic triggers hypothesised to be associated with SAMF events. 5. Solar equinox was the most frequently identified trigger of both flowering and fruiting among the species sampled. 6. The same potential abiotic triggers of flowering and fruiting for a species were shared between the two sites in some, but not the majority, of species. In most cases where there was no congruence, the species in question either flowered/fruited very rarely or very often. 7. In cases with multiple species analysed in a genera, most species were identified as sharing the same potential abiotic trigger of flowering and/or fruiting in both sites (e.g., Calophyllum: onset of wet season; Cratoxylon: spring equinox). The exception to this was Palaquium, which appears to respond to a number of different potential abiotic triggers. Highlights: • •

Most likely abiotic triggers of flowering and fruiting identified (p179) Calendar showing likelihood of fruiting throughout the year (p185)

Phenology Paper 2

Page ii


CONTENTS EXECUTIVE SUMMARY............................................................................................................................................ ii CONTENTS ..............................................................................................................................................................iii ACKNOWLEDGEMENTS .......................................................................................................................................... v ACRONYMS .............................................................................................................................................................vi INTRODUCTION ...................................................................................................................................................... 1 Fruiting Phenology in Tropical Peat-Swamp Forests .......................................................................................... 1 Studies of Abiotic Triggers of Flowering and Fruiting ........................................................................................ 2 METHODS ............................................................................................................................................................... 3 Definitions .......................................................................................................................................................... 3 Site Locations and Descriptions.......................................................................................................................... 3 Sabangau ........................................................................................................................................................ 4 Tuanan ............................................................................................................................................................ 4 Phenology Sampling Methods ............................................................................................................................ 4 Sabangau ........................................................................................................................................................ 4 Tuanan ............................................................................................................................................................ 5 Sampling Period .............................................................................................................................................. 5 Enumeration ................................................................................................................................................... 5 Species Identification ..................................................................................................................................... 5 Monthly data collection ................................................................................................................................. 5 Tree Species Included in the Sample .................................................................................................................. 6 Abiotic Data Collection and Analysis Methods................................................................................................... 9 Rainfall ............................................................................................................................................................ 9 Temperature ................................................................................................................................................... 9 Occurrence of El Ni単o conditions and events ................................................................................................ 9 Haze from forest fires ..................................................................................................................................... 9 Day length..................................................................................................................................................... 10 Data Analysis .................................................................................................................................................... 10 Analysis of Potential Phenological Triggers ...................................................................................................... 11 RESULTS ................................................................................................................................................................ 14 Variations in Abiotic Variables.......................................................................................................................... 14 Temperature and Rainfall ............................................................................................................................. 14 Visibility ........................................................................................................................................................ 22 Comparison between sites ........................................................................................................................... 22 Fruiting and Flowering Phenology Species Accounts ....................................................................................... 24 Relationships Between Sites for Onset of Flowering and Fruiting ................................................................. 162 DISCUSSION ........................................................................................................................................................ 163 Comparison of Timing of Flowering and Fruiting Onset Between Sites......................................................... 163 Potential Abiotic Triggers of Flowering and Fruiting ...................................................................................... 163

Phenology Paper 2

Page iii


Anomalous Environmental Events that may Influence Flowering and Fruiting ............................................. 168 REFERENCES ....................................................................................................................................................... 169 APPENDIX I - CALENDAR OF SPECIES WITH HIGH LIKELIHOOD OF FRUITING IN DIFFERENT MONTHS AND FRUITING REGULARITY CATEGORIES .................................................................................................................. 172

Phenology Paper 2

Page iv


ACKNOWLEDGEMENTS We thank KFCP for providing the necessary support for data analysis and the writing of this report. The following individuals and organisations are thanked for their assistance at the respective sites. Sabangau • The Centre for the International Cooperation in Management of Tropical Peatlands (CIMTROP), Lembaga Ilmu Pengetahuan Indonesia (LIPI) and Kementerian Negara Riset dan Teknologi Republik Indonesia (RISTEK) for research permissions. • Dr Ir. Suwido H. Limin and the CIMTROP administrative staff. • Twentinolosa Firtsman, Dewa, Andri Thomas, Zeri Yeen, Sahdo, Ahmat, Ella, Ben Buckley, Lindy Thompson, Reychell Chadwick and especially Santiano, Ari Purwanto and Fransiscus Agus Harsanto for phenology data collection and data entry. • The above assistants, plus all OuTrop staff and volunteers that helped with plot creation and enumeration. • Zeri Yeen, Santiano, Selpi and especially Erna Shinta from the CIMTROP Herbarium for tree identification. • The L.S.B. Leakey Foundation, Wildlife Conservation Society, US Fish and Wildlife Service Great Ape Conservation Fund, Wingate Foundation, Primate Conservation Inc., Conservation International Primate Action Fund, Columbus Zoo, the Rufford Foundation and the Orang-utan Tropical Peatland Project Volunteer Programme for financial support. • Jack Rieley, Susan Page, David Chivers and the Wildlife Research Group, University of Cambridge for their support. Tuanan • Director General Departemen Kehutanan (PHKA), Departamen Dalam Negri, BKSDA Palangka Raya, Lembaga Ilmu Pengetahuan Indonesia (LIPI) and Kementerian Negara Riset dan Teknologi Republik Indonesia (RISTEK) for research permissions. • Borneon Orangutan Survival Foundation (BOS) and MAWAS Palangka Raya for permission to work in the MAWAS Reserve. • Tatang Mitra Setia, Sri Suci Utami Atmoko and students of the Universitas Nasional (UNAS) • Arbainsyiah, Ambriansyah and Kade Sidiasa from the Wanariset Herbarium for tree identification. • Meredith L. Bastian, Erin R. Vogel and Serge A. Wich for creating the plots, data entry and fruitful discusssions. • Nadi, Rahmat, Ganda, Yandi Tono Idun and Awan for data collection. • Zürich University and the A. H. Schultz Foundation for financial support.

Phenology Paper 2

Page v


ACRONYMS CIMTROP

Centre for International Cooperation in the Management of Tropical Peatlands

DBH

Diameter (of a tree) at breast height (1.3 m above ground level, or 1.3m above emergent stilts/buttresses)

ENSO

El Ni単o Southern Oscillation

KFCP

Kalimantan Forests and Climate Partnership

MSF

Mixed (peat) swamp forest

NLPSF

Natural Laboratory of Peat-Swamp Forest (Sabangau)

OuTrop

The Orang-utan Tropical Peatland Project

PSF

(Tropical) peat-swamp forest

SA

Sabangau

SAMF

Supra-annual mast fruiting

SD

Standard deviation

TU

Tuanan

Phenology Paper 2

Page vi


INTRODUCTION In this report, we present the results of a six-year study of the fruiting phenology of selected Bornean peatswamp forest trees, identified as potentially suitable for restoration activities at two sites: Sabangau and Tuanan. This report was commissioned by the Kalimantan Forests and Climate Partnership under a subcontract agreement between the Orangutan Tropical Peatland Project (OuTrop) and HK Logistics Pty Ltd. The objective of this project is “to determine the phenology of a broad range of tropical peat-swamp forest tree species found in Central Kalimantan”, in order to support KFCP’s forest restoration activities. In an earlier progress report (Harrison et al., 2010), we provided information on the basic fruiting phenology of all of the tree species within the phenology plots at the two research sites. Analyses were presented for each species at each site and included density estimates, identification of the timing of fruiting peaks and the variability surrounding these peaks. Each species was further classified as having either regular/predictable or non-regular/unpredictable fruiting patterns. Based on this, a calendar detailing when each species could be reliably expected to bear fruit in each month was drawn up. Following this first report, KFCP identified a shortlist of potentially suitable tree species for peat-swamp forest restoration activities, for which a more detailed phenological analyses is presented herein. This includes analysis of potential abiotic triggers, fruiting durations and crop sizes, plus a discussion of ‘anomalous environmental events’ that could potentially disrupt flowering and fruiting. The aim of this report is to provide KFCP with the information necessary to develop fruit collection protocols for the selected species. Fruiting Phenology in Tropical Peat-Swamp Forests While studies of forest phenology (i.e., the study of periodic biological phenomena, in this case fruiting events) have been prevalent throughout many areas of Borneo, few such studies have been conducted in PSF. Dry lowland evergreen forests cover large areas of Borneo and are notable for their high abundance of Dipterocarpaceae, which are associated with supra-annual mast-fruiting (SAMF) events, occurring at irregular intervals every 2-10 years. These SAMF events primarily involve dipterocarps, but also many other tree families (Janzen, 1974; Appanah, 1985; Ashton et al., 1988; Appanah, 1993; Sakai et al., 1999; Sakai, 2002). During SAMF events, 55-60% of tree species may bear fruit, compared to only 20-25% in non-mast years (McClure, 1966; Medway, 1972). SAMF events lead to huge peaks and troughs in fruit and flower availability in the forest, which in turn have dramatic effects on many species of forest fauna, including important seed dispersers (Leighton and Leighton, 1983; Curran and Leighton, 2000; Ostfeld and Keesing, 2000; Frederiksson et al., 2006; Harrison and Chivers, 2007). In contrast, PSF has a much lower abundance of Dipterocarpaceae (van Schaik, 1996; Shepherd et al., 1997; Page et al., 1999; Harrison et al., 2010) and, consequently does not participate in SAMF events (Cannon et al., 2007a, b). As a result, fruit production in PSF is relatively consistent and lacks the dramatic peaks and troughs in fruit availability that occur in dipterocarp forests (Cannon et al., 2007a, b). Previous phenology studies in Indonesian PSF have been conducted in the highly-fertile PSF of Suaq Balimbing, Sumatra (van Schaik, 1999; Marshall et al., 2009) and in small patches of PSF in Gunung Palung, West Kalimantan (Knott, 1999; Marshall and Leighton, 2006; Cannon et al., 2007a, b). Short-term phenological data have also been reported from the extensive PSFs of Central Kalimantan, collected at the two sites included in this report (Marshall et al., 2009; Harrison et al., 2010). With the exception of Cannon et al.’s work in Gunung Palung, however, the remaining phenology analyses cited above have been performed from a primate-centric viewpoint, and all have been conducted at a forest-wide scale. Thus, detailed species-by-species accounts of fruiting phenology in Bornean PSF, which are potentially of great use to PSF conservation practitioners and restoration ecologists, have yet to be conducted. This gap in knowledge was addressed in our earlier progress report (Harrison et al., 2010). Information on the fruiting phenology of tree species is important for developing protocols and schedules for forest restoration, particularly for planning schedules for seed collection (FORRU 2005). While analysis of temporal fruiting patterns are necessary, the abiotic conditions that may trigger flowering and fruiting in a species vary between years, and so prediction of the timing of flowering and fruiting events based on these

Phenology Paper 2

Page 1


variables is necessary. Similarly, extreme anomalous abiotic events, such as drought and forest fires may also influence phenology patterns; for example, evidence exists to suggest that litter-fall in PSF increases dramatically during El Niño-induced forest fires (Harrison et al., 2007). Such analyses are presented in this report, in order to support KFCP’s restoration activities. Studies of Abiotic Triggers of Flowering and Fruiting A number of authors have attempted to assess potential abiotic triggers of flowering and fruiting in tropical forests, with a particular focus on triggers for SAMF events (e.g., Ashton et al., 1988; Tutin and Fernandez, 1993; van Schaik et al., 1993; Chapman et al., 1999; Yasuda et al., 1999; Anderson et al., 2005; van Schaik and Pfannes, 2005; Brearley et al., 2007). Fruiting must always follow flowering (though all flowering events may not necessarily turn into fruiting events) and so, while many authors have attempted to assess potential abiotic triggers of fruiting, assessment of triggers for flowering have been more common. Potential abiotic triggers of flowering and fruiting come in many forms, and most abiotic variables have been suggested as potential triggers in at least one study. These include rainfall (Brearley et al., 2007) and prolonged drought (Medway, 1972; Janzen, 1974), temperature (Wycherley, 1973), in particular minimum temperature (Ashton et al., 1988; Tutin and Fernandez, 1993; Yasuda et al., 1999), solar irradiation (i.e., peaks in irradiation associated with solar zenith, van Schaik et al., 1993; Wright and van Schaik, 1994; van Schaik and Pfannes, 2005) and day length (Audus, 1972; Vince Prue et al., 1984). Scientists have long been aware of a link between the occurrence of El Niño events and SAMF in the dipterocarp forests of South-east Asia, and sustained drops in minimum night-time temperatures associated with the penetration of dry air masses into the aseasonal tropics that occur during these events are hypothesised to be the most likely trigger for floral induction in these SAMF events, with a two month lag between the occurrence of the trigger and the actual emergence of flowers (Ashton et al., 1988; Appanah, 1993). While there have been many studies of potential triggers of flowering and fruiting in the tropics, the majority of these studies have been conducted at a forest, rather than a species, level and none have been conducted in Bornean peat-swamp forest. Those studies that have looked at potential triggers of flowering and fruiting have often found that all species do not appear to be induced by the same abiotic trigger. For instance, Günter et al. (2008) found that of 13 tree species studied at two sites in Cordillera Real, Ecuador, the initiation of flowering was associated with the spring or autumn equinox in eight species in at least one site, and with rainfall, temperature or day length in other cases, and that flowering and fruiting were best predicted with various time lags ranging from 0-6 months for flowers and 1-9 months for fruits. In some species, potential differences were found between the abiotic variables best associated with flowering at the two sites (ibid.). Thus, it is important to assess the full range of potential abiotic variables that may trigger flowering and fruiting, and to perform analyses with varying time lags.

Phenology Paper 2

Page 2


METHODS Definitions Fruiting stem

A stem that has produced fruit at least once during the study period and, hence, is known to be capable of fruit production.

New flowering event

The first month of a flowering event

New fruiting event

The first month of a fruiting event

Reproductive stem

A stem that has produced flowers (and may or may not have produced fruit) at least once during the study period and, hence, is fertile and potentially capable of reproduction.

Stem

An individual tree

Site Locations and Descriptions The two study sites included in this report – Sabangau and Tuanan – are each located in the extensive belt of peatland that stretches across the lowlands of southern Kalimantan (Figure 1).

Figure 1. Location of Sabangau (NLPSF) and Tuanan in Central Kalimantan, Indonesia.

Phenology Paper 2

Page 3


Sabangau The Sabangau study site (2o 19’ S and 113o 54’ E) is within the 500 km2 Natural Laboratory of Peat-Swamp Forest (NLPSF), located in the north-east of the 6,300 km2 Sabangau Forest. The phenology plots monitored for this study are located in an area of ombrogenous mixed-swamp forest towards the edge of the peat dome, where peat depth varies from 1 - 4 m (Page et al., 1999). This is an area of intermediate productivity, compared to the low-productivity low-pole forest further towards the centre of the peat dome and more highly-productive tall-interior forest on the deepest peat at the centre of the peat dome, which is unique to Sabangau (Page et al., 1999; Morrogh-Bernard et al., 2003; Sulistiyanto et al., 2004). The study area was subject to selective logging from 1991-1997, followed by indiscriminate illegal logging until March 2004 (Husson et al., 2007). Tuanan The Tuanan research station (2o09’ S and 114o26’ E) is located in the 5,010 km2 Mawas Reserve, along the Kapuas Murung River. The peat here is shallower than in Sabangau (≤ 2 m deep) and partly alluvial (Wich et al., 2009), and thus is expected to show higher levels of productivity than the PSF in Sabangau. The site is recovering from selective commercial logging during the early 1990s, followed by opportunistic (illegal) logging until the end of 2002 (van Schaik et al., 2005) and ongoing extraction of ‘gemur’ (Alseodaphne coriacea). Phenology Sampling Methods Sabangau Sabangau phenology data was collected from six permanent phenology plots and one phenology circuit (Figure 2). The six plots are located at roughly equi-distant intervals from 1-4 km from the forest edge, running parallel to the black-water River Sabangau. Each plot is divided into two parallel sub-plots: Sub-Plot A measuring 5 x 300 m and containing stems ≥ 6 cm DBH, and Sub-Plot B measuring 5 x 500 m containing stems ≥ 20 cm DBH. This gave a total plot area of 2.4 ha, with 0.9 ha of plot including stems ≥ 6 cm DBH and 1.5 ha of plot including stems ≥ 20 cm DBH. Species density estimates for stems of DBH ≥ 10 cm were thus made using only data from Sub-Plots A, whereas density estimates for stems ≥ 20 cm DBH were made using data from both Sub-Plots. The phenology circuit was established to include certain important orang-utan and gibbon food species that were relatively rare in the phenology plots (e.g., Diospyros siamang, Parartocarpus venenosus, Tetrameristra glabra). The stems in the phenology circuit were not used to estimate stem density, as they are not within a plot with specified area, but were used to aid in identifying fruiting peaks for each species. Phenology surveys generally began on the 15th of each month and lasted for seven days.

Figure 2. Map of phenology plot and circuit locations in Sabangau. East-west transects start at the railway / boardwalk on the western side of the grid and are coded using numbers indicating the approximate distance (in km) of each transect’s start point from the field station. The north-south transects, which run parallel to the railway / boardwalk, are coded with letters and are separated by 250-m intervals.

Phenology Paper 2

Page 4


Tuanan Tuanan phenology data was collected from a single, large phenology plot located along two transects in the centre of the study area. One transect ran parallel, and the other perpendicular, to the Kapuas River. All stems ≥ 10 cm DBH within 5 m of the transects were included, covering a total plot area of 2 ha. Phenology surveys generally began on the 12th of each month (occasionally on the 10th) and lasted for 3 – 4 days. Sampling Period Data collection started simultaneously at both sites in September 2003, has continued monthly and is ongoing. Data for this report cover the period September 2003 to August 2009 for Sabangau, and June 2009 for Tuanan, representing 72 and 70 months for which data are available from the two sites, respectively. The methods used have been outlined previously (Marshall et al., 2009; Harrison et al., 2010). Enumeration In each plot, all trees above the designated minimum DBH limits were enumerated, measured (DBH, height) and identified using both Latin and local names. Stems were re-measured every two (SA) to three (TU) years to account for growth. At these times, any new trees satisfying the minimum size requirements for the plots were enumerated, identified, measured and recruited into the phenology surveys thereafter. Numerous trees died during the surveys (both as a result of natural and, less frequently, anthropogenic causes) and these trees were removed from the sample once death was confirmed. Some stems fell over during the study and, unless they died, were still included in the phenology sample as they generally remained reproductive. Stems that snapped and exhibited coppice re-growth from the snapped stem were removed from the sample, as these stems were no longer capable of producing fruit from their small immature shoots. Non-tree species (lianas and Ficus) were sampled, but are excluded from this analysis. Species Identification Tree species identification was performed with the assistance of expert local botanists and members of the local communities with extensive knowledge of PSF fauna. Collaboration in confirming consistency in species identifications between the two sites have been extensive and are ongoing. This has involved visits of researchers from both sites to the other, use of the same expert botanists (E. Shinta, Z. Yeen, Arbasinsyiah, Ambriansyah and Kade Sidiasa), joint visits to herbariums, and extensive discussions and sharing of photographs. Nevertheless, despite these efforts, there is still uncertainty regarding the shared identity of certain species in our data set. This is often due to high numbers of similar species in certain genera (e.g., Tristaniopsis, Litsea). Thus, in such cases, we have adopted a cautious approach and presented results under the assumption each of the species at the two sites are different. It is therefore possible that a small number of species classed here as different may in fact be the same at the two sites. This has resulted in one or two instances where species given the same Latin name are presented on different sheets for the two sites, as it was clear that these did not represent the same species. Despite this, the Latin nomenclature will be correct in the large majority of cases. Because local names frequently differ between the two sites (e.g., “tagula”, which refers to a species of the genera Xylopia, Annonaceae in Sabangau, but to species of Alseodaphne and Litsea, Lauraceae in Tuanan) and can often refer to more than one true species (e.g., “jambu burung” in Sabangau, which refers to a number of species of Syzygium), they are problematic for comparisons between sites. Nevertheless, even the best local village tree experts familiar with > 100 different species will have no knowledge of Latin nomenclature. Thus, local names are also presented for each site, but should not be used with caution and should not be assumed to accurately indicate taxonomy. Monthly data collection Each stem was observed monthly from ground level using binoculars to record reproductive activity. The presence and abundance of fruit, flowers and leaf flush (not included in this report) was recorded for each stem. Spotting and identifying reproductive stages in tall PSF trees requires considerable skill, so the same highly-experienced observers were used each month to reduce this variability. Distinguishing ripe/unripe fruit, and immature/mature flowers, is difficult, even for experienced observers familiar with each species, as (i) the

Phenology Paper 2

Page 5


observer is usually viewing the fruit/flowers from afar, making confirmation of maturity stages difficult, even with binoculars; (ii) the similarity of ripe and unripe stages for many fruit species; and (iii) the observer is frequently looking up at silhouetted shapes of the fruit/flowers against the sun or a green backdrop, making inspection of colour difficult. Inspection of fruit/flowers that have fallen to the ground, and inference from this, is also problematic, as ripe fruits/mature flowers are more likely to fall to the ground – either as a result of being dislodged by a predator (indehiscent) or being released by the parent tree once a set stage of ripeness/maturity has been achieved (dehiscent) – than unripe fruits/immature flowers, creating potential bias. Thus, no attempt to distinguish ripe/unripe fruits and immature/mature flowers has been made in our analyses. For each stem recorded as bearing flowers and/or fruit in a month, crop size (i.e., the number of fruits and/or flowers in the canopy) was recorded by counting the number of fruits/flowers in a sub-sample of the crown and then estimating the number in the total tree crown by extrapolation. The following crop-size categories were used: A B C D E

= = = = >

1-10 11-100 101-1,000 1,001-10,000 10,001

Tree Species Included in the Sample Following assessment of the restoration potential of the different PSF tree species by KFCP and consideration of the fruiting phenology information in our first report (Harrison et al., 2010), a final list of tree species to be analysed in this report was drawn up (Table 1). In a few cases, the species suggested for inclusion by KFCP were not present in our phenology plots and, hence, a suitable substitute species present within the plots was used wherever possible. Furthermore, in some instances, results are presented with the different species in the genera merged, rather than broken down into species. These species/genera were determined after consultation with KFCP and are noted.

