Keywords

1 Introduction

Native riparian forests are distributed along the Ooyamazawa stream in the Chichibu Mountains of the Kanto region, central Japan. In these forests, Fraxinus platypoda Oliv. (Oleaceae) is one of the dominant canopy tree species, coexisting with Pterocarya rhoifolia and Cercidiphyllum japonicum. The Chichibu Mountains harbor a very complex topography characterized by steep slopes, with tree diversity distributed among the resulting microhabitats. For example, Tsuga sieboldii and Chamaecyparis obtusa forests are distributed along ridges, Fagus crenata and Fagus japonica on mountain slopes, and F. platypoda, P. rhoifolia, and C. japonicum in valleys (Maeda and Yoshioka 1952; Tanaka 1985). The regeneration mechanisms of T. sieboldii, F. crenata, and F. japonica have been extensively studied by Suzuki (1979, 1980, 1981a, b), Nakashizuka and Numata (1982a, b) and Nakashizuka (1983, 1984a, b), and Ohkubo et al. (1988, 1996), respectively. These researchers examined forest regeneration within the context of gap dynamics theory. Gaps are open spaces within the forest canopy layer formed by die back, trunk breakage, and uproot ed trees. In general, gaps occur on a small scale and are not often accompanied by soil disturbance, with the exception of uprooted trees.

On the other hand, disturbance s within riparian zones vary in type, frequency, magnitude, and size compared to canopy gap formation on hillslopes. In steeper mountain regions, valley floor landforms are sculpted by fluvial process es and a variety of mass soil movement processes from tributaries and adjacent hillslopes (Gregory et al. 1991).

The ash species, Fraxinus platypoda, is a late successional species in riparian habitats of cool temperate forest s distributed along the Pacific coast of Japan. This species is distributed from Tochigi Prefecture in the north to Miyazaki Prefecture in the south (Fig. 2.1). Kisanuki et al. (1992) and Ann and Oshima (1996) examined regeneration mechanisms within a gap dynamics framework in mixed forests of F. Platypoda and P. rhoifolia. However, the regeneration of F. platypoda may also be related to large-scale natural disturbances such as debris flow s and landslide s in the riparian zone . In this chapter, I present long-term research related to the life history and regeneration process of F. platypoda with respect to natural disturbances within the riparian zone.

Fig. 2.1
figure 1

Distribution of F. platypoda. Modified after Kawahara et al. (2009)

Full size image

2 Study Species

Fraxinus platypoda is a deciduous canopy species that can reach up to 40 m in height and 150 cm in diameter at breast height (DBH, 130 cm) (Fig. 2.2). This species is well adapted to stream disturbance s that vary in frequency and size, and it dominates forests in riparian zones (Sakio 1997).

Fig. 2.2
figure 2

F. platypoda forest in the Ooyamazawa riparian forest

Full size image

Branching occurs above the trunk of the tree. Most individuals have a single trunk and rarely exhibit sprouts, similar to species such as P. rhoifolia and C. japonicum. Leaves are impari-pinnate compound, consisting of 7–9 leaflets, and leaf length is about 25–35 cm and decussate-opposite (Fig. 2.3). The base of the petiole markedly bulges and holds the stem. Open hairs occur along the middle vein on the back of the leaf, but others are hairless. The apical leaflet is oblong-ellipsoid oblanceolate in shape, and is 8–20 cm in length and 3–7 cm in width, with a small petiole of 1–2 cm in length. The side leaflets lack a petiole, and the base is wedged with fine serrations. Twigs are thick, gray-brown, and hairless, and many are oval lenticel. The pith is thick. Branches grow rapidly in early spring and stop growing in June (Sakio 1993). The root system is concentrated in the shallow part of the ground surface. Fraxinus platypoda has medium to large-diameter straight roots and horizontal root s and is classified as a deep root type. Fine roots are dense, but root hairs are rare (Karizumi 1979). The species is easily uprooted by scouring running water. In saplings, sedimentation by soil and sand easily produces adventitious root s (Sakio 2002).

Fig. 2.3
figure 3

Leaves of F. platypoda

Full size image

3 Reproductive Traits

3.1 Flower

Sexual expression of F. platypoda is not obvious. The species exhibits two morphological types: one with male flower s and the other with hermaphrodite flower s (Fig. 2.4). Both flowers lack a perianth. Male flowers have one stamen with a pair of anther s, while hermaphrodite flowers have one pistil with a pair of anthers. Whether the male portion of hermaphrodite flowers of F. platypoda is functional is unknown. However, sexual expression of F. platypoda may be functionally considered to be androdioecy , as the pollen of hermaphrodite flowers exhibits germination ability. Because the breeding characteristics of F. platypoda are not clearly understood, I treat the sexual expression of F. platypoda as dioecy , i.e., female trees and male trees, in this chapter.

