How heterogeneous are the cloud forest communities in the mountains of central Veracruz, Mexico? Guadalupe Williams-Linera, María Toledo-Garibaldi & Claudia Gallardo Hernández Plant Ecology An International Journal ISSN 1385-0237 Volume 214 Number 5 Plant Ecol (2013) 214:685-701 DOI 10.1007/s11258-013-0199-5 1 23 Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media Dordrecht. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”. 1 23 Author's personal copy Plant Ecol (2013) 214:685–701 DOI 10.1007/s11258-013-0199-5 How heterogeneous are the cloud forest communities in the mountains of central Veracruz, Mexico? Guadalupe Williams-Linera • Marı́a Toledo-Garibaldi • Claudia Gallardo Hernández Received: 15 January 2013 / Accepted: 25 March 2013 / Published online: 3 April 2013 Ó Springer Science+Business Media Dordrecht 2013 Abstract The montane forest in central Veracruz, Mexico is a hotspot of biodiversity. We asked whether lower and upper montane forests could be distinguished in this ecoregion. Variables of vegetation and seasonality in precipitation were tested across 14 sites between 1,250- and 2,550-m elevations. A total of 1,639 individuals and 128 tree species was recorded. There was a unimodal pattern in the richness of species, genera, and families; their richness was positively correlated with precipitation in the wettest quarter of the year, though there were no differences in the basal area and density. Rarefaction, species turnover, nonmetric multidimensional scaling, and a cluster histogram suggest two major groups: lower elevation forests that are less diverse, have low beta diversity and are more similar in composition, with Clethra macrophylla, Liquidambar styraciflua, and Quercus lancifolia as indicator species; and higher elevation forests that are more diverse, have high species turnover, and include forests with Quercus corrugata and Prunus rhamnoides, and forests with Fagus grandifolia, Persea americana, and Ternstroemia sylvatica as indicator species. However, other communities (an Oreomunnea mexicana at the upper site, and a limestone site in the lower forests), G. Williams-Linera (&)
M. Toledo-Garibaldi
C. G. Hernández Instituto de Ecologı́a, A.C, Carretera antigua a Coatepec No. 351, 91070 Xalapa, VER, Mexico e-mail: guadalupe.williams@inecol.edu.mx exemplify the high regional heterogeneity. We conclude that elevation and seasonality in precipitation produce a directional change in richness and indicator species, but not in vegetation structure. Lower montane forests differed from cloud forests at upper elevations. However, other factors should be included—mainly biogeographic affinities, historic and recent anthropogenic disturbance—to conclusively distinguish them. Montane forest can still be considered very heterogeneous and very high in beta diversity. Keywords Alpha diversity
Beta diversity
Elevation gradient
Indicator species
Lower montane forest
Upper montane forest Introduction The remarkably diverse physiognomy of tropical montane cloud forest (TMCF) has been acknowledged worldwide (Grubb 1977; Bruijnzeel et al. 2010). The forest structure and species composition of TMCF have been studied extensively, and related to changes in temperature, precipitation, seasonal rainfall, edaphic conditions, topography, natural and anthropogenic disturbance, forest area, and elevation (e.g., Grubb 1977; Gentry 1988; Lieberman et al. 1996; Bruijnzeel et al. 2010; Homeier et al. 2010; Martin et al. 2010; Bach and Gradstein 2011; López-Mata et al. 2012), and recently with climate change (TéllezValdés et al. 2006; Rojas-Soto et al. 2012). 123 Author's personal copy 686 Changes in structural characteristics and woody species richness have been related to elevation in TMCF. The most commonly reported trend is a decrease in the average height of trees and basal area of the stands (Tang and Ohsawa 1997; Homeier et al. 2010), although other studies have reported increased basal area (Kitayama 1992; Lieberman et al. 1996; Vázquez and Givnish 1998; Aiba and Kitayama 1999), and a hump-back curve with a tendency toward smaller basal areas with elevation (Reich et al. 2010). Several studies have reported that woody plant species richness decreased as elevation increased (Rincón 