Phenology Paper 2

Page 6


Table 1. PSF tree species included in analysis

Family Anacardiaceae Anisophyllaceae

Species suggested Campnosperma coriaceum Combretocarpus rotundatus Dyera lowii / polyphylla

Species included C. coriaceum C. rotundatus

Clusiacea / Guttiferae

Calophyllum hosei

C. hosei (SA) / nodosum (TU)

Clusiacea / Guttiferae Dipterocarpaceae

Calophyllum sclerophyllum

C. sclerophyllum

Shorea balangeran

Dipterocarpaceae Ebenaceae Elaeocarpaceae

Vatica rassak Diospyros siamang Elaeocarpus petiolatus

S. balangeran, parvifolia (TU) and teysmanniana (SA) Vatica rassak D. siamang E. mastersii

Fabaceae / Leguminosae Hypericaceae

Koompassia malaccensis

K. malaccensis

Cratoxylon arborescens

C. arborescens / spp.

Hypericaceae

Cratoxylon glaucum

C. glaucum / spp.

Lauraceae Melastomataceae Meliaceae Myristicaceae Myrtaceae Myrtaceae

Litsea spp. Pternandra galeata Aglaia rubiginosa Horsfieldia crassifolia Eugenia spicata Syzygium oblatum

L. spp. P. cf. coerulescens / galeata A. rubiginosa H. crassifolia E. cf. spicata S. cf. garcinifolia / havilandii

Apocynaceae

Phenology Paper 2

D. lowii / polyphylla

Notes

Classification revised since last report (Harrison et al., 2010) from D. lowii (synonymous) C. hosei not present in Tuanan plots (though possibility this may be the same species as C. hosei in Sabangau), so substituted by C. nodosum, the most numerous species of this genera in Tuanan ID confusion between sites; same Latin used at each site for different species. Sabangau team confident of ID, Tuanan ID less certain. S. balangeran included in plots, but sample size low at both sites, so commonest Shorea species at each site also included. V. mangachopai E. periolatus not present in plots, E. mastersii has largest sample size in genus and so best available substitute size, but is probably less flood tolerant than E. periolatus and hence may differ in phenology.

Cratoxylon species presented both individually and combined, as very difficult to distinguish species in the field. Cratoxylon species presented both individually and combined, as very difficult to distinguish species in the field Individual species combined, as very difficult to distinguish in field ID between sites possibly the same, but uncertain

ID between sites possibly the same, but uncertain Species not present in plots, so substituted by the two most common species in genus combined

Page 7


Myrtaceae

Tristaniopsis obovata

T. spp.

Sapotaceae Sapotaceae

Madhuca motleyana Palaquium spp.

Tetrameristaceae Theaceae

Tetramerista glabra Ploiarium alternifolium

M. motleyana P. coclearifolium, leiocarpum, pseudorostratum and ridleyii / cf. xanthochymum T. glabra P. alternifolium

Thymelaeaceae Anacardiaceae Anacardiaceae Anacardiaceae Apocynaceae Apocynaceae Dipterocarpaceae Dipterocarpaceae Euphorbiaceae Euphorbiaceae Euphorbiaceae Lauraceae Myrtaceae Myrtaceae Myrtaceae

Gonystylus bancanus Mangifera altissima Gluta renghas Gluta wallichii Alstonia spatulata Alstonia pneumatophora Shorea leprosula Shorea pauciflora Macaranga hypoleuca Macaranga pruinosa Mallotus muticus Alseodaphne coriacea Melaleuca cajuputi Eugenia cerina Syzygium zippeliana

G. bancanus NA NA NA NA NA NA NA NA NA NA NA NA NA NA

Podocarpaceae

Dacrydium pectinatum

NA

Phenology Paper 2

Tristaniopsis phylogeny complex and species very difficult to distinguish in field, so results presented for all species combined Species kept separate, as relatively easy to distinguish in the field

In Tuanan plots, but probably not in optimum conditions, as this is a light lover common in sedge, so note may not exhibit normal phenology in swamps Genus not in sample Genus not in sample Genus not in sample Genus not in sample Genus not in sample Species not in sample Species not in sample Genus not in sample Genus not in sample Genus not in sample Species did not fruit (large stems absent) Genus not in sample Species not in sample Species not in sample (flood resistant and so probably doesn’t match any Syzygium species present in our plots) Genus not in sample

Page 8


Abiotic Data Collection and Analysis Methods All abiotic data are expressed in terms of a ‘phenology month’ (i.e., from the usual start day of phenology surveys in each month), rather than calendar months, to ensure comparability with the phenological data. Rainfall Rainfall was recorded daily using a rain gauge positioned in an open area near the respective research camps. Results were expressed as mean daily rainfall, rather than total monthly rainfall, because (i) the number of days in a month varies between months, which will influence figures for total monthly rainfall, but not mean daily rainfall; (ii) for various reasons, rainfall data were not collected on every single day of the study period (particularly around Christmas and Idul Fitri holidays), which would also lead to bias in figures of total monthly rainfall; and (iii) given a perfect data set, the two variables would be virtually perfectly correlated, as they essentially measure the same thing. Wet and dry seasons, and drought months, were defined based upon inspection of the data sets for the two sites. The critical rainfall value used to distinguish wet and dry season months was 6 mm/day, and drought months were defined as those months with less than 4 mm rainfall/day. On occasions, one anomalous month of higher or lower rainfall occurred in the middle of a dry or wet season, respectively. Such anomalous months were classified as belonging to the season in which they fell. The first month of each wet and dry season, and drought period, were identified and the subsequent months were listed as month two, three, etc., after the onset of the relevant season. This numbering continued until the next season of that type (i.e., months following the onset of the dry season were numbered up until the start of the next dry season, and therefore included wet season months, as flowering/fruiting may occur six months after the onset of the dry season, regardless of whether this still falls within the dry season, or the wet season has already begun). As prolonged drought has been suggested as a potential abiotic trigger of fruiting in some studies (e.g., Wood, 1956; Medway, 1972; Janzen, 1974), we also identified periods with two, three, etc., consecutive drought months. Temperature Temperature at both sites was recorded daily using minimum-maximum thermometers positioned in shaded areas just inside the forest near the respective research camps. Mean 24-h temperatures were then calculated from the minimum and maximum values. In addition to mean temperatures over a period, four-day rolling mean minimum temperatures were determined, as sustained drops in minimum temperature below 20˚C are hypothesised to trigger mastflowering and -fruiting events in dipterocarp forests (Ashton et al., 1988). Thus, if one or more mean rolling four-day temperatures were less than 20˚C, then that month was classified as a sustained low minimum temperature month. Similarly, four-day rolling mean maximum temperatures were also calculated, as this is another variable that could potentially influence phenology patterns. Occurrence of El Niño conditions and events The El Niño Southern Oscillation (ENSO) involves changes in sea-surface temperatures between the western and eastern Pacific ocean (e.g., (McPhaden and Picaut, 1990). The preferred measure of sea-surface temperature in this part of the world is the NINO3 anomaly, which responds best to the state of the tropical ocean-atmospheric system (Latif et al., 1998; Shukla, 1998). Data on the NINO3 index during the survey period was obtained through National Weather Service Climate Prediction Centre (http://www.cpc.noaa.gov/data/indices). We classified a month as experiencing El Niño conditions if the NINO3 index exceeded -0.5 for half or more of the weeks in the month (cf. Wich and van Schaik, 2000). Following standard procedure, an El Niño event was only judged to occur if El Niño conditions persisted for five months or more. Haze from forest fires Fire is becoming an increasingly frequent problem in the peat-swamps of Indonesia, as human activities drain the peatlands, leading to increased incidence and severity of fires and thick palls of haze, especially

Phenology Paper 2

Page 9


during El Niño (Page et al., 2002; Harrison et al., 2009). We were not able to measure haze from forest fires directly, but were able to obtain a suitable proxy of this: visibility records from the Cjilik Riwut airport in Palangka Raya, located 11 km from the Sabangau research area and 56 km from Tuanan. The airport records visibility each day and flights are frequently delayed or cancelled in severe smoke seasons. Data on mean, minimum and maximum daily visibility were obtained for the period June 2004-August 2009 courtesy of the Weather Underground website (http://www.wunderground.com). Day length Day length for both sites was calculated using the NewMoon version 1.0 computer programme, which calculates astronomical data based on the GPS location of a site (cf. Cheyne, 2007). This software is available for free download at www.psgb.org. Data Analysis The same timeframe (September 2003-June 2009) was used for both sites in calculation of the majority of statistics presented herein, in order to avoid potentially skewing fruiting times for one site relative to the other. Slightly different timeframes (September 2003-August 2009 for Sabangau and September 2003-June 2009 for Tuanan) were used for the phenological triggers analyses in the two sites, as Tuanan data for July and August 2009 were not available for use in this report. These extra two months were included in the Sabangau data set in order to improve statistical power. Thus, our monthly estimates of mean fruit production and variability in fruit for September-June are comprised of six data points, whereas our estimates for July and August are comprised of five data points. This may reduce the precision of our estimates for these latter two months for species that frequently come into fruit at these times. For each of the selected species, the number of “flowering” and “fruiting” stems were recorded (note that the former term is equivalent to a “reproductive stem”; see definitions). These were defined as those stems that had produced flowers or fruit at least once during the study period and, hence, were capable of flower or fruit production, respectively. Non-flowering and non-fruiting stems were defined as those stems that had never produced flowers or fruit, respectively, during the study period. These latter two figures therefore did not vary from month to month, with the exception of stems that were added or died during the course of the study. The number of flowering (reproductive) or fruiting stems was the denominator in calculations of “percentage flowering stems” and “percentage fruiting stems”. In some cases, a stem may have matured and fruited for the first time in the middle/end of the study period, and been incapable of fruit production before that time, reducing the precision of our estimates. This will be relatively rare, however, and it will be much more common that stems potentially capable of flower or fruit production that did not produce flowers or fruit early on in the study (or at all) did so for reasons unrelated to their age/maturation; e.g., that species’ reproductive cycle results in flower/fruit production only once every three years. Similarly, for each month, each stem was recorded as bearing flowers, bearing fruit or having ‘no phenology’. These figures formed the numerator in calculations of “percentage flowering stems with flowers” and “percentage fruiting stems with fruit”. Thus, the figure for percentage fruiting stems with fruit was calculated as (the percentage of flowering stems with flowers was calculated in exactly the same way): Percentage of fruiting stems with fruit = (Number of stems of species x with fruit in month y / number of fruiting stems of species x) x 100 This was determined to be the most suitable figure for this analysis because (i) stem densities vary between the two sites (and will also differ in KFCP’s site); (ii) percentages are more comparable between species that occur at different densities within the respective sites; and (iii) species that do not produce flowers or fruit add nothing to efforts to identify the timing of fruiting peaks and can potentially skew percentages if they die/are added to the plots mid-way through the study period. All reproductive stems, regardless of DBH, were analysed as different tree species begin to produce flowers and fruit at different sizes (see Results). Thus, for one species, a minimum DBH of for example 15 cm may

Phenology Paper 2

Page 10


include a number of stems that will never produce fruit (as fruit is not produced in this species until over ca. 20 cm DBH), and exclude a number of stems that are capable of producing fruit for another species (as fruit production begins at ca. 10 cm DBH for this species). Thus, we included all stems for each species that produced flowers and fruit, respectively, to produce our estimates, and state the minimum DBH that produced fruit in the species results sheets. The minimum sample size for inclusion in the analysis was one reproductive stem. Four considerations regarding sample size should be taken into account when interpreting our results: (i) the larger the sample size of stems monitored, the more confidence can be placed in the results for a particular species; (ii) variability in figures for percentage flowering/fruiting stems with flowers/fruit will be higher for species with very low sample size (e.g., if there is only one fruiting stem of a species, then percentage fruiting stems with fruit in a month can only ever be 0 or 100%); (iii) flowering and fruiting events for a species generally do not involve all stems of that species producing flowers/fruit at the same time (see Results), so some flowering and fruiting events may be missed if the small number of stems included in the sample do not participate in a particular flowering or fruiting event, resulting in an impression of less regular flowering/fruiting for species with smaller sample size; and (iv) the larger the sample size, the higher the probability of small trees producing fruit, because production of fruit by very small stems is uncommon. Based on the species accounts created via the above methods, we listed those species/genera that have a high probability of fruiting in each month. Using this calendar, we listed all the species at both sites and categorised them according to the regularity and consistency of their fruiting. A detailed description of these methods is provided in our preliminary report (Harrison et al., 2010) and classifications for those species included in this report are provided in Appendix I. Each species/genera was then classified into the following fruiting categories for each site: 1. Year-round – high likelihood of fruit in 9 or more months of the year; 2. Frequent – high likelihood of fruit in 4-8 months of the year; 3. Regular - high likelihood of fruit in 1-3 identified months of the year; 4. Irregular – no high likelihood of fruit in any particular month of the year; and 5. Never fruiting – never fruited during the study period. For both fruiting duration and crop sizes, median averages were calculated and frequency charts presented, which show the number of times that each fruiting duration or crop size was recorded for each species. Medians were calculated by converting the categories into single numbers (i.e., category “A” becomes “1”, category “B” becomes “2”, etc.). Median averages are preferable to medians in this case, because they are not skewed by the occasional very high recording (e.g., if the ‘typical’ crop size was 11-100, and a crop size of 1,001-10,000 was recorded on one occasion, a mean average would be elevated severely by this one high reading). The frequency charts also allow the modal average (i.e., the value that occurred most frequently) to be identified. In some cases, the median average covers two categories (this occurs if the central value falls in between two categories), and this is indicated by either a decimal place (in the case of fruiting duration) or a wider range (in the case of crop size, e.g., 11-1,000, which covers the categories 10-100 and 101-1,000). All fruiting events that included either the first or last month of the study period were excluded for the purposes of this analysis, because fruiting events may have started before or after the study period and, similarly, the highest crop size may have occurred before or after the study period. Analysis of Potential Phenological Triggers Identification of the triggers of fruiting and flowering events is complex, due to the wide range of potential triggers and different ‘lag periods’ with which these may operate (e.g., the trigger may occur one, two or three months prior to the actual onset of flowering). For the purposes of this study, we assessed the effect of each potential abiotic trigger on flowering and fruiting for a species in both the same month as the phenological data, and with a lag periods ranging from 1-6 months for flowers, and 1-9 months for fruit

Phenology Paper 2

Page 11


(other researchers have found similar lag periods to be effective, e.g., Günter et al., 2008). We refrain from performing analyses with longer lag periods, as have been conducted by some researchers (e.g., Anderson et al., 2005, who performed analyses with lags of up to 19 months), as, although these researchers found significant results using these lags, we consider it unlikely that trees in the aseasonal tropics wait nearly two years after the occurrence of their flowering/fruiting trigger to produce flowers/fruits and that such results are more likely due to chance. We therefore consider six months to be a sensible maximum lag period between trigger occurrence and flower production, with a further three months maximum lag before fruit production. Because solar equinoxes occur once every six months, the maximum lag period used for assessing the potential role of solar equinoxes as an abiotic trigger was five months. In order to assess potential triggers of flowering and fruiting, it was necessary to isolate the first month of each flowering or fruiting event in each individual stem (termed a “new flowering/fruiting event”). This is because flowering and fruiting events typically last longer than one month (see Results) and, hence, the continued presence of flowers or fruit in months following the first month of a flowering or fruiting events is almost certainly a mere consequence of the fact that that stem had flowers or fruits in the month before, rather than a response to any abiotic trigger. Months following the first month of a flowering or fruiting event where the stem still had flowers or fruits were therefore excluded from the analysis, as it was impossible for them to have new flowering or fruiting events in that month. Similarly, fruit phenology data for the first month were also excluded from the analysis, because it is very likely that stems with fruit in the first month of the study still retained fruit from the month before our data collection started and, hence, that this was not “new fruit”. The first month’s data was not excluded for flowers, because PSF trees tend to flower for shorter durations and, hence, this is less likely to influence the results of regressions for analysis of potential abiotic triggers for flowering. In many cases, visual inspection of figures showing the relationships between different variables may be the best method for identifying potential triggers, but this is difficult to employ on a large scale. The total combination of all the different potential triggers for all the species included in the analysis was ca. 10,900, and producing this number of figures and carefully inspecting them all visually would be prohibitively time consuming. Thus, we employed a three, and in some cases four, stage approach to identifying potential abiotic triggers: Stage 1.

Stage 2.

Stage 3.

Phenology Paper 2

Initial screening analysis: This involved using simple non-parametric statistical tests (Spearman’s correlations for continuous variables, and Mann-Whitney or Kruskal-Wallis tests for categorical variables) to identify independent (abiotic) variables that produced a significant effect on the dependent (fruiting/flowering) variable for further analysis. Non-parametric tests were used as many of the variables violated the necessary normality assumptions for parametric tests. Regression analysis: These statistically more powerful tests focused on those variables that produced significant results in the screening tests. Regression analysis allows for equivalent comparisons between the effects of continuous and categorical independent variables on the dependent variable (categorical variables can be included in regression analyses as either dichotomous variables or blocks of dummy variables, and the predictive power of different regression models comparing categorical and continuous variables can be compared via the model’s R2 values (e.g., Zar, 1999). The model with the highest predictive power (i.e., the highest R2 value) was then selected as the most likely abiotic trigger. Where necessary, we transformed variables (arcsine for proportions, log for all others) to ensure normality. We checked the tests’ assumptions were met by examination of plots of standardised residuals and standardised predicted values, distribution of standardised residuals, expected against observed probabilities of standardised residuals. Visual inspection: Inspection of graphical representations of these test results. This was performed to (a) aid interpretation of the statistical results; and (b) to assess whether the results of the test made ‘biological sense’. In some cases, this led to rejection of the regression model with the strongest predictive power and acceptance of a model with slightly less predictive power that made ‘biological sense’.

Page 12


Stage 4.

Refinement. In some cases, the regression analyses and inspection of graphical representations of the results yielded results where, upon close inspection, it was apparent that a different approach may yield stronger predictive power. While such a post-hoc analysis (i.e., picking out potential tests that look like they might yield significant results and then demonstrating that they do) is not generally accepted scientific practice, it was considered appropriate for the current analysis, in which the aim was to attempt to identify the best predictor of flowering and fruiting, rather than testing specific hypotheses on potential abiotic triggers for different species.