Fig. 2.4
figure 4

Two types of flowers of F. platypoda

Full size image

The timing of flowering in F. platypoda differs depending on altitude but, typically, occurs in mid-April (700 m a.s.l) to mid-May (1500 m a.s.l.). In the Ooyamazawa riparian forest (1500 m a.s.l.), flowering occurs from the beginning of May to mid-May. However, the exact timing fluctuates annually.

Flowering data collected over 28 years have demonstrated clear fluctuations of flower values for F. platypoda in the core research plot (Fig. 2.5). The numbers of female and male trees in the core plot (0.54 ha) were 26 and 20, respectively. The rank of flowering for female and male trees was scored from 1 to 5 by observation using binoculars. The average flowering rank of all individuals was 3.22 for females and 4.11 for males; values were significantly higher in males. The coefficient of variation (CV) was 0.48 for females and 0.28 for males; values were significantly higher in females. In this forest, although the flowering interval was 2–3 years, the interval changed over the 28 years. The flowering of the two sexes exhibited clear synchronization until 2002, but synchronization ceased thereafter. After 2002, more males have flowered every year, while females have retained a distinct interval. The change in flowering fluctuation after 2002 may have been driven by external factors, such as climate warming.

Fig. 2.5
figure 5

Mean rank of flowering of female and male trees of F. platypoda from 1990 to 2017. The numbers of female and male trees were 26 and 20, respectively

Full size image

3.2 Seed Production

After flowering, fruit (samara ) of F. platypoda continues to grow, causing abortion . The pericarp finishes growing at the end of August, at which point seeds begin to grow. Oven-dried matured fruit of F. platypoda weighed 144 ± 24 mg, and the dry weight of seeds was 80 ± 17 mg (Fig. 2.6; Sakio et al. 2002). Seeds mature in mid-October and are dispersed by wind and stream water in November (Fig. 2.7). Many mature seeds occur in mast year s, but a large number of seeds are empty or insect-damaged during non-mast years.

Fig. 2.6
figure 6

Fruits and seeds of F. platypoda in autumn

Full size image
Fig. 2.7
figure 7

Fruits of F. platypoda in the stream in autumn (Sakio 2008)

Full size image

Clear fluctuations in seed production and flowering of F. platypoda occurred over the 28 years (Fig. 2.8). Because seed production was strongly positively correlated with the extent of flowering, the former is presumed to be regulated by the latter (Fig. 2.9).

Fig. 2.8
figure 8

Mean rank of seed production of F. platypoda from 1990 to 2017

Full size image
Fig. 2.9
figure 9

Relationship between flower and seed rank

Full size image

4 Germination

The seeds of F. platypoda germinate from the end of June to mid-July. Current-year seedling s are found on litter , gravel, mineral soil, and fallen logs, except after non-mast years. As long as the light environment is not very strong, current-year seedlings of F. platypoda only have cotyledon s (Fig. 2.10), and the true leaf does not expand during the germination year. The germination site of F. platypoda is not strongly restricted by the soil and light environment, unlike C. japonicum, whose germination sites are limited. F. platypoda does not exhibit seed dormancy . The seeds of F. platypoda produced in autumn germinate in the early summer of the following year and do not germinate thereafter.

Fig. 2.10
figure 10

Current seedling of F. platypoda

Full size image

5 Seedling Survival

The seeds of F. platypoda are randomly dispersed. Therefore, the distribution of current-year seedling s of F. platypoda is not affected by microenvironments such as soil and light conditions. However, the distribution pattern of seedlings of F. platypoda changes with the growth of seedlings, shifting from random or uniform distribution to aggregated throughout growth (Fig. 2.11). Small-sized seedlings (height < 20 cm) tend to be distributed around the active channel (Figs. 2.11 and 2.12), while the distribution of larger seedlings (20 cm ≤ height < 1 m) is more closely related to the microtopography than to canopy gap s. These larger seedlings tend to be aggregated in abandoned channel s. On hillslopes where the forest floor vegetation is dense, F. platypoda seedlings disappear after several years due to the effects of shade. The mean longevity of F. platypoda seedlings in various environments is 1.19 ± 0.58 years (Sakio et al. 2002). On the other hand, seedlings near active channels exhibit long life spans and high density due to the absence of forest floor vegetation (Fig. 2.11). However, seedling communities can be destroyed by flooding, often caused by large typhoon s. Gravel deposit s formed during flooding events can serve as new seedling establishment sites. When the seedling community stabilizes due to channel fluctuations, it continues to grow into a large-sized seedling community (20 cm ≤ height).