2007; Homeier et al. 2010; Salas-Morales and Meave 2012) and other studies report a unimodal pattern of species richness with elevation (Kitayama 1992; Sánchez-González and López-Mata 2005). Elevation trends in species richness vary among taxonomic groups and regions; recently the most supported trend is a peak in richness at midelevations (Rahbek 2005). Several authors have focused on detecting discrete limits to vegetation belts in tropical mountains (Holdridge et al. 1971; Grubb 1977; Hemp 2010; Martin et al. 2010; Bach and Gradstein 2011), and upper and lower montane forests have been recognized as formation types in several countries (Holdridge et al. 1971; Grubb 1977; Bruijnzeel et al. 2010). The most important factor determining the distribution of montane forests in tropical mountains is the frequency of fog (Grubb 1977). The limit between upper and lower montane rainforests may correlate well with the altitude of cloud formation, a high incidence of fog (Grubb 1977), and may be defined using mean annual temperature, mean annual precipitation, and altitudinal belts (Holdridge et al. 1971). Still, it is unclear whether it is possible to detect limits between belts. Some authors speculate that species composition changes continuously over the gradient, and therefore do not feel that tropical forest vegetation can be divided into discrete zones (Lieberman et al. 1996). Others have reported that discrete limits to vegetation belts in tropical mountains are usually lacking, however, the statistical analysis of species turnover is the most-effective method for detecting the elevation limits of vegetation (Bach and Gradstein 2011). Abrupt or discrete ecotones in vegetation patterns have been detected in some 123 Plant Ecol (2013) 214:685–701 tropical montane forests. In the mountains of the Dominican Republic, ecotones are produced and maintained primarily by catastrophic disturbances, such as fire and hurricanes (Martin et al. 2010), whereas on Mount Kilimanjaro, Tanzania, forest belts have been explained by the enduring influence of people, precipitation and the occurrence of frost or by the minimum temperature (Hemp 2010). In the TMCF of Mexico, there are no reports of ecotones controlled by natural or anthropogenic-related disturbance. In Mexico, the national vegetation classification scheme explicitly includes a category corresponding to TMCF, which is roughly equivalent to the term bosque mesófilo de montaña, sensu Rzedowski (1978) (Bruijnzeel et al. 2010; González-Espinosa et al. 2011). As described by Rzedowski (1978), bosque mesófilo de montaña is the most accepted classification of the humid forests. Rzedowski (1996) argued that TMCF is an individual vegetation type, because of the exclusive genera and family composition that have maintained its floristic integrity over time. Challenger (1998) chose to use this name for the temperate humid ecological zone, and the National Institute of Statistics and Geography (INEGI for its name in Spanish) recognized only one type of bosque mesófilo de montaña. Although there is no tradition of separating montane forest into upper and lower TMCF in the Mexican classification system, several authors have recognized subtypes of vegetation or associations, and differences in the species distribution along the elevation gradient, which depend on physiographic factors and the prevailing climate (Luna et al. 1989; Cházaro 1992; Rincón 2007; Challenger and Soberón 2008). This study was designed to look at changes in the structure of the vegetation, tree species composition and diversity in relation to changes in precipitation and temperature along the range of elevations, where tropical montane cloud forests grow in central Veracruz, Mexico. We investigated whether it is possible to detect—based on structure, composition, and diversity—a discrete limit between upper and lower montane forests. We hypothesized that with increasing elevation, there is a directional change in vegetation structure, and in the richness of species, genera and families, and that some species are indicators of the different types of forest. Author's personal copy Plant Ecol (2013) 214:685–701 687 Methods Study area The study area is located in the humid montane forest of central