Fruiting or flowering may correlate with a number of independent variables. This may often be the result of correlations between the different independent variables; i.e., an independent variable that does not influence flowering/fruiting is correlated with this, simply because it is correlated with the actual trigger. Multiple regressions are not of use here, as the aim is to identify a single trigger. This is because (i) while a combination of independent variables may produce a more powerful statistical model, identifying a combination of, say, four independent variables that together predict flowering or fruiting patterns would be of little practical use; (ii) assumptions of non-collinearity between independent variables are frequently violated, as the different abiotic variables are highly inter-correlated, and inclusion of inter-correlated variables in the same model produces spurious relationships, preventing accurate interpretation of statistical results (Graham, 2003); and (iii) from an ecological and evolutionary perspective, it is probably advantageous to react to a single abiotic cue, rather than a complex combination of different triggers. Outliers are points that lie outside the ‘normal’ range of distribution of a variable (e.g., if 100 points of a variable lay within the range 10-20 units, then one point lying at 50 units would be considered an outlier). Due to the way in which regression statistics are calculated, outliers have an unproportional effect on regression statistics; i.e., one outlier can drastically influence the fit, power and significance of a model. Thus, it is standard practice in regression analyses to exclude such outliers, in order to identify the true relationships between variables (e.g., Zar, 1999). Our aim in the present analysis, however, is to predict the timing of those months with high flower and fruit availability, which will be the statistical outliers (i.e., to distinguish those few months with very high flower/fruit availability from the majority of months with low/zero availability). Consequently, outliers have not been excluded from this analysis. Statistical significance values relate to the probability of the observed result being due to chance or an actual interaction between the two variables. The standard significance (or alpha) value employed is 0.05. A significance less than 0.05 indicates there is a one or less in twenty chance of that particular outcome happening by chance. Thus, when conducting a large number of statistical tests, as in this study, there is a strong possibility that false positive (Type I) results will be obtained (i.e., where the result of the statistical test is non significant, but a true interaction exists between the two variables. Thus, in such situations, significance levels are usually adjusted (through use of, e.g., the Bonferroni correction, (Hochberg, 1988), in order to reduce the risk of false positives occurring. However, while corrections of this nature are recommended when conducting rigorous testing of a specific hypothesis, they are not recommended when the aim is hypothesis generation or the screening of results for further testing (Roback and Askins, 2004). Thus, corrections for multiple comparisons were not performed in Stage 1 of the analysis described above, but were performed in Stage 2. In Hochberg’s (1988) procedure, all significance (p) values from hypothesis tests (in this case, regression analyses) are ranked in inverse order according to their p-value and the observed p-values are compared to critical values derived from dividing the observed p-value by its inverse rank. A specific hypothesis is then rejected if the observed p-value is greater than this critical value. These corrections were performed for Stage 2 tests at the level of species for each site, rather than including all statistical tests performed for all species at both sites, because the aim here is to identify potential abiotic triggers of flowering/fruiting for each species at each site. Thus, adjusted significance levels therefore varied between each species and site, depending on the number of regression analyses performed for that species at that site. The results of trigger tests that were significant pre correction, but not significant post correction, are noted and included in the species accounts, though these are unlikely to be useful for predictive purposes.

Phenology Paper 2

Page 13


RESULTS Variations in Abiotic Variables Temperature and Rainfall Variations in abiotic variables at the two sites during the study period are shown in Figures 3-8. Mean monthly temperatures generally fell within a fairly narrow range at both sites throughout the study period, whereas rainfall fluctuated widely both within and between years. Although the overall variation in temperature was low, the four-day rolling mean minimum temperature dropped below 20 C in only 14/73 (19%) of the months in Sabangau and 2/72 (3%) of months in Tuanan. Eight of these events in Sabangau occurred either in or one month before/after El Ni単o conditions, and three of the five groups of these events occurred around El Ni単o events, indicating a link between these two occurrences. None of these events were associated with El Ni単o in Tuanan.

Phenology Paper 2

Page 14


Phenology Paper 2

6 4 5

2 0 0

Aug-Sep 03 Sep-Oct 03 Oct-Nov 03 Nov-Dec 03 Dec 03-Jan 04 Jan-Feb 04 Feb-Mar 04 Mar-Apr 04 Apr-May 04 May-Jun 04 Jun-Jul 04 Jul-Aug 04 Aug-Sep 04 Sep-Oct 04 Oct-Nov 04 Nov-Dec 04 Dec 04-Jan 05 Jan-Feb 05 Feb-Mar 05 Mar-Apr 05 Apr-May 05 May-Jun 05 Jun-Jul 05 Jul-Aug 05 Aug-Sep 05 Sep-Oct 05 Oct-Nov 05 Nov-Dec 05 Dec 05-Jan 06 Jan-Feb 06 Feb-Mar 06 Mar-Apr 06 Apr-May 06 May-Jun 06 Jun-Jul 06 Jul-Aug 06 Aug-Sep 06 Sep-Oct 06 Oct-Nov 06 Nov-Dec 06 Dec 06-Jan 07 Jan-Feb 07 Feb-Mar 07 Mar-Apr 07 Apr-May 07 May-Jun 07 Jun-Jul 07 Jul-Aug 07 Aug-Sep 07 Sep-Oct 07 Oct-Nov 07 Nov-Dec 07 Dec 07-Jan 08 Jan-Feb 08 Feb-Mar 08 Mar-Apr 08 Apr-May 08 May-Jun 08 Jun-Jul 08 Jul-Aug 08 Aug-Sep 08 Sep-Oct 08 Oct-Nov 08 Nov-Dec 08 Dec 08-Jan 09 Jan-Feb 09 Feb-Mar 09 Mar-Apr 09 Apr-May 09 May-Jun 09 Jun-Jul 09 Jul-Aug 09 Aug-Sep 09

Mean rainfall (mm/d) 30

26

22 20 18

16 20

14

12 10 8 15

Temperature (C)

a.

35

28 30

24

25

10

Months

Page 15


Phenology Paper 2

4 2 5

0 0

Mar-Apr 03 Apr-May May-Jun 03 Jun-Jul 03 July-Aug 03 Aug-Sep 03 Sep-Oct 03 Oct-Nov 03 Nov-Dec 03 Dec 03-Jan Jan-Feb 04 Feb-Mar 04 Mar-Apr 04 Apr-May May-Jun 04 Jun-Jul 04 Jul-Aug 04 Aug-Sep 04 Sep-Oct 04 Oct-Nov 04 Nov-Dec 04 Dec 04-Jan Jan-Feb 05 Feb-Mar 05 Mar-Apr 05 Apr-May May-Jun 05 Jun-Jul 05 Jul-Aug 05 Aug-Sep 05 Sep-Oct 05 Oct-Nov 05 Nov-Dec 05 Dec 05-Jan Jan-Feb 06 Feb-Mar 06 Mar-Apr 06 Apr-May May-Jun 06 Jun-Jul 06 Jul-Aug 06 Aug-Sep 06 Sep-Oct 06 Oct-Nov 06 Nov-Dec 06 Dec 06-Jan Jan-Feb 07 Feb-Mar 07 Mar-Apr 07 Apr-May May-Jun 07 Jun-Jul 07 Jul-Aug 07 Aug-Sep 07 Sep-Oct 07 Oct-Nov 07 Nov-Dec 07 Dec 07-Jan Jan-Feb 08 Feb-Mar 08 Mar-Apr 08 Apr-May May-Jun 08 Jun-Jul 08 Jul-Aug 08 Aug-Sep 08 Sep-Oct 08 Oct-Nov 08 Nov-Dec 08 Dec 08-Jan Jan-Feb 09 Feb-Mar 09 Mar-Apr 09 Apr-May May-Jun 09 Jun-Jul 09 Jul-Aug 09 Aug-Sep 09

Rainfall (mm) 30 28 35

26 24 30

22 25

20

18 16 20

14

12 10 8 15

Temperature (C)

b.

10

6

Months

F Figure 3. Monthly variation in mean rainfall (mm/d) and mean 24-h, minimum and maximum temperature in Sabangau (a) and Tuanan (b). Bars represent rainfall (filled = wet season; not filled = dry); solid lines monthly averages of 24-h means; dotted lines daily min/ max.

Page 16


0

Phenology Paper 2 Aug-Sep 03 Sep-Oct 03 Oct-Nov 03 Nov-Dec 03 Dec 03-Jan 04 Jan-Feb 04 Feb-Mar 04 Mar-Apr 04 Apr-May 04 May-Jun 04 Jun-Jul 04 Jul-Aug 04 Aug-Sep 04 Sep-Oct 04 Oct-Nov 04 Nov-Dec 04 Dec 04-Jan 05 Jan-Feb 05 Feb-Mar 05 Mar-Apr 05 Apr-May 05 May-Jun 05 Jun-Jul 05 Jul-Aug 05 Aug-Sep 05 Sep-Oct 05 Oct-Nov 05 Nov-Dec 05 Dec 05-Jan 06 Jan-Feb 06 Feb-Mar 06 Mar-Apr 06 Apr-May 06 May-Jun 06 Jun-Jul 06 Jul-Aug 06 Aug-Sep 06 Sep-Oct 06 Oct-Nov 06 Nov-Dec 06 Dec 06-Jan 07 Jan-Feb 07 Feb-Mar 07 Mar-Apr 07 Apr-May 07 May-Jun 07 Jun-Jul 07 Jul-Aug 07 Aug-Sep 07 Sep-Oct 07 Oct-Nov 07 Nov-Dec 07 Dec 07-Jan 08 Jan-Feb 08 Feb-Mar 08 Mar-Apr 08 Apr-May 08 May-Jun 08 Jun-Jul 08 Jul-Aug 08 Aug-Sep 08 Sep-Oct 08 Oct-Nov 08 Nov-Dec 08 Dec 08-Jan 09 Jan-Feb 09 Feb-Mar 09 Mar-Apr 09 Apr-May 09 May-Jun 09 Jun-Jul 09 Jul-Aug 09 Aug-Sep 09

Rainfall (mm) 22

18

16 20

14

12 15

10

8 10

4

Temperature (C)

26 30

24

25

20

6

5

2

0

Months

F Figure 4. Mean monthly rainfall and 24-h temperature in Sabangau and Tuanan compared. Bars represent rainfall and lines represent temperature (Sabangau = black; Tuanan = red).

Page 17


27 26

4-day rolling min T (C

25 24 23 22 21 20

15/08/2009

15/06/2009

15/04/2009

15/02/2009

15/12/2008

15/10/2008

15/08/2008

15/06/2008

15/04/2008

15/02/2008

15/12/2007

15/10/2007

15/08/2007

15/06/2007

15/04/2007

15/02/2007

15/12/2006

15/10/2006

15/08/2006

15/06/2006

15/04/2006

15/02/2006

15/12/2005

15/10/2005

15/08/2005

15/06/2005

15/04/2005

15/02/2005

15/12/2004

15/10/2004

15/08/2004

15/06/2004

15/04/2004

15/02/2004

15/12/2003

15/10/2003

18

15/08/2003

19

Date Figure 5. Four-day rolling mean minimum temperatures in relation to El Ni単o conditions and events in Sabangau and Tuanan. The solid lines represents the four-day rolling mean minimum temperature (black = Sabangau; red = Tuanan), the coarsely-dotted line is set at 20 C, black sections of the horizontal bar equate to non-El Ni単o conditions/events, green sections indicate El Ni単o conditions that did not last long enough to be classed as an event, and blue sections indicate El Ni単o events.

Phenology Paper 2

Page 18


36 35 34

4-day rolling max T (C)

33 32 31 30 29 28 27 26

15/08/2009

15/06/2009

15/04/2009

15/02/2009

15/12/2008

15/10/2008

15/08/2008

15/06/2008

15/04/2008

15/02/2008

15/12/2007

15/10/2007

15/08/2007

15/06/2007

15/04/2007

15/02/2007

15/12/2006

15/10/2006

15/08/2006

15/06/2006

15/04/2006

15/02/2006

15/12/2005

15/10/2005

15/08/2005

15/06/2005

15/04/2005

15/02/2005

15/12/2004

15/10/2004

15/08/2004

15/06/2004

15/04/2004

15/02/2004

15/12/2003

15/10/2003

24

15/08/2003

25

Date F Figure 6. Four-day rolling mean maximum temperatures in relation to El Ni単o conditions and events in Sabangau and Tuanan. The solid lines represents the four-day rolling mean maximum temperature (black = Sabangau; red = Tuanan), the coarsely-dotted line is set at 31 C, black sections of the horizontal bar equate to non-El Ni単o conditions/events, green sections indicate El Ni単o conditions that did not last long enough to be classed as an event, and blue sections indicate El Ni単o events.

Phenology Paper 2

Page 19


0

Phenology Paper 2 Jun-Jul 04 Jul-Aug 04 Aug-Sep 04 Sep-Oct 04 Oct-Nov 04 Nov-Dec 04 Dec 04-Jan 05 Jan-Feb 05 Feb-Mar 05 Mar-Apr 05 Apr-May 05 May-Jun 05 Jun-Jul 05 Jul-Aug 05 Aug-Sep 05 Sep-Oct 05 Oct-Nov 05 Nov-Dec 05 Dec 05-Jan 06 Jan-Feb 06 Feb-Mar 06 Mar-Apr 06 Apr-May 06 May-Jun 06 Jun-Jul 06 Jul-Aug 06 Aug-Sep 06 Sep-Oct 06 Oct-Nov 06 Nov-Dec 06 Dec 06-Jan 07 Jan-Feb 07 Feb-Mar 07 Mar-Apr 07 Apr-May 07 May-Jun 07 Jun-Jul 07 Jul-Aug 07 Aug-Sep 07 Sep-Oct 07 Oct-Nov 07 Nov-Dec 07 Dec 07-Jan 08 Jan-Feb 08 Feb-Mar 08 Mar-Apr 08 Apr-May 08 May-Jun 08 Jun-Jul 08 Jul-Aug 08 Aug-Sep 08 Sep-Oct 08 Oct-Nov 08 Nov-Dec 08 Dec 08-Jan 09 Jan-Feb 09 Feb-Mar 09 Mar-Apr 09 Apr-May 09 May-Jun 09 Jun-Jul 09 Jul-Aug 09 Aug-Sep 09

Rainfall (mm/d) 20

15 6

5

10 4

5

Visibility (km)

25 10

9

8

7

3

2

1

0

Month

Figure 7. Mean visibility at the Palangka Raya airport in relation to mean daily rainfall at both sites and the occurrence of El Ni単o conditions/events. The solid line represents visibility at the airport, the bars represent rainfall (black for Sabangau, red for Tuanan), black sections of the horizontal bar equate to non-El Ni単o conditions/events, green sections indicate El Ni単o conditions that did not last long enough to be classed as an event, and blue sections indicate El Ni単o events.

Page 20


735 730 725

Aug-Sep 09

May-Jun 09

Feb-Mar 09

Nov-Dec 08

Aug-Sep 08

May-Jun 08

Feb-Mar 08

Nov-Dec 07

Aug-Sep 07

May-Jun 07

Feb-Mar 07

Nov-Dec 06

Aug-Sep 06

May-Jun 06

Feb-Mar 06

Nov-Dec 05

Aug-Sep 05

May-Jun 05

Feb-Mar 05

Nov-Dec 04

Aug-Sep 04

May-Jun 04

Feb-Mar 04

715

Nov-Dec 03

720 Aug-Sep 03

Mean day length (min)

740

Month Figure 8. Mean monthly day length and equinox timings. The line represents day length, black sections of the horizontal bar represent non-equinox months, blue sections represent autumn equinox months and green sections represent spring equinox months. Equinox timings are the same at both sites and day length is practically identical between sites.

Phenology Paper 2

Page 21


The four-day rolling mean maximum temperature exceeded 31 C in 14/73 (19%) of months in Sabangau and 13/72 (18%) of months in Tuanan. Six of these events in Sabangau occurred either in or one month before/after El Niño conditions occurred, and three of the six groups of these events occurred around El Niño events; in Tuanan, nine of these events occurred either in or one month before/after El Niño conditions occurred, and three of the six groups of these events occurred around El Niño conditions/events, indicating a probable link between these two occurrences. Visibility Visibility at the airport was clearly related to both rainfall and the occurrence of El Niño conditions/events and the resultant wild fires. Visibility decreased drastically during the two El Niño events that occurred during the study period, as a result of the very dry conditions that provide ideal conditions for fire. As a result of the high particulate level of peat fires, visibility also dropped on two other occasions, following periods of unusually low rainfall, which both occurred in close proximity to El Niño conditions. Comparison between sites The large majority of abiotic variables showed a positive relationship between the two sites (Table 2), indicating very similar seasonal trends. The only exceptions to this were for the onset of the dry season (generally earlier at Sabangau) and the four-day rolling minimum temperature > 20 C. Further evidence of climatic similarity between the two sites comes from tests for differences in the different abiotic variables, which showed that the two sites did not differ in terms of rainfall, or length or dry season, wet season and drought periods, but do differ in terms of temperature, with Tuanan being warmer than Sabangau (Table 2). The strong relationship between temperature at the two sites indicates that this difference may be either a true climatic difference, or merely a subtle difference in the positioning of the thermometers/level of shade at the two sites.

Phenology Paper 2

Page 22


Table 2. Relationships and differences between abiotic variables in Sabangau and Tuanan. Variable

Mean Sabangau

Mean Tuanan

Difference significant?

Difference test result

Correlation significant? Yes, positive Yes, positive Yes, positive Yes, positive Yes, positive

Correlation coefficient 0.731 *** (0.679) *** (0.679) *** (0.747) *** (0.407) ***

Mean rain (mm/d) 8.16 7.50 No 1.587 Dry season month 4.67 4.5 No (2,628.0) Wet season month 7.33 7.5 No (2,628.0) Drought month 3.16 3.00 No (2,628.0) Months since start dry NA NA NA NA season Months since start wet NA NA NA NA Yes, positive (0.723) *** season Onset dry season NA NA NA NA No (0.092) Onset wet season NA NA NA NA Yes, positive (0.411) *** Onset drought NA NA NA NA Yes, positive (0.411) *** Mean min temp 22.05 22.96 Yes 16.651 *** Yes, positive 0.698 *** Mean max temp 28.49 28.74 Yes 3.314 ** Yes, positive 0.820 *** Mean 24-h temp 25.26 25.83 Yes 12.822 *** Yes, positive 0.785 *** Rolling 4-d min temp < 14 2 Yes (2,197.0) ** Yes, positive (0.131) 20 C? Rolling 4-d max temp > 14 14 No (2,664.5) Yes, positive (0.381) *** 31 C? Parentheses indicate the use of non-parametric tests (Spearman’s correlation coefficient for relationships and Mann-Whitney for tests of difference) where the assumptions of normality for parametric tests (Pearson’s correlation and t-test) were not met. The “Difference test result” column indicates the t statistic (t-test) or Mann-Whitney U statistic (Mann-Whitney). Asterisks indicate statistical significance: * p < 0.05, ** p < 0.01, *** p < 0.001. Correlation coefficients are Pearson’s correlation coefficient or Spearman’s rank correlation coefficient. Sample size was 73 months in almost all cases. Mean values for “Dry season month”, “Wet season month” and “Drought month” indicate the mean length per year. Mean values for rolling four-day temperatures indicate the total number of months in which these spikes occurred throughout the study period.