Fig. 2.11
figure 11

Distribution of F. platypoda seedlings. Upper and lower graphs show the relationship with microtopography and canopy gaps, respectively. The six plots in the figure are each 30 × 32 m, and numbers indicate the number of individuals in 1 m2

Full size image
Fig. 2.12
figure 12

Advanced sapling communities of F. platypoda along the stream

Full size image

In other words, the distribution of F. platypoda seedlings is regulated by the microtopographic variation in forest floor vegetation . In riparian forests, the light environment depends not only on the presence of canopy gap s but also on gaps in the forest floor vegetation due to stream disturbance . The dynamics of F. platypoda seedlings are thought to be strongly influenced by the latter.

6 Seedling Growth

Large-sized seedlings (1 m ≤ height) that have established on stable sites exhibit variation in growth rates depending on the light environment . New shoots of F. platypoda begin to elongate rapidly in early spring and stop growing in June (Sakio 1993). Seedling s under canopy gap s grow faster than seedlings under the canopy (Fig. 2.13) and have more leaves. In many forests, canopy gaps appear to be necessary for the growth of canopy trees (Suzuki 1980, 1981a; Nakashizuka and Numata 1982a, b; Nakashizuka 1983, 1984a). Therefore, even in riparian forests, gap formation via the death of canopy trees may be necessary for seedlings of F. platypoda to grow into the canopy.

Fig. 2.13
figure 13

Seasonal changes of current shoot lengths of F. platypoda saplings in canopy gaps and under the canopy. Modified after Sakio (1993)

Full size image

7 Forest Structure and Spatial Distribution

In the 4.71-ha study plot within the Ooyamazawa riparian forest, we observed 418 F. platypoda individuals out of a total of 2214 trees (4 cm ≤ DBH) (Sakio et al. 2002). Canopy tree s, subcanopy trees (10 cm ≤ DBH), and the shrub layer (DBH < 10 cm) accounted for 304, 59, and 55 individuals, respectively. The mean DBH of F. platypoda canopy trees was 56.9 ± 19.0 cm, with a maximum of 140.5 cm (Sakio et al. 2002). The DBH distribution of F. platypoda was continuous from saplings to large canopy trees (Fig. 2.14). F. platypoda exhibited two peaks in DBH distribution: one formed by small trees (DBH < 10 cm) and the other formed by the 40-cm DBH class. These data suggest that F. platypoda maintains sapling banks. The peak of the 40-cm class suggests synchronous regeneration caused by a large-scale disturbance . The relative density of dominant canopy tree s of F. platypoda is high and does not exhibit distinct fluctuations along the stream.

Fig. 2.14
figure 14

DBH distribution of F. platypoda individuals in the Ooyamazawa riparian plot (4.71 ha). Only individuals over 4 cm in DBH were measured. Modified after Sakio (2008)

Full size image

In the core plot (60 × 90 m), one peak of F. platypoda individuals occurred within the 40–60-cm DBH class, similar to the pattern observed in the 4.71-ha plot (Sakio 1997). In addition, increment core s of all F. platypoda individuals larger than 4-cm DBH were obtained using an increment borer in November 1998, in the core plot. The age distribution of F. platypoda individuals was continuous from saplings to older canopy-aged trees (Sakio 1997). Figure 2.15 presents the spatial distribution of tree age for F. platypoda in the core plot. The age of most individuals was aggregated around 200 years, but several young aggregated groups also occurred (e.g., small patches A, C, and F). These results suggest that a large-scale disturbance occurred around 200 years ago in the Ooyamazawa riparian forest and that F. platypoda regenerated simultaneously. Since that time, F. platypoda advanced sapling s have regenerated under canopy gaps. Thus, F. platypoda is likely to become an overwhelmingly dominant species within the basin because it can regenerate within sites of large-scale disturbance as well as within small gaps.

Fig. 2.15
figure 15

Spatial distribution of F. platypoda individuals in the core plot (60 × 90 m). Numbers show the age of F. platypoda individuals. The size of circles reflects the DBH. Gray and black circles show F. platypoda individuals, and open circles show other species

Full size image

8 Conclusion

Fraxinus platypoda produces a large number of seeds once every few years, thus forming young advanced sapling communities within various microtopographic habitats. In particular, advanced sapling communities are formed in gravel deposit s along mountain streams. These sapling communities continue to be regenerated after repeated destruction by mountain stream disturbances and subsequent regeneration on new gravel deposits. After large disturbances, F. platypoda regenerates within all river basins and also fills canopy gap s by advanced saplings when small gaps are formed. Thus, F. platypoda succeeds as a dominant species in riparian forest s by regenerating in response to various scales of disturbances throughout its life history .