Phenology Paper 2

Page 23


Fruiting and Flowering Phenology Species Accounts These accounts are presented alphabetically by Latin name, ordered by family, then genera and species. They contain analyses of variations in fruiting between months, comparison of the timing of fruiting and flowering events, fruiting durations, fruit crop sizes, and potential abiotic triggers of flowering and fruiting. The following notes should be considered when interpreting these accounts: 1. The definitions listed above are adhered to in the accounts. 2. “NA” indicates that there are no available data for that species at that site (i.e., that there were no flowering and/or fruiting stems in the plots). 3. A stem density of “0 / ha” indicates that the species is not present at a site and has never been recorded, either inside or outside of the phenology plots. 4. A stem density of “< 1 / ha” indicates that either (a) the species is found in the phenology plots, but is at very low density (indicated by a total number of stems sampled ≥ 1); or (b) that the species (or DBH class) is found in the site, but is either very rare or not present in the phenology plots (indicated by a total number of stems sampled equal to zero). 5. Fruiting frequency categories match those described above and in (Harrison et al., 2010). 6. Where data from both Sabangau and Tuanan are presented in the same figure, Sabangau is represented in black and Tuanan in red. 7. Boxplots showing the variation in fruit availability between months show the median (line in middle of box), inter-quartile range (box; i.e., the range within which the 50% of cases closest to the median lie), and the minimum/maximum points (outliers, see key below). 8. The results of the abiotic trigger analyses should not necessarily be interpreted as allowing prediction of all fruiting/flowering events, but rather as allowing the forecast of predictable events within the total number of events analysed. Key to figures = minimum or maximum point lying between 1.5 and 3 inter-quartile ranges from the median = minimum or maximum point lying over 3 inter-quartile ranges from the median = fruiting = flowering

Phenology Paper 2

Page 24


Family Latin Name Local name

: Anacardiaceae : Campnosperma coriaceum : Terontang (SA), tarantang (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 40 13 69 53 77 45 65 40 / 76

Tuanan 17 2 33 33 100 14 42 8 / 57

1 11-100 9 Frequent

1 11-100 10 Regular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 25


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 80

70

60

50

40

30

20

10

Month

Flower

Tuanan

45

40

35

30

20 25

15

10

5

Month

Flower

Page 26


Potential Triggers of Flowering Sabangau The best regression model that could be produced was with mean maximum temperature (R2 = 0.151, df = 71, p = 0.001) as the independent variable, which produced a negative relationship. This species flowers virtually continuously and, consequently, this relationship is probably not useful for predictive purposes.

Tuanan The best regression model that could be produced included the mean maximum temperature, with a one month lag (R2 = 0.202, df = 68, p < 0.001): flowering was more likely one month following months with a low mean maximum temperature, though inspection of the scatter plot reveals that flowering events occurred across a wide temperature range. Furthermore, as in Sabangau, this species flowers very regularly and all year round and, therefore, this relationship is probably not very useful for predictive purposes.

Phenology Paper 2

Page 27


Potential Triggers of Fruiting Sabangau No regression model remained significant post-correction for multiple comparisons. The strongest trend was with the first month of the dry season with a two month lag, which had fairly low predictive power (R2 = 0.106, df = 70, p = 0.006). This may due to the regular fruiting of this species throughout the year.

Tuanan The most powerful predictive model contained day length as the potential abiotic trigger (R2 = 0.160, df = 68, p =0.001), with increasing day length two months previously being associated with increased likelihood of fruiting. The predictive power of this model was modest, however, and fruiting events also occurred in months following those months with some of the shortest day lengths.

Phenology Paper 2

Page 28


Potential Triggers Summary Sabangau The predictive models produced were of relatively low predictive power, which is probably a result of the relatively regular fruiting and flowering of this species throughout all months and seasons. It would appear, however, that smoke reduces flowering potential and that new fruiting peaks may be likely to occur in the third month of the wet season (though this latter observation is questionable statistically). Tuanan The regression analyses yielded the mean maximum temperature (operating with a onemonth lag; negative relationship) as the most likely abiotic trigger of flowering and day length (two-month lag; positive relationship) as the most likely trigger of fruiting. As in Sabangau, the predictive power of these models was modest and this species produced flowers and fruits regularly throughout the study period, so these models are of limited practical use.

Percentage occurrence

Fruiting Duration

80 70 60 50 40 30 20 10 0 1

2

3

4

5

6

8

7

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Percentage occurrence

Crop Size

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 29


Family Latin Name Local name

: Anisophyllaceae : Combretocarpus rotundatus : Tumih (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 11 5 22 19 86 18 82 57 / 83

Tuanan 0 0 0 NA NA NA NA NA

2 1,001-10,000 6 / 21* Frequent

NA NA NA NA

Variation in Fruit Availability Between Months

* Minimum DBH producing fruit 6 cm, but on 369 observations of fruiting in this species, 366 (99%) were in trees over 21 cm DBH

Phenology Paper 2

Page 30


Timing of New Flowering and Fruiting Events During Each Month Sabangau

% stems new fruit/flower event

80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Jun-Jul 08

Sep-Oct 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Jun-Jul 06

Sep-Oct 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Jun-Jul 04

Sep-Oct 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Sabangau No significant regression models could be made from those independent variables shown to be significant in the screening tests. This is probably because this species has new flowering events constantly throughout the year in all months/seasons. Potential Triggers of Fruiting Sabangau No significant regression model could be built from those variables that were significant in the screening tests and no obvious patterns emerged from the graphs. This is presumably because this species fruits frequently in all months/seasons. Potential Triggers Summary Sabangau There is a trend towards higher flower production in months 2-4 of the dry season, but no statistically significant models could be built to predict either flowering or fruiting. Again, this is probably because this species produces both flowers and fruits during all months/seasons of the year. This agrees with the flowering vs. fruiting graph.

Phenology Paper 2

Page 31


Percentage occurrence

Fruiting Duration

60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Percentage occurrence

Crop Size

70 60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 32


Family Latin Name Local name

: Apocynaceae : Dyera lowii / polyphylla : Jelutong (SA), pantung (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 16 11 43 28 65 15 35 23 / 80

Tuanan 12 6 23 13 57 7 30 14 / 43

2 11-100 6 / 23* Frequent

1.5 1-10 15 Regular

Variation in Fruit Availability Between Months

* Minimum DBH producing fruit 6 cm, but on 123 observations of fruiting in this species, 120 (97%) were in trees over 23 cm DBH

Phenology Paper 2

Page 33


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month Sabangau

60

50

40

30

20

10

Month

Flower

Tuanan

50

45

40

35

30

25

20

15

10

5

Month

Flower

Page 34


Potential Triggers of Flowering Sabangau The best predictive model was for minimum temperature with a two-month lag (R2 = 0.137, df = 70, p = 0.002). However, although this model had reasonable statistical power, this is probably not an effective model for practical use, due to the relatively unpredictable nature of these temperature changes, and the fact that this species flowers regularly/throughout the year in Sabangau.

Tuanan The most powerful model contained first month of the dry season as a potential trigger for flowering, with a three months lag before flower production (R2= 0.151, df = 69, p = 0.001).

Phenology Paper 2

Page 35


Potential Triggers of Fruiting Sabangau The model with the highest predictive power was minimum visibility with a three-month lag (R2 = 0.437, df = 59, p < 0.001), which shows increased incidence of new fruiting events following periods of low visibility. Possibly the most useful model in practical terms (and with the fourth highest predictive power, after mean minimum temperature with four- and five-month lag) was that of day length as the independent variable (R2 = 0.255, df = 69, p < 0.001), which shows higher incidence of new fruiting events as day length increases.

Tuanan Fruiting was best predicted by the regression model including spring equinox as the independent variable, operating with a three-month lag (R2 = 0.193, df = 68, p < 0.001): higher incidence of fruiting occurred three month after the spring equinox.

Phenology Paper 2

Page 36


Potential Triggers Summary Sabangau The fact that this species flowers and fruits regularly throughout the year in Sabangau again complicates the interpretation of the analysis of abiotic triggers. Minimum temperature is a possible trigger for flowering, with a two-month lag; a strong negative relationship exists between visibility and fruiting, with a two-month lag; and a strong positive relationship exists between fruiting and day length, which is probably the most useful model in practical terms. This does not disagree with the flowering vs. fruiting graph. Tuanan The strongest models that emerged from the analysis included the onset of the dry season as the potential trigger of flowering, with a three-month lag, and the spring equinox as the potential trigger for fruiting, with a three-month lag. This agrees with the flowering vs. fruiting graph, yet inspection of this graph didn’t reveal any consistent and easily-identifiable time-lag between flowering and fruiting.

Percentage occurrence

Fruiting Duration 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

70 60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 37


Family Latin Name Local name

: Clusiaceae / Gutterifae : Calophyllum hosei : Jinjit / bintangor (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 71 11 118 62 53 54 46 0 / 56

Tuanan 0 0 0 NA NA NA NA NA

2 1,001-10,000 7 Regular

NA NA NA NA

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 38


Timing of New Flowering and Fruiting Events During Each Month Sabangau

60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Jun-Jul 08

Sep-Oct 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Sep-Oct 06

Dec 06-Jan

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Jun-Jul 04

Sep-Oct 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

70

Month Fruit

Flower

Potential Triggers of Flowering Sabangau The best regression model was identified post-hoc and included 3 or 4 months lags since the start of the wet season as one category and all the remaining months as another (R2 = 0.259, df = 71, p < 0.001).

Phenology Paper 2

Page 39


Potential Triggers of Fruiting Sabangau The best regression model was identified post-hoc and included 4 or 5 months lag since the start of the wet season as one category and all the remaining months as another (R2 = 0.326, df = 71, p < 0.001).

Potential Triggers Summary Sabangau The regression analyses and graphs show strongly that a large peak in new flowering events occurs 3-4 months after the start of the wet season. This, in turn, is followed by a large peak in new fruiting events one month later. This agrees with the fruiting vs. flowering graphs.

Percentage occurrence

Fruiting Duration

50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Page 40


Percentage occurrence

Crop Size

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 41


Family Latin Name Local name

: Clusiaceae / Gutterifae : Calophyllum nodosum : Mahadingan (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 0 0 0 NA NA NA NA NA

Tuanan 11 3 21 15 71 13 62 0 / 77

NA NA NA NA

2 11-100 13 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 42


Timing of New Flowering and Fruiting Events During Each Month Tuanan

50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

60

Month Fruit

Flower

Potential Triggers of Flowering Tuanan The best predictive model that could be constructed was built post-hoc and included the onset of the wet season as the independent variable, operating with a three-four-month lag (R2 = 0.167, df = 69, p < 0.001).

Phenology Paper 2

Page 43


Potential Triggers of Fruiting Tuanan The onset of fruiting was best predicting using the spring equinox as the potential trigger, with a delay of 5 months until flowering (R2 = 0.165, df = 68, p = 0.001).

Potential Triggers Summary Tuanan Flowering in this species is most likely to be predicted by the onset of the wet season, with a three-four month delay until flower production in around December-February. Fruiting then follows in February-March, five months after the spring equinox. This agrees with the flowering vs. fruiting graph.

Percentage occurrence

Fruiting Duration

40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Tuanan

Phenology Paper 2

Page 44


Percentage occurrence

Crop Size

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Tuanan

Phenology Paper 2

Page 45


Family Latin Name Local name

: Clusiaceae / Gutterifae : Calophyllum schlerophyllum (SA) : Kapurnaga (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 7 <1 19 8 42 8 42 43 / 100

Tuanan 0 0 0 NA NA NA NA NA

3 1,001-10,000 30 Regular

NA NA NA NA

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 46


Timing of New Flowering and Fruiting Events During Each Month Sabangau

% stems new fruit/flower event

100 90 80 70 60 50 40 30 20

Jun-Jul 09

Dec 08-Jan

Mar-Apr 09

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Sep-Oct 05

Dec 05-Jan

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Sabangau There were no significant results in the screening tests of potential abiotic triggers. However, upon inspection of graphical representations of the data, the start of the wet season with a 1-3 month lag appeared to be a likely potential candidate and regression of this produced a reasonable predictive model (R2 = 0.130, df = 71, p = 0.002).

Phenology Paper 2

Page 47


Potential Triggers of Fruiting Sabangau There were no significant results in the screening tests of potential abiotic triggers. However, upon inspection of graphical representations of the data, the start of the wet season with a 2-4 month lag appeared to be a likely potential candidate and regression of this produced a reasonable predictive model (R2 = 0.175, df = 68, p < 0.001).

Potential Triggers Summary Sabangau This analysis suggests that a large peak in new flowering events occurs 1-3 months after the start of the wet season, followed by a large peak in new fruiting events one month later. This agrees with the fruiting vs. flowering graphs.

Percentage occurrence

Fruiting Duration

70 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Page 48


Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 49


Family Latin Name Local name

: Clusiaceae / Gutterifae : Calophyllum sclerophyllum (TU) : Kapurnaga kakal (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number fruiting stems Percentage of fruiting stems Total number flowering stems Percentage of flowering stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 0 0 0 NA NA NA NA NA

Tuanan <1 <1 1 0 0 1 100 NA

NA NA NA NA

NA NA NA Never fruited

Timing of New Flowering and Fruiting Events During Each Month Tuanan 100

% stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Dec 04-Jan

Mar-Apr 05

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Tuanan This species only flowered once during the study period and, hence, a potential abiotic trigger could not be identified. Visual inspection of the data revealed that this flowering event did occur four months after a drop in the four-day rolling minimum temperature below 20 C, but flowering did not occur on the other two occasions when such temperature drops occurred and the sample size of only one stem precludes assessment of whether this is a true relationship or mere coincidence.

Phenology Paper 2

Page 50


Potential Triggers Summary Tuanan As a result of this species only flowering once during the study period, no potential abiotic triggers of flowering could be identified. This species did not produce fruit during the study period.

Phenology Paper 2

Page 51


Family Latin Name Local name

: Dipterocarpaceae : Shorea balangeran : Kahui / balangeran (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 2 2 2 2 100 2 100 0 / 100

Tuanan 0 0 0 NA NA NA NA NA

1 > 10,001 22 Irregular

NA NA NA NA

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 52


Timing of New Flowering and Fruiting Events During Each Month Sabangau 100 % stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Jun-Jul 08

Sep-Oct 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Dec 06-Jan

Mar-Apr 07

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Jun-Jul 05

Sep-Oct 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Dec 03-Jan

Mar-Apr 04

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Sabangau No predictive model remained significant following correction for multiple comparisons. The best trend model that could be constructed revealed a positive relationship between mean daily rainfall and percentage new flowering events, with a three-month lag (R2 = 0.125, df = 55, p = 0.008); three out of five flowering events occurred three months after months with mean daily rainfall greater than 17 mm.

Phenology Paper 2

Page 53


Potential Triggers of Fruiting Sabangau The strongest predictive model contained mean minimum temperature as the independent variable, operating with a seven-month lag (R2 = 0.175, df = 55, p = 0.001): all four fruiting events occurred seven months after months with mean minimum temperature ≤ 21.5 C.

Potential Triggers Summary Sabangau Flowering in this species appears more likely to be related to mean daily rainfall than any other abiotic variable, with the probability of flowering highest three months after months with mean daily rainfall greater than 17 mm, though this relationship is questionable statistically. The probability of fruiting was highest seven months after months with mean minimum temperature less than 21.5 C. This does not disagree with the flowering vs. fruiting graph.

Percentage occurrence

Fruiting Duration 100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Page 54


Crop Size

Percentage occurrence

70 60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 55


Family: Latin Name: Local name:

Dipterocarpaceae Shorea parvifolia Meranti daun kecil (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 0 0 0 NA NA NA NA NA

Tuanan 36 9 72 41 57 1 1 0 / 100

NA NA NA NA

1 11-100 11 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 56


Timing of New Flowering and Fruiting Events During Each Month Tuanan

% stems new fruit/flower event

45 40 35 30 25 20 15 10

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

5

Month Fruit

Flower

Potential Triggers of Flowering Tuanan The only significant regression model that could be built to predict flowering in this species included the first month of an El Ni単o event, operating with a six-month lag (R2 = 0.112, df = 69, p = 0.005): flowering was more likely six months after the start of an El Ni単o event.

Potential Triggers of Fruiting Tuanan This species only fruited once during the study period and, hence, no potential abiotic trigger could be identified.

Phenology Paper 2

Page 57


Potential Triggers Summary Tuanan Flowering appears to respond to El Ni単o events, with increased likelihood of flowering half a year after the start of an event. No potential abiotic trigger could be identified for fruiting.

Percentage occurrence

Fruiting Duration

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Tuanan

Crop Size

Percentage occurrence

100 80 60 40 20 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Tuanan

Phenology Paper 2

Page 58


Family Latin Name Local name

: Dipterocarpaceae : Shorea teysmanniana : Meranti semut (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 57 11 149 43 29 9 6 0 / 63

Tuanan 0 0 0 NA NA NA NA NA

1 11-100 7 Irregular

NA NA NA NA

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 59


Timing of New Flowering and Fruiting Events During Each Month Sabangau

% stems new fruit/flower event

70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Jun-Jul 08

Sep-Oct 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Jun-Jul 04

Sep-Oct 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Sabangau No predictive model remained significant following correction for multiple comparisons. The trend model with the strongest predictive value contained mean 24-h temperature (R2 = 0.109, df = 68, p = 0.006), which showed a positive relationship with a four-month lag, but which had wide scatter and was not appear useful for predictive purposes. Potential Triggers of Fruiting Sabangau The best model of new fruiting events contained mean 24-h temperature (same month, R2 = 0.158, df = 70, p = 0.001), which showed a positive relationship.

Phenology Paper 2

Page 60


Potential Triggers Summary Sabangau Fruiting patterns in this species are most closely related to 24-h temperature, with the probability of high-percentage new fruiting events increasing in tandem with temperature. It is unclear whether this relationship agrees with the flowering vs. fruiting graph.

Percentage occurrence

Fruiting Duration

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Crop Size

Percentage occurrence

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 61


Family Latin Name Local name

: Dipterocarpaceae : Vatica mengachopai : Rasak napu (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 1 <1 4 3 75 2 50 0 / 100

Tuanan 0 0 0 NA NA NA NA NA

2.5 101-10,000 19 Irregular

NA NA NA NA

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 62


Timing of New Flowering and Fruiting Events During Each Month Sabangau 100

% stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Dec 06-Jan

Mar-Apr 07

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Jun-Jul 05

Sep-Oct 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Sabangau No potential abiotic triggers of flowering could be identified for this species. New flowering events were relatively rare, but occurred at inconsistent times of year, preventing accurate prediction of the timing of fruiting events. Potential Triggers of Fruiting Sabangau No statistically significant models for potential abiotic triggers of fruiting could be produced. The only trend model that could be identified included mean daily rainfall, operating with a lag of eight months (R2 = 0.065, df = 62, p = 0.044). This relationship is weak and not practically useful. Potential Triggers Summary Sabangau No potential abiotic triggers of flowering or fruiting could be identified, as phenological events occurred at unpredictable times of year.

Phenology Paper 2

Page 63


Percentage occurrence

Fruiting Duration

30 25 20 15 10 5 0 1

2

3

4

5

6

8

7

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 64


Family Latin Name Local name

: Ebenaceae : Diospyros siamang : Ehang (SA), pinding pandan (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 3 <1 16 8 50 8 50 63 / 100

Tuanan 3 1 7 7 100 5 71 0 / 67

3 101-1,000 21 Year-round

1.5 1-100 10 Regular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 65


Timing of New Flowering and Fruiting Events During Each Month Sabangau 100

% stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Jun-Jul 07

Sep-Oct 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Tuanan

50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

60

Month Fruit

Flower

Potential Triggers of Flowering Sabangau No predictive models remained significant following correction for multiple comparisons. The trend with the highest predictive value included the four-day rolling mean maximum temperature > 31 C (R2 = 0.096, df = 71, p = 0.008), where the percentage of new flowering events decreased when rolling maximum temperature reached above 31 C. The fact that this model was not significant and that rolling maximum temperature did not exceed 31 C in most months, makes this relationship impractical for predictive purposes. The lack of predictive model found is presumably because this species flowers throughout the year in Sabangau.

Phenology Paper 2

Page 66


Tuanan No significant predictive models could be built for this species. Potential Triggers of Fruiting Sabangau The only significant predictive models that could be built contained the four-day rolling maximum temperature > 31 C, operating with a lag period of seven months (R2 = 0.133, df = 60, p = 0.004). This model implies that the probability of high-percentage new fruiting events is greater seven months following such a temperature spike. This relationship is of questionable relevance, however, as this species fruits virtually continually throughout the year in Sabangau and so fruiting cannot be triggered by temperature spikes of this nature, which occurred during only five months of the study period. The lack of useful predictive relationship detected is presumably a result of this species year-round fruiting habits in Sabangau.

Tuanan No significant predictive models could be built for this species. Potential Triggers Summary Sabangau This species flowers and fruits year round in Sabangau and, hence, no abiotic variables could be identified as potential triggers. Tuanan No abiotic variables could be identified as potential triggers for fruiting of flowering in this species, and inspection of flowering vs. fruiting graph does not reveal any clear time lag between flowering and fruiting.

Phenology Paper 2

Page 67


Percentage occurrence

Fruiting Duration

60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 68


Family Latin Name Local name

: Elaeocarpaceae : Elaeocarpus mastersii : Mangkinang (SA), mangkinang blawau (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 6 <1 16 10 63 4 25 0 / 50

Tuanan 15 0 30 28 93 12 40 0 / 60

1 101-1,000 9 Irregular

2 101-1,000 10 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 69


Timing of New Flowering and Fruiting Events During Each Month Sabangau

% stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Jun-Jul 08

Sep-Oct 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Tuanan

60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

70

Month Fruit

Flower

Potential Triggers of Flowering Sabangau Only one significant result emerged from the screening test (rainfall with a three-month lag), but this was not significant when subjected to regression analysis. Thus, no potential trigger could be identified for this species. This is not surprising upon inspection of the graph of the timing of new flowering events, which shows the three events each happening at different times of the year.

Phenology Paper 2

Page 70


Tuanan The only abiotic variable that predicted could be identified as a potential trigger for flowering was two consecutive months of drought operating with a six-month lag (R2 = 0.113, df = 69, p = 0.004).

Potential Triggers of Fruiting Sabangau The abiotic variable identified as the most likely potential trigger for fruiting in this species was the four-day rolling maximum temperature > 31 C, operating with a eight-month lag (R2 = 0.118, df = 64, p = 0.005): three out of four fruiting events occurred eight months after such temperature spikes. This relationship was weak, however, and two of the fruiting events occurred very close to the median value, so this relationship is probably not useful for predictive purposes.

Phenology Paper 2

Page 71


Tuanan

Fruiting was virtually equally well predicted by the onset of the wet season, with a four-month lag (R2 = 0.102, df = 68, p = 0.007), and the onset of the dry season, with an eight-month lag (R2 = 0.101, df = 67, p = 0.008). These two months often coincide and, thus, it is difficult to establish which is most likely to be the trigger.

Potential Triggers Summary Sabangau The three new flowering events observed occurred at unpredictable times of year and, hence, it was not possible to isolate any useful potential abiotic triggers for flowering in this species. Fruiting appears most likely to be triggered by spikes in temperature > 31 C for four days, with fruit appearing eight months after such spikes.

Phenology Paper 2

Page 72


Tuanan Flowering in this species appears to be related to two consecutive months of drought, operating with a six-month lag period. Fruiting was equally and most closely related to the onset of the wet season, with a four-month lag, and the onset of the dry season, with an eight-month lag.

Percentage occurrence

Fruiting Duration

70 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 73


Family Latin Name Local name

: Fabaceae / Leguminosae : Koompasia malaccensis : Bungaris / kempas (SA), bengaris (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 6 4 13 5 38 5 38 0 / 25

Tuanan 12 8 23 12 52 3 13 0 / 33

1 101-10,000 43 Irregular

1 1-10 21 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 74


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 90

80

70

60

50

40

30

20

10

Month

Flower

Tuanan

18

16

14

12

10

8

6

4

2

Month

Flower

Potential Triggers of Flowering

Sabangau The model with the highest predictive value contained minimum visibility (same month), but the best fit line produced from this was very inconclusive. The predictive value of the model based on the presence of an ENSO event was only very slightly lower (R2 = 0.145, df = 71, p = 0.001) and is much more useful in practical terms.

Page 75


Tuanan The onset of an ENSO event, operating with a three-month lag, was the best predictor of flowering in this species (R2 = 0.181, df = 69, p < 0.001), although inspection of the data reveals that many flowering events were clearly not related to ENSO events. Nevertheless, this remains the best available predictor of flowering in this species.

Potential Triggers of Fruiting Sabangau The most powerful model was based on the mean rolling maximum temperature exceeding 31 C with a three-month lag. While potentially useful, the model including the presence of an ENSO event was only slightly less powerful (R2 = 0.145, df = 71, p = 0.001) and practically more useful: fruiting was much higher during El Ni単o events.

Phenology Paper 2

Page 76


Tuanan The only significant regression model for predicting the timing of fruiting included the spring equinox, operating with a four-month lag ( R2 = 0.192, df = 68, p < 0.001). This model held in two out of three of the fruiting events that occurred.

Potential Triggers Summary Sabangau The flowering and fruiting of this species both appear to respond to ENSO events, and it is likely that, for fruiting at least, the actual abiotic trigger may be continual high temperatures over a period of several days, which is generally associated with ENSO events. This relationship with El Ni単o is not surprising, considering that this species is listed by Appanah (1985) as participating in SAMF events in Pasoh, mainland Malaysia. This agrees with the flowering vs. fruiting graph.

Phenology Paper 2

Page 77


Tuanan As in Sabangau, flowering was best predicted by the onset of an ENSO event, operating with a three-month lag, although fruiting was best related to the spring equinox, operating with a four-month lag. Fruiting Duration

Percentage occurrence

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

2

1

0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 78


Family Latin Name Local name

: Hypericaceae : Cratoxylum spp. (arborescens and glaucum combined) : Geronggang / geronggang merah (SA), garunggang / mipa (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 36 6 55 31 56 27 49 7 / 89

Tuanan 12 <3 23 18 78 1 4 0/5

1 101-1,000 7 Regular / irregular

2.5 11-1,000 37 Irregular / never fruited

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 79


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 90

80

70

60

50

40

30

20

10

Month

Flower

Tuanan

90

80

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The regression model including spring equinox (same month) was very powerful (R2 = 0.543, df = 71, p < 0.001), with much higher incidence of new flowering events occurring in the month of the spring equinox.

Page 80


Tuanan As in Sabangau, a regression model including spring equinox (same month) had high predictive power, with higher incidence of new flowering events in the month of the spring equinox (R2 = 0.283, df = 69, p < 0.001). This was the only significant predictive model for flowering at this site.

Potential Triggers of Fruiting Sabangau Similar to flowers, the best regression models that could be built for this species contained the spring equinox as the independent variable (same month and lag one month combined, R2 = 0.214, df = 70, p < 0.001).

Phenology Paper 2

Page 81


Tuanan The regression model including the spring equinox (operating with a one-month lag) also had highest predictive power for fruiting (R2 = 0.313, df = 68, p < 0.001). The two occasions when this species came into fruit during the study period coincided with the month following the spring solar zenith, which was also one month after the species had come into flower.

Potential Triggers Summary Sabangau Flowering and fruiting both appear to be very strongly related to the spring equinox, with new flowering events coinciding with the solar zenith and new fruiting events following 0-1 month later. This agrees with the fruiting vs. flowering graph.

Phenology Paper 2

Page 82


Tuanan As in Sebangau, flowering and fruiting appears to be related to the spring equinox, with new flowering events coinciding with the zolar zenith and new fruits appearing one month later. This agrees with the flowering vs. fruiting graph.

Percentage occurrence

Fruiting Duration

80 70 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 83


Family Latin Name Local name

: Hypericaceae : Cratoxylum arborescens : Geronggang (SA), garunggang (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 12 4 22 11 50 10 45 9 / 40

Tuanan 1 <1 2 1 50 0 0 NA

1 101-10,000 15 Regular

NA NA NA Never fruited

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 84


0

Phenology Paper 2 Oct-Nov 03

Fruit

100

90

80

70

60

50

40

30

20

10 Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jan-Feb 06

Dec 05-Jan 06

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Nov-Dec 05

Sep-Oct 08

Sep-Oct 05

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Oct-Nov 05

Aug-Sep 05

Jul-Aug 05

Jun-Jul 05

May-Jun 05

Sep-Oct 07

Tuanan

Apr-May 05

Flower

Jun-Jul 07

Month

Mar-Apr 05

Feb-Mar 05

Fruit

Jan-Feb 05

Dec 04-Jan 05

Nov-Dec 04

Oct-Nov 04

Sep-Oct 04

Aug-Sep 04

Jul-Aug 04

Jun-Jul 04

May-Jun 04

Apr-May 04

Mar-Apr 04

Feb-Mar 04

Jan-Feb 04

Dec 03-Jan 04

Nov-Dec 03

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 100

90

80

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The regression model including spring equinox (same month) was very powerful (R2 = 0.592, df = 71, p < 0.001), with much higher incidence of new flowering events in the month of the spring equinox.

Page 85


Tuanan Possibly as a result of the only two stems in the plots dying mid-way through the study period, only one flowering event occurred and, hence, it was impossible to identify a potential abiotic trigger for flowering in this species at Tuanan. Potential Triggers of Fruiting Sabangau The regression model including spring equinox (same month and lag one month) was fairly powerful (R2 = 0.240, df = 70, p < 0.001), with much higher incidence of new fruiting events surrounding the spring equinox.

Tuanan This species only flowered once during the study period and did not produce fruit and, hence, it was impossible to identify potential abiotic triggers. This is probably due to the death of the only two stems of this species in the plots mid-way through the study period.

Phenology Paper 2

Page 86


Potential Triggers Summary Sabangau Flowering and fruiting both appear to be very strongly related to the spring equinox, with new flowering events coinciding with the solar zenith and new fruiting events following 0-1 month later. This agrees with the fruiting vs. flowering graph. Tuanan This species did not fruit during the study period.

Percentage occurrence

Fruiting Duration

80 70 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 87


Family Latin Name Local name

: Hypericaceae : Cratoxylon glaucum : Geronggang merah (SA), mipa (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 24 2 33 20 61 17 52 6 / 82

Tuanan 11 2 21 17 81 1 5 0 / 100

1 101-1,000 7 Irregular

2.5 11-1,000 37 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 88


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 80

70

60

50

40

30

20

10

Month

Flower

Tuanan

90

80

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The regression model including spring equinox (same month) was very powerful (R2 = 0.454, df = 71, p < 0.001), with much higher incidence of new flowering events in the month of the spring equinox.

Page 89


Tuanan Flowering was very well predicted by the regression model including spring equinox (same month, R2 = 0.270, df = 69, p < 0.001), with much higher incidence of new flowering events occurring in the month of solar zenith.

Potential Triggers of Fruiting Sabangau The best regression model included spring equinox (same month and lag one month, R2 = 0.169, df = 70, p < 0.001), with a higher incidence of new flowering events in the month of the spring equinox.

Phenology Paper 2

Page 90


Tuanan The regression model including the spring equinox (operating with a one-month lag) also had highest power for flowering (R2 = 0.313, df = 68, p < 0.001). The only two times this species came into fruit during the study period coincided with the month following spring solar zenith, which was also one month after the species had come into flower.

Potential Triggers Summary Sabangau Flowering and fruiting both appear to be very strongly related to the spring equinox, with new flowering events coinciding with the solar zenith and new fruiting events following 0-1 month later. This agrees with the fruiting vs. flowering graph.

Phenology Paper 2

Page 91


Tuanan As in Sabangau, flowering and fruiting appears to be related to the spring equinox, with new flowering events occurring during the month of the solar zenith and new fruiting events following one month later. This agrees with the fruiting vs. flowering graph.

Percentage occurrence

Fruiting Duration

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Percentage occurrence

Crop Size

70 60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 92


Family Latin Name Local name

: Lauraceae : Litsea spp. combined : Medang / tampang species (SA), kamehas / tagula / pahawas species (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 28 <4 66 33 50 26 39 22 / 81

Tuanan 14 0 28 27 96 15 54 0 / 75

1 1,001-10,000 7 Regular / irregular

2 11-100 10 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 93


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 70

60

50

40

30

20

10

Month

Flower

Tuanan

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The most powerful model included mean daily rainfall, with a five-month lag (R2 = 0.187, df = 67, p < 0.001), which increased likelihood of flower production following months with high rainfall. In our sample, flowers were produced on each of the seven occasions when rainfall, eight months prior, averaged around 15 mm/day or over.

Page 94


Tuanan The onset of flowering was best predicted by a regression model including mean rainfall, with a four-month lag (R2 = 0.163, df = 69, p = 0.001): the likelihood of flowering increased four months after months with high mean daily rainfall, though flowering also occurred following many months without high rainfall.

Potential Triggers of Fruiting Sabangau The most powerful model contained first month of the dry season as a potential trigger for fruiting, with a two month lag before fruit production (i.e., with a peak in the third month of the dry season, R2 = 0.173, df = 70, p < 0.001).

Phenology Paper 2

Page 95


Tuanan Use of the spring equinox (operating with a one-month lag) as the independent variable produced the only significant predictive model for the onset of fruiting (R2 = 0.241, df = 68, p < 0.001), with higher incidence of fruiting one month after the solar zenith.

Potential Triggers Summary Sabangau The strongest models that emerged from the analysis included mean daily rainfall as a potential trigger for flowering, with a five-month lag, and the start of the dry season as a potential trigger for fruiting, with a two-month lag. This indicates flowering at the peak of the wet season, followed by fruiting in the dry season, though it is possible that, due to the regular flowering and fruiting patterns of this genus, which had both new flowering and fruiting events in most months, these models will be of little practical use.

Phenology Paper 2

Page 96


Tuanan Flowering was best predicted by a regression model including mean daily rainfall, operating with a four-month lag, whereas the spring equinox (operating with a one-month lag) was the most likely abiotic trigger of fruiting. Fruiting Duration

Percentage occurrence

60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 97


Family Latin Name Local name

: Melastomataceae : Pternandra cf. coerulescens/galeata : Kemuning yang bergaris tiga (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 2 <1 3 2 67 2 67 0 / 100

Tuanan 0 0 0 NA NA NA NA NA

1 11-1,000 10 Irregular

NA NA NA NA

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 98


Timing of New Flowering and Fruiting Events During Each Month Sabangau

50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

60

Month Fruit

Flower

Potential Triggers of Flowering Sabangau Flowering in this species appear best predicted by the four-day rolling maximum temperature > 31 C (same month, R2 = 0.118, df = 71, p = 0.003), with flowering only occurring in months where the four-day rolling maximum temperature exceeded 31 C. Due to the very small number of flowering events, it is difficult to tell whether this is a real relationship or merely coincidence.

Phenology Paper 2

Page 99


Potential Triggers of Fruiting Sabangau Statistical analysis of potential abiotic triggers of fruiting was impossible, due to this species fruiting only once during the study period. Inspection of the data revealed that the single fruiting event occurred in the second month of the wet season, one month after a spring equinox and in a month where the four-day rolling maximum temperature exceeded 31 C, though it is impossible to tell which, if either, of these variables was the single trigger for fruiting. Potential Triggers Summary Sabangau Due to the low number of flowering and fruiting events exhibited by this species, it was not possible to identify potential abiotic triggers with any precision. Both flowering events occurred in months where the four-day rolling maximum temperature exceeded 31 C, and the single fruiting event coincided with the second month of the dry season, fell one month after a spring equinox and occurred in a month where the four-day rolling maximum temperature exceeded 31 C.

Percentage occurrence

Fruiting Duration 100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Crop Size

Percentage occurrence

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Page 100


Family Latin Name Local name

: Meliaceae : Aglaia rubiginosa : Kajalaki (SA), kajalaki bawi (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 7 5 18 10 56 9 50 44 / 78

Tuanan 6 2 12 8 67 1 8 0 / 100

2 101-1,000 23 Year-round

1 1-10 24 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 101


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 80

70

60

50

40

30

20

10

Month

Flower

Tuanan

80

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The best predictive model for the onset of flowering was identified post-hoc and contained the first two months of the dry season as the independent variable (R2 = 0.215, df = 71, p < 0.001).

Page 102


Tuanan The best predictive model for the onset of flowering contained the autumn equinox as the independent variable, operating with a two-month lag (R2 = 0.172, df = 69, p < 0.001).

Potential Triggers of Fruiting Sabangau The best predictive model was produced using the first month of dry season with a twomonth lag as the independent variable (R2 = 0.254, df = 69, p < 0.001).

Phenology Paper 2

Page 103


Tuanan Statistical analysis of potential abiotic triggers of fruiting was impossible, due to this species fruiting only once during the study period and this single fruiting event comprising only one stem. Inspection of the data revealed that this fruiting event occurred one month after a flowering event of identical magnitude and, thus, it is most likely that fruiting in this species merely follows flowering with a lag of about one month, though all flowering events do not turn into fruiting events at this site. Potential Triggers Summary Sabangau Species flowering is probably triggered by the start of the dry season, with the occurrence of new flowering events peaking in the first two months of the dry season. Fruiting then follows one or two months later, in the third month of the dry season. This agrees with the graph of fruiting vs. flowering for this species. Tuanan Flowering seems to follow two months after the autumn equinox. As only one fruiting event occurred during the study period, no potential trigger of fruiting could be identified. This agrees with the fruiting vs. flowering graph.

Phenology Paper 2

Page 104


Percentage occurrence

Fruiting Duration

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Percentage occurrence

Crop Size

100 80 60 40 20 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 105


Family Latin Name Local name

: Myristicaceae : Horsfieldia crassifolia : Mendarahan daun besar (SA), kumpang daun perak (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 73 11 123 90 73 55 45 7 / 62

Tuanan 13 2 25 19 76 10 40 20 / 70

1 101-1,000 9 Irregular

2 11-100 11 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 106


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 45

40

35

30

20 25

15

10

5

Month

Flower

Tuanan

80

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The best predictive model contained mean minimum temperature with a lag of two months (R2 = 0.316, df = 70, p < 0.001). This produced a negative relationship between the two variables, indicating that the probability of a high-percentage flowering event diminishes following months with high mean minimum temperatures.

Page 107


Tuanan The model with the highest predictive power for flowering contained four months of consecutive drought (R2= 0.223, df = 69, p < 0.001), but, although flowering occurred during most droughts, it also occurred during non drought months.

Potential Triggers of Fruiting Sabangau The best predictive model contained mean minimum temperature with a lag of four months (R2 = 0.314, df = 68, p < 0.002). This produced a negative relationship between the two variables, again indicating that the probability of a high-percentage fruiting event diminishes following months with high mean minimum temperatures.

Phenology Paper 2

Page 108


Tuanan No significant predictive model could be built for fruiting in this species at Tuanan and, hence no potential trigger could be identified. This is probably a result of this species’ regular fruiting during different times of year. Potential Triggers Summary Sabangau The flowering and fruiting patterns of this species are best related to mean minimum temperature, which operate with a two and four month lag, respectively: the probability of high-percentage flowering and fruiting events is lower following months with high mean minimum temperature. This does not disagree with the flowering vs. fruiting graph. Tuanan Flowering in this species was best predicted by four consecutive months of drought; no potential abiotic trigger could be identified for fruiting. This does not disagree with the flowering vs. fruiting graph. Inspection of the flowering vs. fruiting graph also did not reveal any specific time lag between flowering and fruiting events.

Phenology Paper 2

Page 109


Fruiting Duration

Percentage occurrence

70 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 110


Family Latin Name Local name

: Myrtaceae : Eugenia cf. spicata : Kayu lalas (SA)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 4 <1 10 2 20 0 0 NA

Tuanan 0 0 0 NA NA NA NA NA

NA NA NA Never fruited

NA NA NA NA

Timing of New Flowering and Fruiting Events During Each Month Sabangau

50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Jun-Jul 06

Sep-Oct 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

60

Month Fruit

Flower

Potential Triggers of Flowering Sabangau Only two flowering events occurred during the study period at completely different times of year: one in Nov-Dec 2003 and one in Apr-May 2006. Unsurprisingly, therefore, no independent variables were significant in the screening tests and no potential flowering triggers could be identified. Potential Triggers of Fruiting Sabangau This species did not come into fruit during the study period.

Phenology Paper 2

Page 111


Potential Triggers Summary Sabangau Due to flowering events occurring at completely different times of year and a lack of fruiting events during the study period, no potential abiotic triggers of flowering/fruiting could be identified.

Phenology Paper 2

Page 112


Family Latin Name Local name

: Myrtaceae : Syzygium spp. combined (cf. garcinifolia and havilandii) : Jambu burung / tatumbu (SA), tatumbu kasar / tatumbu putih (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 36 <8 86 51 59 41 48 39 / 75

Tuanan 19 5 36 35 97 25 69 6 / 45

2 1,001-10,000 8 Frequent / regular

2 11-100 10 Frequent / irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 113


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 70

60

50

40

30

20

10

Month

Flower

Tuanan

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The most powerful predictive model contained day length (same month) as the potential trigger (R2 = 0.325, df = 70, p < 0.001), with decreasing day length associated with increased likelihood of flowering.

Page 114


Tuanan Flowering was predicted equally powerfully by two different regression models. The first included three consecutive months of drought, operating with a five-month lag, and the second included four consecutive months of drought, operating with a four-month lag (R2 = 0.252, df = 69, p < 0.001). In both cases the likelihood of flowering was higher following the prolonged period of drought. These two independent variables are essentially the same, however, and, thus, it is impossible to determine which is the most likely trigger.

Potential Triggers of Fruiting Sabangau The most powerful predictive model contained day length as the potential trigger (R2 = 0.303, df = 70, p < 0.001), with decreasing day length two months previously being associated with increased likelihood of flowering.

Phenology Paper 2

Page 115


Tuanan The best predictive model that could be constructed revealed a negative relationship between mean daily rainfall and percentage of new fruiting events, with a two-month lag (R2 = 0.207, df = 68, p < 0.001).

Potential Triggers Summary Sabangau Strong relationships exist between day length and flowering and fruiting patterns in this species, with decreased day length associated with increased probability of flowering and fruiting events, with lags of zero and two months, respectively. This agrees with the flowering vs. fruiting graph.

Phenology Paper 2

Page 116


Tuanan A prolonged period of drought (three or four months) is the most likely abiotic trigger for flowering, with flowers appearing after five or four months, respectively. Fruiting appears best related to decreased rainfall two months previously. This agrees with the flowering vs. fruiting graph. Fruiting Duration

Percentage occurrence

60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14

17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Percentage occurrence

Crop Size

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 117


Family Latin Name Local name

: Myrtaceae : Tristaniopsis spp. : Blawan spp. (SA), balawan spp. (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 10 1 31 10 32 7 23 0 / 57

Tuanan 16 4 31 19 61 8 26 0 / 63

2 101-10,000 12 Irregular

1.5 1-100 11 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 118


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 60

50

40

30

20

10

Month

Flower

Tuanan

35

30

25

20

15

10

5

Month

Flower

Potential Triggers of Flowering

Sabangau The most powerful statistical model produced included the four-day rolling maximum temperature > 31 C, with a six-month lag (R2 = 0.140, df = 66, p = 0.002).

Page 119


Tuanan The best predictive model for flowering was produced using the onset of the wet season as the independent variable, which produced a positive relationship, operating with a threemonth lag (R2 = 0.181, df = 69, p < 0.001).

Potential Triggers of Fruiting Sabangau The most powerful statistical model produced included the four-day rolling maximum temperature > 31 C, with a eight-month lag (R2 = 0.141, df = 64, p = 0.002). Tuanan The most powerful regression model predicted that fruiting was most likely to occur five months after the spring equinox (R2 = 0.233, df = 68, p < 0.001).

Phenology Paper 2

Page 120


Potential Triggers Summary Sabangau The most likely potential candidate for an abiotic trigger of flowering and fruiting was the four-day rolling maximum temperature exceeding 31 C, which led to higher new flowering and fruiting probabilities six and eight months later, respectively. Tuanan The percentage of flowering stems was found to be highest three months after the onset of the wet season, whereas fruiting was best predicted by the spring equinox, operating with a five-month delay. Fruiting Duration

Percentage occurrence

60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13

14 17 18 19

20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Tuanan

Page 121


Percentage occurrence

Crop Size

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 122


Family Latin Name Local name

: Sapotaceae : Madhuca mottleyana : Katiau (SA/TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 9 2 18 9 50 8 44 38 / 88

Tuanan 4 3 8 6 75 5 63 0 / 80

2 101-1,000 20 Frequent

2 11-100 11 Regular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 123


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 100

90

80

70

60

50

40

30

20

10

Month

Flower

Tuanan

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The most powerful model included the autumn equinox as a potential trigger, with a fivemonth lag (R2 = 0.204, df = 71, p < 0.001).

Page 124


Tuanan The most powerful regression model for flowering in this species included mean daily rainfall (positive relationship), operating with a six-month lag (R 2 = 0.195, df = 69, p < 0.001). Inspection of the resulting scatter plot reveals that this relationship is far from perfect, however, with many flowering events (including the highest percentage event) occurring following months with rainfall close to/below average. The next best regression model included spring equinox (same month, R2 = 0.103, df = 69, p = 0.007), which may be more useful for practical purposes.

Phenology Paper 2

Page 125


Potential Triggers of Fruiting Sabangau The most likely potential abiotic trigger of fruiting identified for this species was the start of the dry season, operating with a lag of five months (R2 = 0.241, df = 67, p < 0.001).

Tuanan The best regression model that could be built for this species contained the spring equinox as the independent variable, operating with a five-months lag ( R2 = 0.125, df = 68, p = 0.003). There was a much higher likelihood of flowering five months after the solar zenith, compared to other months.

Phenology Paper 2

Page 126


Potential Triggers Summary Sabangau The autumn equinox would appear to be the trigger for flowering in this species, with flowering occurring five months after the overhead passage of the sun (i.e., in AugustSeptember). The fifth month following the autumn equinox generally falls in the fourth of fifth month of the dry season; this may be expected, considering the delicate nature of this species’ flowers, which may be easily dislodged by heavy rain. New fruiting events then peak one month later, in the sixth month of the dry season. Thus, it is probable that there is no real trigger for fruiting in this species, and that fruiting merely follows a month or two after flowering, which is triggered by the autumn equinox. This agrees with the flowering vs. fruiting graph. Tuanan The most powerful predictive model for flowering included mean daily rainfall (six-months lag) as the dependent variable, but, although less powerful, the spring equinox may be a more useful predictor of flowering. This latter model also agrees with the results for Sabangau; i.e., flowering in Tuanan occurs one month later than flowering in Sabangau, which hints that the species may use the same abiotic trigger for flowering at the two sites. Fruiting was best predicted by the spring equinox, operating with a five month lag, and inspection of the flowering vs. fruiting graph shows that fruiting indeed occurs about five months after flowering, supporting for the notion that spring equinox is the most likely trigger for flowering in this species.

Phenology Paper 2

Page 127


Fruiting Duration

Percentage occurrence

35 30 25 20 15 10 5 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 128


Family Latin Name Local name

: Sapotaceae : Palaquium coclearifolium : Nyatoh gagas (SA), nyatoh undus daun besar (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 6 <1 18 9 50 8 44 25 / 100

Tuanan 7 2 13 13 100 5 38 0 / 50

2 101-1,000 10 Irregular

2.5 11-100 17 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 129


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 60

50

40

30

20

10

Month

Flower

Tuanan

45

40

35

30

25

20

15

10

5

Month

Flower

Potential Triggers of Flowering

Sabangau The best model included the first month of the wet season, with a one month lag (R2 = 0.123, df = 71, p = 0.003).

Page 130


Tuanan Flowering was best predicted be a regression model including the autumn equinox (threemonth lag), with a higher incidence of flowering following the autumn equinox (R2 = 0.248, df = 69, p < 0.001).

Potential Triggers of Fruiting Sabangau No model remained significant following correction for multiple comparisons. The strongest trend included the third and fourth months of the wet season (R2 = 0.073, df = 70, p = 0.023).

Phenology Paper 2

Page 131


Tuanan The best regression model that could be built for fruiting in this species contained the spring equinox as the independent variable, operating with a one-month lag (R2 = 0.27, df = 68, p < 0.001).

Potential Triggers Summary Sabangau Both flowering and fruiting in this species would appear to be determined by the start of the wet season: with predictable flowering occurring two months into the wet season and a predictable fruiting event one to two months later (though this latter relationship is questionable statistically). Tuanan Flowering appeared to occur most likely three months after the autumn equinox and fruiting one month after the spring equinox. Fruiting therefore seems most likely to occur four months after flowering. Inspection of the flowering vs. fruiting graph confirms this finding.

Phenology Paper 2

Page 132


Fruiting Duration

Percentage occurrence

50 40 30 20 10 0 1

2

3

4

5

6

8

7

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Crop Size

Percentage occurrence

70 60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 133


Family Latin Name Local name

: Sapotaceae : Palaquium leiocarpum : Hangkang (SA/TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 127 32 293 106 36 101 34 51 / 87

Tuanan 15 10 30 30 100 24 80 48 / 92

2 101-1,000 6 / 20* Frequent

3 11-100 10 Frequent

Variation in Fruit Availability Between Months

* Minimum DBH producing fruit 6 cm, but of 1,880 observations of fruiting in this species, 1,835 (98%) were in trees over 20 cm DBH

Phenology Paper 2

Page 134


Timing of New Flowering and Fruiting Events During Each Month Sabangau

60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Jun-Jul 04

Sep-Oct 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

70

Month Fruit

Flower

Tuanan

% stems new fruit/flower event

80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Sabangau No predictive model remained significant following correction for multiple comparisons. The model with the strongest trend contained the first month of the dry season as the independent variable; unsurprisingly the predictive power of this model was low (R2 = 0.066, df = 71, p = 0.030). Again, this is most likely due to the relatively consistent flowering pattern of this species.

Phenology Paper 2

Page 135


Tuanan The only significant regression model that could be built for flowering included two consecutive months of drought, operating with a three-month lag period (R2 = 0.148, df = 69, p = 0.001).

Potential Triggers of Fruiting Sabangau The most powerful predictive model contained mean 24-h temperature (same month, R2 = 0.118, df = 70, p = 0.003), with high-percentage fruiting events being more likely during months with a relatively low mean 24-h temperature.

Tuanan The only significant regression model predicting fruiting events contained the onset of drought, with increased probability of fruiting three months after the first drought month (R2 = 0.209, df = 68, p < 0.001).

Phenology Paper 2

Page 136


Potential Triggers Summary Sabangau The power of the predictive models for potential abiotic triggers were not high, as a consequence of the relatively consistent flowering and fruiting patterns in this species. No reliable abiotic trigger of flowering could be identified, whereas the probability of a new fruiting event is highest in months with low mean 24-h temperature. This does not disagree with the flowering vs. fruiting graph. Tuanan The onset of drought appeared to be associated with both flowering and fruiting in this site, with responses occurring with a three-month delay. Despite these statistical results, inspection of the flowering vs fruiting graph shows that flowering and fruiting do not always occur at the same time. Consequently, it is possible that these potential triggers are not real and are merely coincidental, due to the frequent and year-round flowering and fruiting patterns of this species. Fruiting Duration

Percentage occurrence

50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Tuanan

Page 137


Percentage occurrence

Crop Size

60 50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 138


Family Latin Name Local name

: Sapotaceae : Palaquium pseudorostratum : Nyatoh babi (SA), nyatoh puntik (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 17 2 31 12 39 7 23 43 / 100

Tuanan 34 11 67 56 84 42 63 59 / 82

1 101-1,000 10 Frequent

3 11-100 10 Frequent

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 139


0

Phenology Paper 2 Fruit Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Fruit

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

Sep-Oct 03

0

Sep-Oct 03

% stems new fruit/flower event

% stems new fruit/flower event

Timing of New Flowering and Fruiting Events During Each Month

Sabangau 90

80

70

60

50

40

30

20

10

Month

Flower

Tuanan

80

70

60

50

40

30

20

10

Month

Flower

Potential Triggers of Flowering

Sabangau The only significant predictive model that could be built contained day length as the potential trigger, with a lag of two months (R2 = 0.120, df = 71, p = 0.003): the probability of high percentage new flowering events was lower following months with long day length.

Page 140


Tuanan No significant predictive models could be built and no potential flowering identified for this species. This is probably due to the frequent and year-round flowering of this species. Potential Triggers of Fruiting Sabangau The potential abiotic variable most likely to trigger fruiting in this species is the start of the wet season, which produced the highest predictive power with a three-month lag (i.e., with a peak in the fourth month of the wet season, R2 = 0.164, df = 68, p = 0.001). The second best model contained the first month of the wet season with a one-month lag, yet, interestingly, there was no effect with a two-month lag, and very few new fruiting events occurred in the third month of the wet season, presumably because many stems still carried fruit from the second month of the wet season.

Phenology Paper 2

Page 141


Tuanan Fruiting was best predicted by the regression model including the spring equinox as the independent variable, operating with a one-month lag (R2 = 0.196, df = 68, p < 0.001), with increased likelihood of fruiting one month after the solar zenith.

Potential Triggers Summary Sabangau The onset of flowering in this species appears related to day length, with peaks in flowering being least likely two months after months with the highest day length (i.e., around February). Fruiting appears to be triggered by the onset of the wet season, with fruiting peaking two to four months into the wet season (the percentage of new fruiting events is low in the third month of the wet season, as many stems still carry fruit from the second month). This agrees with the flowering vs. fruiting graph. Tuanan No potential abiotic trigger could be identified for flowering in this species, though fruiting was most likely to occur one month after the spring equinox. This agrees with the flowering vs. fruiting graph.

Phenology Paper 2

Page 142


Fruiting Duration

Percentage occurrence

60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Percentage occurrence

Crop Size

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 143


Family Latin Name Local name

: Sapotaceae : Palaquium ridleyii / cf. xanthochymum : Nyatoh burung (SA), nyatoh undus buah merah (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 14 5 27 15 56 12 44 25 / 67

Tuanan 11 3 21 19 90 7 33 13 / 43

1 101-1,000 15 Regular

1 11-100 13 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 144


Timing of New Flowering and Fruiting Events During Each Month Sabangau

% stems new fruit/flower event

40 35 30 25 20 15 10

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Jun-Jul 06

Sep-Oct 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

5

Month Fruit

Flower

Tuanan

12 10 8 6 4

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

2

Sep-Oct 03

% stems new fruit/flower event

14

Month Fruit

Flower

Potential Triggers of Flowering Sabangau The most powerful model included drought lasting three months, with a three-month lag, as the potential trigger (R2 = 0.190, df = 69, p < 0.001), with flowering events more following such drought periods. While strongest statistically, a model including the start of the dry season as the potential trigger, operating with a six-month lag, may be more practically useful, as flowering occurred in most years of the study, whereas three-month droughts only occurred on three occasions, and was only marginally less powerful (R2 = 0.174, df = 66, p < 0.001). Following a peak in the sixth month following the onset of the dry season, the percentage of new flowering events drops off smoothly (with the exception of twelve months after the onset of the dry season).

Phenology Paper 2

Page 145


Tuanan The only abiotic variable that could be identified as a potential trigger of flowering in this species was the onset of El Ni単o conditions, with flowering being more likely to occur four months after the beginning of El Ni単o conditions (R2 = 0.1214, df = 69, p = 0.003).

Potential Triggers of Fruiting Sabangau The best model included minimum visibility as a potential trigger (R2 = 0.237, df = 57, p < 0.001), with an increased likelihood of new fruiting events five months after months with low visibility. This relationship is not useful for predictive purposes, however, as fruiting occurred regularly during the study period, whereas hazy periods of visibility only dropped below its usual seven or more kilometres in a small number of months. From visual inspection, the probability of new fruiting events does appear higher eight months following the onset of the dry season, though this was not borne out statistically.

Phenology Paper 2

Page 146


Tuanan No potential abiotic trigger of fruiting could be identified for this species. Potential Triggers Summary Sabangau Flowering in this species appears most likely to be related to the onset of the dry season, with a peak in new flowering events six months after the start of the dry season a gradual decline in the percentage of new flowering events thereafter. Useful statistical relationships to detect potential triggers for fruiting were not forthcoming, but it appears most likely that fruiting may simply follow two months after flowering. This agrees with the flowering vs. fruiting graph. Tuanan Flowering of this species in Tuanan was most closely related to the onset of El Ni単o conditions, with flowering peaking four months later. No potential abiotic trigger of fruiting could be identified, as fruiting in this species does not appear to be concentrated at any particular time of year in Tuanan.

Phenology Paper 2

Page 147


Fruiting Duration

Percentage occurrence

70 60 50 40 30 20 10 0 1

2

3

4

5

6

7

8

9

10

11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Tuanan

Percentage occurrence

Crop Size

50 40 30 20 10 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 148


Family Latin Name Local name

: Tetrameristaceae : Tetramerista glabra : Ponak (SA), tantimun (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 3 1 25 8 32 8 32 33 / 75

Tuanan 2 <1 4 2 50 1 25 0 / 100

2 101-1,000 20 Frequent

1 1-10 17 Regular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 149


Timing of New Flowering and Fruiting Events During Each Month Sabangau

50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Dec 03-Jan

0

Mar-Apr 04

10

Sep-Oct 03

% stems new fruit/flower event

60

Month Flower

Fruit

Tuanan

50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

10

Sep-Oct 03

% stems new fruit/flower event

60

Month Fruit

Flower

Potential Triggers of Flowering Sabangau The most powerful model contained the four-day rolling minimum temperature < 20 C as the independent variable (R2 = 0.256, df = 66, p < 0.001), but this model is not useful for predictive purposes, as it shows a lower likelihood of flowering in months with such a temperature drop, which occur only rarely. The next most useful model contained day length as the potential abiotic trigger, with a lag of two months (R2 = 0.161, df = 70, p = 0.001). Again, although the strength of this relationship was reasonable, the relationship is also of little help in predicting the timing of flowering events, as a result of this species’ relatively consistent flowering patterns.

Phenology Paper 2

Page 150


Tuanan No potential abiotic triggers of flowering could be identified for this species. This is probably a result of the relative scarcity of new flowering events and their occurrence at inconsistent times of year. Potential Triggers of Fruiting Sabangau The only significant regression model included the first month of the dry season (same month, R2 = 0.136, df = 69, p = 0.002) as the independent variable, with higher percentage stems with new fruiting events occurring in the first month of the dry season. The percentage for the third month of the dry season is also relatively high and the low percentage of new fruiting events for the second month of the dry season month is likely a result of stems continuing to bear fruit from the first dry season month. Thus, the first three months of the dry season are the optimum period for searching for the seeds of this species.

Phenology Paper 2

Page 151


Tuanan The single stem of this species only fruited once during the study period, preventing accurate prediction of the timing of fruiting events. This single event coincided with the autumn equinox, though fruiting did not follow the other five autumn equinoxes that occurred during the study period, and none of the other stems of this species fruited during this autumn equinox. It is therefore impossible to determine beyond doubt whether or not the autumn equinox is the trigger for fruiting in this species. Potential Triggers Summary Sabangau This species produces both flowers and fruits relatively consistently, making analysis of potential abiotic triggers difficult. For flowers, it was impossible to identify a potentially useful abiotic trigger. Although fruit production is also relatively consistent, there is generally a peak in fruit availability in the first three months of the dry season. This agrees with the flowering vs. fruiting graph. Tuanan No potential abiotic triggers could be identified, as phenological events occurred at unpredictable times of year. It is possible that fruiting is related to the autumn equinox, but this is based on a single fruiting event in a single stem, so this relationship cannot be confirmed. Fruiting Duration

Percentage occurrence

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Tuanan

Page 152


Percentage occurrence

Crop Size

100 80 60 40 20 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 153


Family Latin Name Local name

: Theaceae : Ploiarium alternifolium : Asam-asam (SA), kayu asem (TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau <1 <1 0 NA NA NA NA NA

Tuanan 1 <1 2 1 50 1 50 0 / 100

NA NA NA NA

2 101-1,000 18 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 154


Timing of New Flowering and Fruiting Events During Each Month Tuanan 100 % stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

Dec 03-Jan

0

Sep-Oct 03

10

Month Fruit

Flower

Potential Triggers of Flowering Tuanan The most powerful regression model found for flowering in this species included mean visibility as the dependent (R2 = 0.197, df = 54, p = 0.001), with increased likelihood of flowering during months of decreased visibility. Although flowering occurred in both periods of normal and low visibility, flowering occurred in four out of six months in which visibility was less than 5 km.

Phenology Paper 2

Page 155


Potential Triggers of Fruiting Tuanan This species only fruited once during the study period and, hence, a potential abiotic trigger cannot be identified with confidence. This single fruiting event occurred three months after a drought period of two months, indicating that this may be a potential trigger, but it is impossible to say whether this result was mere chance. Potential Triggers Summary Tuanan Mean visibility is the most likely potential abiotic trigger for flowering. Fruiting only occurred once during the study period and, thus, a potential abiotic trigger cannot be identified with confidence, though there is a possibility that a drought period of two consecutive months of drought (operating with a three-month lag) could be the trigger for fruiting. Fruiting Duration

Percentage occurrence

120 100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Tuanan

Crop Size

Percentage occurrence

100 80 60 40 20 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Tuanan

Phenology Paper 2

Page 156


Family Latin Name Local name

: Thymelaeaceae : Gonystylus bancanus : Ramin (SA/TU)

Variable Stem density > 10 cm DBH/ha Stem density > 20 cm DBH/ha Total number stems sampled Total number flowering stems Percentage of flowering stems Total number fruiting stems Percentage of fruiting stems Minimum / maximum percentage fruiting stems bearing fruit/year Median fruiting duration (months) Median crop size Minimum DBH bearing fruit (cm) Fruiting regularity category

Sabangau 6 1 37 9 24 2 5 0 / 100

Tuanan 9 1 17 3 18 1 6 0 / 100

1.5 101-1,000 31 Irregular

2 11-100 28 Irregular

Variation in Fruit Availability Between Months

Phenology Paper 2

Page 157


Timing of New Flowering and Fruiting Events During Each Month Sabangau 100 % stems new fruit/flower event

90 80 70 60 50 40 30 20

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Dec 06-Jan

Mar-Apr 07

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Dec 03-Jan

Mar-Apr 04

0

Sep-Oct 03

10

Month Fruit

Flower

Tuanan

30 25 20 15 10

Jun-Jul 09

Mar-Apr 09

Dec 08-Jan

Sep-Oct 08

Jun-Jul 08

Mar-Apr 08

Dec 07-Jan

Sep-Oct 07

Jun-Jul 07

Mar-Apr 07

Dec 06-Jan

Sep-Oct 06

Jun-Jul 06

Mar-Apr 06

Dec 05-Jan

Sep-Oct 05

Jun-Jul 05

Mar-Apr 05

Dec 04-Jan

Sep-Oct 04

Jun-Jul 04

Mar-Apr 04

0

Dec 03-Jan

5

Sep-Oct 03

% stems new fruit/flower event

35

Month Fruit

Flower

Potential Triggers of Flowering Sabangau Only two flowering events occurred during the study period at completely different times of year: Sep-Dec 2003 and Feb-Mar 2006. Although these both occurred around equinox events, no statistical association between these equinox and new flowerring events emerged. No other independent variables were significant in the screening tests and, hence, no potential flowering triggers could be identified.

Phenology Paper 2

Page 158


Tuanan Flowering was best predicted by the regression model including autumn equinox with a lag of three months (R2 = 0.167, df = 69, p < 0.001). Despite this significant result, flowering only coincided with the autumn equinox in four out of seven cases of flowering, and occurred in two of the six autumn equinox months during the study period. Therefore it is far from guaranteed that flowering will follow three months after the autumn equinox.

Potential Triggers of Fruiting Sabangau Only one fruiting event occurred during the study period: in Mar-Apr 2006 (the first flowering event did not lead to any fruit production). Although no independent variables were significant in the screening tests, this event did occur in an autumn equinox month and this was also significant when subjected to regression analysis (R2 = 0.155, df = 70, p = 0.001). Despite this significant result, fruiting only occurred in one out of the six autumn equinoxes that occurred during the study period and so it is far from guaranteed that fruit will follow the autumn equinox.

Phenology Paper 2

Page 159


Tuanan This species only fruited once during the study period and, hence, it was impossible to identify a potential abiotic trigger with confidence. This sole fruiting event occurred seven months following a two-month drought, but it is impossible to establish whether this is a real relationship or merely coincidence. Potential Triggers Summary Sabangau Although no potential abiotic triggers of flowering could be confirmed, it is possible that flowering is timed to enable fruiting to coincide with the autumn equinox, although fruit is not produced in most years. This agrees with the flowering vs. fruiting graph. Tuanan Due to the irregularity of flowering events in this species and only one single fruiting event during the study period, we could not clearly identify a potential abiotic trigger for flowering or fruiting in this species, though flowering does appear most likely to be triggered by the autumn equinox (operating with a three-month lag).

Percentage occurrence

Fruiting Duration

120 100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 17 18 19 20 21 22 23

Fruiting duration (months) Sabangau

Phenology Paper 2

Tuanan

Page 160


Crop Size

Percentage occurrence

100 80 60 40 20 0 1-10

11-100

101-1,000

1,001-10,000

> 10,000

Crop size category Sabangau

Phenology Paper 2

Tuanan

Page 161


Relationships Between Sites for Onset of Flowering and Fruiting The results of Spearman’s correlations between the two sites for the onset of flowering and fruiting (i.e., percentage of stems with new flowering and fruiting events each month) are given in Table 3. Surprisingly, new flowering and fruiting were correlated between the two sites in only 5/18 (28%) and 7/18 (39%) of cases, respectively. Table 3. Spearman’s correlations between percentage stems with new flower and fruit events in the two sites. Family Anacardiaceae Anisophyllaceae Apocynaceae Clusiacea / Guttiferae Clusiacea / Guttiferae

Species Campnosperma coriaceum Combretocarpus rotundatus Dyera lowii / polyphylla Calophyllum hosei Calophyllum nodosum

Clusiacea / Guttiferae Clusiacea / Guttiferae Dipterocarpaceae Dipterocarpaceae Dipterocarpaceae Dipterocarpaceae Ebenaceae Elaeocarpaceae Fabaceae / Leguminosae Hypericaceae Hypericaceae Hypericaceae Lauraceae Melastomataceae Meliaceae Myristicaceae Myrtaceae Myrtaceae Myrtaceae Sapotaceae Sapotaceae Sapotaceae Sapotaceae Sapotaceae Tetrameristaceae Theaceae Thymelaeaceae

Calophyllum sclerophyllum (SA) Calophyllum sclerophyllum (TU) Shorea balangeran Shorea parvifolia Shorea teysmanniana Vatica rassak Diospyros siamang Elaeocarpus mastersii Koompassia malaccensis

Phenology Paper 2

Cratoxylon spp. Cratoxylon arborescens Cratoxylon glaucum Litsea spp. P. cf. coerulescens / galeata Aglaia rubiginosa Horsfieldia crassifolia Eugenia spicata Syzygium spp. Tristaniopsis spp. Madhuca motleyana Palaquium coclearifolium Palaquium leiocarpum Palaquium pseudorostratum Palaquium ridleyii / cf. xanthochymum Tetramerista glabra Ploiarium alternifolium Gonystylus bancanus

New flowers 0.238* NA 0.307** NA NA

New fruit 0.424*** NA 0.250* NA NA

NA NA NA NA NA NA 0.189 0.197 -0.139

NA NA NA NA NA NA 0.001 0.013 NA

0.211 NA -0.150 0.075 NA 0.470*** 0.500*** NA 0.018 0.018 0.216 0.012 0.142 0.451*** -0.179 0.088 NA -0.071

0.362** NA 0.388*** -0.004 NA -0.079 0.333** NA 0.178 0.429*** 0.072 0.050 -0.028 0.258* 0.062 -0.074 NA -0.015

Page 162


DISCUSSION This study presents the first account of species-specific abiotic triggers of flowering and fruiting in PSF trees. We also compare the results of flowering and fruiting phenology, and potential abiotic triggers of these, between two PSF sites, providing insight into potential intra-specific variations between sites. We present novel methods for assessing abiotic triggers of flowering and fruiting (specifically the isolation of the first month of flowering and fruiting events for regression analyses) and, in doing so, provide a useful template for further such studies in PSF and other habitats. Finally, we present information useful for the establishment of seed collection protocols for forest restoration practices. As in our previous report (Harrison et al., 2010), we reiterate the complications in interpretation due to differences in sample sizes between sites, particularly for those species with very low sample size at one site, and from differences in the number and proportion of fruiting stems. More confidence should therefore be placed in fruiting regularity results for those species with larger sample sizes, as there will be lower variability in these estimates and fewer “missed� fruiting events. This is a likely cause behind some of the differences in fruiting classifications for certain species between the two sites; e.g., Aglaia rubiginosa was one of the most consistent fruit producers in Sabangau (consistently producing fruit in 9 months of the year), but produced fruit only rarely in Tuanan (consistently producing fruit in only 1 month of the year). This may be due to either real inter-site difference in the fruiting phenology of these species, or a result of the higher number of fruiting stems (9 vs. 1) sampled in Sabangau compared to Tuanan. Comparison of Timing of Flowering and Fruiting Onset Between Sites As expected, the onset of new flowering and fruiting events were related (i.e., occurred at similar times) between the two sites for a number of species, but this was not true in the majority of cases (Table 3). In most cases, the proximate cause of this appears to be related to differences in the frequency of new flowering and fruiting events between the two sites – some of which are more frequent in Sabangau (e.g., Diospyros siamang, Aglaia rubiginosa) and some in Tuanan (e.g., Gonystylus bancanus). In some cases, the frequency of new flowering and fruiting events is similar, but the onset appears earlier at one site; e.g., Madhuca motleyana, which, upon inspection of the graphs for timing of new flowering and fruiting events during each month, appears to begin flowering and fruiting slightly earlier in Sabangau. It is difficult to speculate about the ultimate cause for these differences, as some species appear to flower/fruit more frequently at one site, and some in the other. Potential Abiotic Triggers of Flowering and Fruiting Six interesting observations emerge from this analysis, which is summarised in Table 4. Firstly, it appears that at least some species may have different abiotic triggers that trigger flowering and fruiting events. In some species, these potentially different abiotic triggers for fruiting and flowering make ecological sense (e.g., Madhuca motleyana in Sabangau, for which flowering appears to occur five months after the autumn equinox in AugustSeptember, with fruiting then following shortly

Phenology Paper 2

Page 163


after, five months after the onset of the dry season; i.e., between September-January) and hint towards the onset of fruiting simply following the onset of flowering with a specified lag period and not actually being triggered by any specific abiotic variable. This is not the case for all species, however; e.g., Dyera lowii / polyphylla in Sabangau, for which flowering appears most likely to be triggered by low mean minimum temperature with a lag of two months, which can occur at unpredictable times, and fruiting potentially triggered by long day length. Secondly, as in other studies (e.g., G端nter et al., 2008), comparison of the most likely potential abiotic triggers of flowering and fruiting indicates that different abiotic variables may trigger flowering and fruiting in different species. This is to be expected in PSF, which does not participate in the SAMF events that characterise dryland forests in the region (Cannon et al., 2007a), and in which many tree species flower and fruit in synchrony in response to (presumably) the same environmental trigger associated with El Ni単o events (Ashton et al., 1988; Appanah, 1993; Sakai et al., 1999). A combination of different abiotic triggers of flowering and fruiting, and of different lag periods between the occurrence of the trigger and the onset of flowering/ fruiting, therefore results in relative asynchrony between species. This likely contributes towards the observation of relatively low levels of fruit availability, and fruiting variability, in Sabangau compared to masting habitats, such as Gunung Palung National Park (Harrison et al., 2010). PSF is a relatively homogenous habitat. Consequently, frugivores do not have the opportunity to move to other habitat types to search for fruit when fruit availability in one habitat type is low (this is noted as an important behavioural response to decreased fruit availability in tropical rainforest animals, Leighton and Leighton, 1983; Hemingway and Bynum, 2005). This apparent use of different abiotic fruiting triggers, and different lag periods between these triggers and fruiting events, for the species in our sample, may therefore be an important factor in permitting the persistence of many species of frugivore in this habitat type, as this ensures that tree species come into fruit at different times of year, preventing fruit availability from dropping too low for too long. In turn, this may also be an important factor allowing the relatively nutrient poor peat soils to support such a high density of floral species. Thirdly, in apparent contrast to the fact that PSF is not a masting habitat, some species in the sample do appear to respond to abiotic triggers hypothesised to be associated with SAMF events. For example, flowering in Shorea parvifolia in Tuanan, and for Koompassia malaccensis at both sites, appears associated with the onset of El Ni単o events, which typically coincide with mast-fruiting events in the region (Ashton et al., 1998; Appanah, 1993). Shorea is a dipterocarp species and Koompassia has also been documented as participating in mast-fruiting events in dryland forests (Appanah, 1985), so the fact that these species also flower in response to El Ni単o events in PSF indicates that these species probably employ a consistent abiotic trigger across their range, even in non-masting habitats. An association between flowering and drops in minimum temperature over a period of at least four days (suggested as the most likely trigger for masting by Ashton et al., 1988) was not apparent in our analysis, however, although we cannot exclude similar drops over a more prolonged period or with more severe temperature reductions as the potential trigger based on the current analysis. Fourthly, despite the relatively small variations in day length and temperature near the equatorial and in these two forests, many of the tree species studied responded either to the spring or autumn equinox, or changes in day length. A key role of equinoxes, particularly the spring equinox (van Schaik et al., 1993; Wright and van Schaik, 1994; van Schaik and Pfannes, 2005), and day length (Audus, 1972; Vince Prue et al., 1984) in the flowering and fruiting patterns of tropical trees has been hypothesised previously, and our results support these hypotheses, although it is clear that these variables are not associated with flowering and/or fruiting in all species. Indeed, the overhead passage of the sun was identified as the most likely abiotic trigger of flowering in more species than for any other variable at both sites (14/28 species in Sabangau and 6/23 species in Tuanan), and of fruiting in many species (4/28 in Sabangau and 9/22 in Tuanan). Presumably, this is related to the predictable nature of these triggers, compared to the other abiotic variables analysed. This is likely to be particularly true for species that fruit each year. Rainfall also appears important in determining the onset of flowering and fruiting in many species, with a large number of species responding to either the start of the wet (flowering: 3/28 in Sabangau and 2/24 in Tuanan; fruiting: 4/28 in Sabangau and 1/22 in Tuanan) or dry (flowering: 2/28 in Sabangau and 1/24 in Tuanan;

Phenology Paper 2

Page 164


fruiting: 5/28 in Sabangau and 1/22 in Tuanan) season. Few species responded directly to mean monthly rainfall. Fifthly, the same potential abiotic triggers of flowering and fruiting for a species were shared between the two sites in some, but not the majority, of species. In many of the species for which incongruence was found between triggers, the species in question either flowered/fruited very irregularly (making accurate identification of potential triggers difficult) or very regularly (more than once a year, in which case there may not even be a specific abiotic trigger of flowering and/or fruiting). This plus the fact that, from an evolutionary perspective, one would expect trees of the same species in different locations to make use of the same abiotic trigger for flowering and fruiting, implies that less confidence should be given to results for species showing apparent differences in abiotic triggers between the two sites. Essentially the same abiotic trigger of flowering between sites was identified for five species and a similar sensitivity in one further (Campnosperma coriaceum, responded to mean maximum temperature in Sabangau and mean minimum temperature in Tuanan). Interestingly, in three of the five cases of shared triggers for flowering between the two sites, the trigger identified was the spring equinox. Only two instances of shared potential triggers between sites were identified for fruiting, both of which involved the spring equinox. In P. pseudorostratum the different fruiting triggers identified actually led to fruiting at very similar times between the two sites: in Sabangau, fruiting appears triggered by the onset of the wet season, with a lag of two months, which would typically result in fruiting in October-January; whereas in Tuanan the response is apparently to the spring equinox, with a lag of one month, which also result in fruiting in October-November. Thus, it is possible that the trigger has been mis-identified in one of these cases, or that the lag period is more variable at one site than the other. Finally, species within certain genera were also identified as sharing the same potential abiotic trigger of flowering and/or fruiting in both of the sites. For example, with the possible exception of Calophyllum nodosum fruiting, the flowering and fruiting of all Calophyllum species appears to respond to the onset of the wet season at both sites. Similarly, all the Cratoxylon species appear to respond to the spring equinox.

Phenology Paper 2

Page 165


Table 4. Most likely abiotic triggers of flowering and fruiting in Sabangau and Tuanan. Parentheses indicate the number of months following the occurrence of the trigger (lag time) until the onset of flowering or fruiting. NA indicates that a species was not present in the phenology plots at a site. All the below tests were statistically significant following correction for multiple comparisons, with the exception of those marked by “*”, which produced a potentially useful relationship and were significant pre correction for multiple comparisons, but not post correction (those that did not produce a potentially useful relationship are considered to have no identifiable potential abiotic trigger). “ǂ” indicates a test result that was statistically significant, but is of limited practical value and/or dubious (see species accounts for further information).

Family Anacardiaceae Anisophyllaceae Apocynaceae Clusiacea / Guttiferae Clusiacea / Guttiferae Clusiacea / Guttiferae Clusiacea / Guttiferae Dipterocarpaceae Dipterocarpaceae Dipterocarpaceae Dipterocarpaceae Ebenaceae Elaeocarpaceae

Species Campnosperma coriaceum Combretocarpus rotundatus Dyera lowii / polyphylla Calophyllum hosei

Fabaceae / Leguminosae Hypericaceae Hypericaceae Hypericaceae Lauraceae Melastomataceae

Koompassia malaccensis

Phenology Paper 2

Calophyllum nodosum Calophyllum sclerophyllum (SA) Calophyllum sclerophyllum (TU) Shorea balangeran Shorea parvifolia Shorea teysmanniana Vatica rassak Diospyros siamang Elaeocarpus mastersii

Cratoxylon spp. Cratoxylon arborescens C. glaucum Litsea spp. Pternandra cf. coerulescens /

Most likely trigger of flowering Sabangau Tuanan Mean max temp Mean min temp (1) ǂ No trigger identified NA Mean min temp (2) Onset dry season (3) Onset wet season (2-3) NA

Most likely trigger of fruiting Sabangau Tuanan Onset dry season (2) * Day length (2) ǂ No trigger identified NA Day length Spring equinox (3) Onset wet season (3-4) NA

NA

Onset wet season (3-4)

NA

Spring equinox (5)

Onset wet season (1-3)

NA

Onset wet season (2-4)

NA

NA

No trigger identified

NA

NA

Rainfall (3) * NA No trigger identified No trigger identified No trigger identified 4-d rolling max temp (8)

NA El Niño event (4) NA NA No trigger identified 2 month drought (6)

Mean min temp (7) NA Mean 24-h temp (1) No trigger identified No trigger identified Mean min temp (2)

El Niño event

El Niño event (3)

El Niño event (3)

NA No trigger identified NA NA No trigger identified Onset wet season (4) / onset dry season (8) Spring equinox (4)

Spring equinox Spring equinox Spring equinox Mean rainfall (5) Rolling max temp

Spring equinox No trigger identified Spring equinox Mean rainfall (4) NA

Spring equinox (0-1) Spring equinox (0-1) Spring equinox (0-1) Onset dry season (2) No trigger identified

Spring equinox (1) NA Spring equinox (1) Spring equinox (1) NA

Page 166


Meliaceae Myristicaceae Myrtaceae Myrtaceae Myrtaceae Sapotaceae Sapotaceae Sapotaceae Sapotaceae Sapotaceae Tetrameristaceae Theaceae Thymelaeaceae

Phenology Paper 2

galeata Aglaia rubiginosa

Onset dry season (1-2)

Autumn equinox (2)

Onset dry season (2)

Horsfieldia crassifolia Eugenia cf. spicata Syzygium cf. garcinifolia / havilandii Tistaniopsis spp. Madhuca motleyana

Mean min T (2) No trigger identified Day length

Mean min T (4) No fruiting during study Day length (2)

Mean 24-h temp (2) Autumn equinox (5)

4 month drought NA 3 month drought (5) / 4 month drought (4) Onset wet season (3) Spring equinox

Palaquium cochlearifolium Palaquium leiocarpum Palaquium pseudorostratum Palaquium ridleyii / cf. xanthochymum Tetramerista glabra Ploiarium alternifolium Gonystylus bancanus

Onset wet season (1) No trigger identified Day length (2) Onset dry season (6)

Autumn equinox (3) 2 month drought (3) No trigger identified El NiĂąo conditions (4)

Day length (2) NA No trigger identified

No trigger identified Mean visibility Autumn equinox (3)

4-d rolling max temp (8) Onset dry season (6) / 1-2 months following flowering Onset wet season (2-3) * Mean 24-h temp Onset wet season (2) 2 months following flowering Onset dry season NA Autumn equinox

1 month following flowering No trigger identified NA Rainfall (2) Spring equinox (5) Spring equinox (5) Spring equinox (1) Onset drought (3) Spring equinox (1) No trigger identified Autumn equinox Ç‚ No trigger identified No trigger identified

Page 167


In addition to those points discussed in the Methods section, four further points should be considered when interpreting this analysis of abiotic triggers for flowering and fruiting: 1. Although many of our analyses are very suggestive, the identified associations between flowering/fruiting cannot be guaranteed to represent actual phenological triggers and may simply be chance associations. Nevertheless, even though it is not possible to say for certain whether an abiotic variable is definitely a trigger for flowering/fruiting or not, these relationships should still enable prediction of the timing of flowering and particularly fruiting events, which is their key role in terms of formulating forest restoration schedules and protocols. 2. Secondly, some species produced flowers and fruits very consistently during the study period (e.g., Diospyros siamang and Campnosperma coriaceum in Sabangau) or very rarely (e.g., Gonystylus bancanus fruit at both sites) and, consequently, it is very difficult to identify a specific potential abiotic trigger (or at least to accurately predict the timing) of flowering and fruiting events for these species. 3. Furthermore, it is probably inappropriate to attempt to identify potential abiotic triggers in these species: if the species is producing new flowers/fruit in most months of the year, then flowering/fruiting are probably not triggered by any specific abiotic variable and it is more likely that percentage of stems with new flowering/fruiting events may be determined by other factors, such as the nutritional status of individual trees (cf. Burgess, 1972; Fox, 1972; Crone et al., 2009). Of course, in terms of developing protocols and schedules for seed collection, this debate is largely irrelevant: prediction of the timing of fruiting events becomes unnecessary because fruit is available in most months. 4. In an unknown number of cases, our identified triggers for fruiting may be merely chance occurrences. In these cases, fruiting may merely follow flowering (which is likely to be triggered by the identified abiotic variable) with an unidentified lag period, which could either be fixed or be dependent on other variables, such as the nutritional status of individual trees (cf. Burgess, 1972; Fox, 1972; Crone et al., 2009). Anomalous Environmental Events that may Influence Flowering and Fruiting Various anomalous environmental events may influence flowering and fruiting, the most obvious of which for PSF are drought and associated fire/haze. There is evidence that haze from forest fires does influence forest-wide productivity levels (Harrison et al., 2007) and it is quite probable that such anomalous abiotic events may influence flowering and fruiting across a variety of tree PSF species, including those analysed herein. Unfortunately, however, it is impossible to make a proper assessment of the influence of these variables on individual species’ flowering and fruiting patterns using the current data set. A much longer data set is required, due to the rarity of these events and the relatively unpredictable nature of flowering and fruiting patterns in many species.

Phenology Paper 2

Page 168


REFERENCES Anderson, D. P., E. V. Nordheim, T. C. Moermond, Z. B. Gone and C. Boesch (2005). Factors influencing tree phenology in Taï National Park, Côte d’Ivoire. Biotropica 37: 631-640. doi: 10.1111/j.17447429.2005.00080.x. Appanah, S. (1985). General flowering of the climax rain forests of South-East Asia. Journal of Tropical Ecology 1: 225-40. Appanah, S. (1993). Mass flowering of dipterocarp forests in the aseasonal tropics. Journal of Biosciences 18: 457-74. Ashton, P. S., T. J. Givnish and S. Appanah (1988). Staggered flowering in the Dipterocarpaceae: New insights into floral induction and the evolution of mast fruiting in the aseasonal tropics. American Naturalist 132: 44-66. Audus, L. J. (1972). Plant Growth Substances. Barnes and Noble, New York. Brearley, F. Q., J. Proctor, Suriantata, L. Nagy, G. Dalrymple and B. C. Voysey (2007). Reproductive phenology over a 10-year period in a lowland evergreen rain forest of central Borneo. Journal of Ecology 95: 828-839. doi: 10.1111/j.1365-2745.2007.01258.x. Burgess, P. F. (1972). Studies on the regeneration of the hill forests of the Malay peninsular: the phenology of dipterocarps. Malayan Forester 35: 103-123. Cannon, C. H., L. M. Curran, A. J. Marshall and M. Leighton (2007a). Long-term reproductive behaviour of woody plants across seven Bornean forest types in the Gunung Palung National Park (Indonesia): suprannual synchrony, temporal productivity and fruiting diversity. Ecology Letters 10: 956-969. doi: 10.1111/j.1461-0248.2007.01089.x. Cannon, C. H., L. M. Curran, A. J. Marshall and M. Leighton (2007b). Beyond mast-fruiting events: Community asynchrony and individual dormancy dominate woody plant reproductive behavior across seven Bornean forest types. Current Science 93: 1558-1566. Chapman, C. A., R. W. Wrangham, L. J. Chapman, D. K. Kennard and A. E. Zanne (1999). Fruit and flower phenology at two sites in Kibale National Park, Uganda. Journal of Tropical Ecology 15: 189-211. Cheyne, S. M. (2007). Effects of meteorology, astronomical variables, location and human disturbance on the singing apes: Hylobates albibarbis. American Journal of Primatology 70: 386-392. doi: 10.1002/ajp.20502. Crone, E. E., E. Miller and A. Sala (2009). How do plants know when other plants are flowering? Resource depletion, pollen limitation and mast-seeding in a perennial wildflower. Ecology Letters 12: 11191126. doi: 10.1111/j.1461-0248.2009.01365.x. Curran, L. M. and M. Leighton (2000). Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting Dipterocarpaceae. Ecological Monographs 70: 101-128. FORRU (2005) How to Plant a Forest: The Principles and Practice of Restoring Tropical Forests. Forest Restoration Research Unit, Biology Department, Science Faculty, Chiang Mai University, Thailand. Fox, J. E. D. (1972). The Natural Vegetation of Sabah and Natural Regeneration of the Dipterocarp Forests. University of Wales Aberystwyth, Aberystwyth. Frederiksson, G., S. A. Wich and Trisno (2006). Frugivory in sun bears (Helarctos malayanus) is linked to El Niño-related fluctuations in fruiting phenology, East Kalimantan, Indonesia. Biological Journal of the Linnean Society 89: 489-508. Graham, M. H. (2003). Confronting multicollinearity in ecological multiple regressions. Ecology 84: 28092815. Günter, S., B. Stimm, M. Cabrera, M. L. Diaz, M. Lojan, E. Ordoñez, M. Richter and M. Weber (2008). Tree phenology in montane forests of southern Ecuador can be explained by precipitation, radiation and photoperiodic control. Journal of Tropical Ecology 24: 247-258. doi: 10.1017/S0266467408005063 Harrison, M. E., S. M. Cheyne, Y. Sulistiyanto and J. O. Rieley (2007). Biological effects of smoke from dryseason fires in non-burnt areas of the Sabangau peat-swamp forest, Central Kalimantan, Indonesia. In: J. O. Rieley, C. J. Banks and B. Radjagukguk (Eds). Carbon-Climate-Human Interaction on Tropical Peatland. Proceedings of The International Symposium and Workshop on Tropical Peatland, Yogyakarta, 27-29 August 2007, EU CARBOPEAT and RESTORPEAT Partnership, Gadjah Mada University, Indonesia and University of Leicester, United Kingdom. www.geog.le.ac.uk/carbopeat/yogyaproc.html.

Phenology Paper 2

Page 169


Harrison, M. E. and D. J. Chivers (2007). The orang-utan mating system and the unflanged male: A product of declining food availability during the late Miocene and Pliocene? Journal of Human Evolution 52: 275-293. doi: 10.1016/j.jhevol.2006.09.005. Harrison, M. E., S. E. Page and S. H. Limin (2009). The global impact of Indonesian forest fires. Biologist 56: 156-163. Harrison, M. E., S. J. Husson, N. Zweifel, L. J. D’Arcy, H. C. Morrogh-Bernard, S. M. Cheyne, M. A. van Noordwijk and C. P. van Schaik (2010). The Fruiting Phenology of Peat-Swamp Forest Tree Species at Sabangau and Tuanan, Central Kalimantan, Indonesia. Report for the Kalimantan Forests and Climate Partnership, Palangka Raya, Indonesia. Harrison, M. E., H. C. Morrogh-Bernard and D. J. Chivers (2010). Orangutan energy intake and the influence of fruit availability in the non-masting peat-swamp forest of Sabangau, Indonesian Borneo. International Journal of Primatology 31: 585-607. Hemingway, C. A. and N. Bynum (2005). The influence of seasonality on primate diet and ranging. In: D. K. Brockman and C. P. van Schaik (Eds). Seasonality in Primates: Studies of Living and Extinct Human and Non-Human Primates. Cambridge University Press, Cambridge. pp. 57-104. Hochberg, Y. (1988). A sharper Bonferroni procedure for multiple tests of significance. Biometrika 75: 800802.. Husson, S., H. Morrogh-Bernard, L. D'Arcy, S. M. Cheyne, M. E. Harrison and M. Dragiewicz (2007). The importance of ecological monitoring for habitat management - A case study in the Sabangau forest, Central Kalimantan, Indonesia. In: J. O. Rieley, C. J. Banks and B. Radjagukguk (Eds). Carbon-ClimateHuman Interaction on Tropical Peatland. Proceedings of The International Symposium and Workshop on Tropical Peatland, Yogyakarta, 27-29 August 2007, EU CARBOPEAT and RESTORPEAT Partnership, Gadjah Mada University, Indonesia and University of Leicester, United Kingdom. www.geog.le.ac.uk/carbopeat/yogyaproc.html. Janzen, D. H. (1974). Tropical blackwater rivers, animals, and mast fruiting by the Dipterocarpaceae. Biotropica 6: 69-103. Knott, C. D. (1999). Reproductive, Physiological and Behavioural Responses of Orangutans in Borneo to Fluctuations in Food Availability. PhD thesis, Harvard University. Latif, M., D. P. Anderson, T. Barnett, M. Cane, R. Kleeman, A. Leetma, J. O'Brien, A. Rosati and E. Schneider (1998). A review of the predictability and prediction of ENSO. Journal of Geophysical Research 103: 14374-14393. Leighton, M. and D. R. Leighton (1983). Vertebrate responses to fruiting seasonality within a Bornean rainforest. In: S. L. Sutton, T. C. Whitmore and C. A. Chadwick (Eds). Tropical Rainforest: Ecology and Management. Blackwell Scientific Publications, Oxford. pp. 181-196. Marshall, A. J. and M. Leighton (2006). How does food availability limit the population density of whitebearded gibbons? In: G. Hohmann, M. M. Robbins and C. Boesch (Eds). Feeding Ecology in Apes and Other Primates: Ecological, Physical and Behavioral Aspects. Cambridge University Press, Cambridge. pp. 311-333. Marshall, A. J., M. Ancrenaz, F. Q. Brearley, G. M. Fredriksson, N. Ghaffar, M. Heydon, S. J. Husson, M. Leighton, K. R. McConkey, H. C. Morrogh-Bernard, J. Proctor, C. P. van Schaik, C. P. Yeager and S. A. Wich (2009). The effects of forest phenology and floristics on populations of Bornean and Sumatran orangutans: are Sumatran forests better orangutan habitat than Bornean forests? In: S. A. Wich, S. S. Utami Atmoko, T. Mitra Setia and C. P. van Schaik (Eds). Orangutans: Geographic Variation in Behavioral Ecology and Conservation. Oxford University Press, Oxford. pp. 97-116. McClure, H. E. (1966). Flowering, fruiting and animals in the canopy of a tropical rain forest. Malaysian Forester 29: 182-203. McPhaden, M. J. and J. Picaut (1990). El Niùo-Southern Oscillation displacements of the Western Equatorial Pacific Warm Pool. Science 250: 1385-88. Medway, L. (1972). Phenology of a tropical rain forest in Malaya. Biological Journal of the Linnean Society 4: 117-46. Morrogh-Bernard, H., S. Husson, S. E. Page and J. O. Rieley (2003). Population status of the Bornean orangutan (Pongo pygmaeus) in the Sebangau peat swamp forest, Central Kalimantan, Indonesia. Biological Conservation 110: 141-52. Ostfeld, R. S. and F. Keesing (2000). Pulsed resources and community dynamics of consumers in terrestrial ecosystems. Trends in Ecology and Evolution 15: 232-237.

Phenology Paper 2

Page 170


Page, S. E., J. O. Rieley, Ø. W. Shotyk and D. Weiss (1999). Interdependence of peat and vegetation in a tropical peat swamp forest. Philosophical Transactions of the Royal Society of London B 354: 18851807. Page, S. E., F. Siegert, H. D. V. Boehm, A. Jaya and S. Limin (2002). The amount of carbon released from peat and forest fires in Indonesia during 1997. Nature 420: 61-5. Roback, P. J. and R. A. Askins (2004). Judicious use of multiple hypothesis tests. Conservation Biology 19: 261-267. Sakai, S., K. Momose, T. Yumoto, T. Nagamitsu, H. Nagamasu and A. A. Hamid (1999). Plant reproductive phenology over four years including an episode of general flowering in a lowland dipterocarp forest, Sarawak, Malaysia. American Journal of Botany 86: 1414-36. Sakai, S. (2002). General flowering in lowland mixed dipterocarp forests of South-east Asia. Biological Journal of the Linnean Society 75: 233-47. Shepherd, P. A., J. O. Rieley and S. E. Page (1997). The relationship between forest structure and peat characteristics in the upper catchment of the Sungai Sebangau, Central Kalimantan. In: J. O. Rieley and S. E. Page (Eds). Biodiversity and Sustainability of Tropical Peatlands. Samara Publishing, Cardigan, UK. pp. 191-210. Shukla, J. (1998). Predictability in the midst of chaos: a scientific basis for climate forecasting. Science 282: 728-731. Sulistiyanto, Y., J. O. Rieley and S. Limin (2004). Nutrient dynamics in different sub-types of ombrotrophic peat swamp forest in Central Kalimantan, Indonesia. In: J. Päivänen (Ed). Wise Use of Peatlands (Vol I). Proceedings of the 12th International Peat Congress, Tampere, Finland, 6-11 June 2004. Saarijärven Offset Oy, Saarijärvi, Finland. pp. 752-759. Tutin, C. E. G. and M. Fernandez (1993). Relationships between minimum temperature and fruit production in some tropical forest trees in Gabon. Journal of Tropical Ecology 9: 241-248. van Schaik, C. P., J. W. Terborgh and S. J. Wright (1993). The phenology of tropical forests: Adaptive significance and consequences for primary consumers. Annual Review of Ecology and Systematics 24: 353-77. van Schaik, C. P. (1996). Strangling figs: their role in the forest. In: C. P. van Schaik and J. Supriatna (Eds). Leuser: A Sumatran Sanctuary. Perdana Ciptamadri, Jakarta. pp. 111-119. van Schaik, C. P. (1999). The socioecology of fission-fusion sociality in orangutans. Primates 40: 69-86. van Schaik, C. P. and K. R. Pfannes (2005). Tropical climates and phenology: a primate perspective. In: D. K. Brockman and C. P. van Schaik (Eds). Seasonality in Primates: Studies of Living and Extinct Human and Nonhuman Primates. Cambridge University Press, Cambridge. pp. 23-54. van Schaik, C. P., S. A. Wich, S. S. Utami and K. Odom (2005). A simple alternative to line transects of nests for estimating orangutan densities. Primates 46: 249-254. doi: 10.1007/s10329-005-0134-z. Vince Prue, D., B. Thomas and K. E. Kockshull (1984). Light and the Flowering Process. Academic Press, New York. Wich, S. A. and C. P. van Schaik (2000). The impact of El Niño on mast fruiting in Sumatra and elsewhere in Malesia. Journal of Tropical Ecology 16: 563-77. Wich, S. A., S. S. Utami Atmoko, T. Mitra Setia and C. P. van Schaik (2009). Introduction. In: S. A. Wich, S. S. Utami Atmoko, T. Mitra Setia and C. P. van Schaik (Eds). Orangutans: Geographic Variation in Behavioral Ecology and Conservation. Oxford University Press, Oxford. pp. xix-xxviii. Wood, G. H. S. (1956). The Dipterocarp flowering season in North Borneo. Malayan Forester 19: 193-201. Wright, S. J. and C. P. van Schaik (1994). Light and the phenology of tropical trees. American Naturalist 143: 192-199. Wycherley, P. R. (1973). The phenology of plants in the humid tropics. Micronesia 9: 75-96. Yasuda, M., J. Matsumoto, N. Osada, S. e. Ichikawa, N. Kachi, M. Tani, T. Okuda, A. Furukawa, A. R. Nik and N. Manokaran (1999). The mechanism of general flowering in Dipterocarpaceae in the Malay Peninsular. Journal of Tropical Ecology 15: 437-49. Zar, J. H. (1999). Biostatistical Analysis: Fourth Edition. Prentice Hall, New Jersey.

Phenology Paper 2

Page 171


APPENDIX I - CALENDAR OF SPECIES WITH HIGH LIKELIHOOD OF FRUITING IN DIFFERENT MONTHS AND FRUITING REGULARITY CATEGORIES Here we indicate the months in which a high likelihood of fruiting exists, plus fruiting regularity classifications for both sites (adapted from Harrison et al., 2010). The calendar of timing of high-likelihood fruiting events and fruiting regularity classification table on which this is based are provided in Appendices I and II of Harrison et al. (2010), respectively. Key Abbreviations for monthly fruiting events: s = Sabangau, high likelihood of ≥ 1% fruiting stems bearing fruit; S = Sabangau, high likelihood of ≥ 10% fruiting stems bearing fruit; t = Tuanan, high likelihood of ≥ 1% fruiting stems bearing fruit; T = Tuanan, high likelihood of ≥ 10% fruiting stems bearing fruit. Abbreviations for fruiting categories: - = no date; 0 = never fruited; F = frequent; I = irregular; R = regular; YR = year round. Species

Category

Month

SA

TU

Jan

Anacardiaceae Campnosperma coriaceum Campnosperma squamatum

F I

R -

s

Anisophyllaceae Combretocarpus rotundatus

F

-

Apocynaceae Dyera lowii / polyphylla

F

R

Clusiaceae / Guttiferae Calophyllum hosei Calophyllum nodusum Calophyllum sclerophyllum (SA) Calophyllum sclerophyllum (TU)

R R -

I 0

Feb

Mar t

S

s t

Apr

s

s

May

Jun

s

s

s

Jul

Aug

Sep

Oct

Nov

s

s

s

Dec S

s

s

S

s

S t

s

s

Dipterocarpaceae

Phenology Paper 2

Page 172


Shorea balangeran Shorea parvifolia Shorea teysmanniana Vatica mangachopai

I I I

I -

Ebenaceae Diospyros siamang

Y

R

Elaeocarpaceae Elaeocarpus mastersii

I

I

Fabaceae / Leguminosae Koompassia malaccensis

I

I

Hypericaceae Cratoxylon arborescens Cratoxylon glaucum

R I

0 I

Lauraceae Litsea cf. elliptica Litsea cf. rufo-fusca Litsea sp. 1 cf. resinosa Litsea spp.

I R R R

I

Linaceae Ctenolophon parvifolius

F

-

Melastomataceae Pternandra cf. coerulescens / galeata

I

-

Phenology Paper 2

S

S

S

S

S

S

S

S

s

S

S

S t

S

S

S

s

S

s s

s

S

S

S

Page 173


Meliaceae Aglaia rubiginosa

Y

I

Myristicaceae Horsfieldia crassifolia

I

I

0 F R I

I F I

F I F F R

R I F F I

S

s t

s T sT s

T s T

Tetrameristaceae Tetramerista glabra

F

R

s

Theaceae Ploiarium alternifolium

-

I

Thymelaeaceae Gonystylus bancanus

I

I

Myrtaceae Eugenia cf. spicata Syzygium garcinifolia Syzygium havilandii Tristaniopsis spp. Sapotaceae Madhuca mottleyana Palaquium cochlearifolium Palaquium leiocarpum Palaquium pseudorostratum Palaquium ridleyii / cf. xanthochymum

Phenology Paper 2

S

s

s

s

s

s

S

S

S

S

S

S

S

S s

S t

s

S t

s t

s t

s

s

S

S

s

S

S t

S t t

S T s T

S T s T

s T s T

s

S

s

s

s

S

s

s

s

Page 174




Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.