Veracruz, Mexico, a rare forest ecoregion within a biodiversity hotspot according to the WWF, with only 4 % of the forest in protected areas (Gillespie et al. 2012). In this region, 14 sites representing TMCF were selected based on the following characteristics: they should have an area of relatively undisturbed forest ([5 ha with no signs of heavy disturbance, except minor wood extraction), offer relatively easy access, and represent locations along the elevation gradient. Forest fragments were surrounded by land use types common to the region including pastures, crops, coffee plantations, commercial tree plantations, and secondary vegetation. Sites are located in the Eastern slopes of the Cofre de Perote, a shield volcano, the seventh highest peak in Mexico, with 4,282-m elevation. They are on the same lithological unit (except one site) on steep slopes exceeding 30° (Table 1). They are under the influence of the Trade Winds and at the cloud formation altitudes where orographic precipitation belt occurs (Williams-Linera 2007). At the lower and upper parts of the elevation gradient, mean annual temperatures are 18 and 12 °C, and annual precipitation values are 1,700 and 1,200 mm, respectively, with precipitation being the highest (2,200 mm) at the middle of the gradient. Climate information included 19 variables extracted from the WorldClim database (Hijmans et al. 2005) for each site. Downloaded climate data were verified in the field with data obtained from the nearest meteorological station on the elevation gradient; WorldClim data are adequate enough to describe climate variation among sites. A stepwise multiple linear regression was used as a tool to select variables predictive of species richness along the elevation gradient. Precipitation seasonality (coefficient of variation of monthly means), and the precipitation of the wettest quarter of the year and that of the warmest quarter of the year were selected using stepwise forward regression (JMP 6.0.0, SAS 2005). Two more climate variables were included in this study, because they have been used in other studies (annual precipitation and annual mean temperature). Vegetation structure, richness, and diversity In each site, we set up ten 10 9 10 m plots and in these plots, we measured all the trees C5-cm diameter at 1.3 m (dbh), counted the number of individuals, and identified species. Vouchers of fertile specimens were collected and deposited at the XAL herbarium. Basal area (m2/ha), density (individuals/ha), richness (S) and the Shannon Diversity Index (H0 , natural log) were Table 1 Characteristics of the study sites in central Veracruz, Mexico S. no N Latitude W Longitude Elevation (m a.s.l.) Slope (°) Aspect Lithology Precipitation (mm) T mean (°C) 1 19°300 96°560 1,250 36 W–NW Andesite 1,671 18.7 2 0 19°31 50.6 96°580 3.300 1,350 35 W Andesite 1,669 17.9 3 19°310 1500 96°590 2700 1,420 30 NW–N Andesite 1,708 17.8 4 19°320 96°580 1,450 12–33 W Andesite 1,669 17.9 5 6 0 19°35 19°310 0 1,475 1,483 19–31 5–30 SW SE–SW Andesite Limestone 1,836 1,653 16.8 18.1 7 19°300 8 0 19°30 40.8 00 96°58 96°570 00 97°00 1,500 31–35 S–W Andesite 1,926 17.2 97°00 5.700 1,630 32 SE Andesite 1,925 17.1 9 19°290 3700 97°00 4800 1,800 31 NE Andesite 2,160 16.5 10 19°330 4000 97°010 400 1,875 37 W–NW Andesite 1,607 14.6 11 19°410 0.900 96°510 15.100 1,900 31–35 NW–N Andesite 1,610 15.2 97°10 5400 1,950 37 NE Andesite 2,189 16.1 97°30 52.200 2,450 34 NW Andesite 1,118 12.4 97°30 3700 2,550 22 SE Andesite 1,189 11.8 0 00 12 19°29 23 13 19°310 43.300 14 0 00 19°31 4.3 Annual precipitation and annual mean temperature values were extracted from the WorldClim database (Hijmans et al. 2005) 123 Author's personal copy 688 calculated as descriptive measures of the tree community for each site. Richness was defined as the number of species. To compare tree species richness for a common number of individuals along the elevation gradient, we used individual-based rarefaction curves with the MaoTau richness function based on all species recorded in a 0.1-ha plot per site. The Shannon Index and rarefaction curves were computed using EstimateS ver. 8.0.0 (Colwell 2006). Species turnover along the elevation gradient was determined using the Chao’s Jaccard abundance-based similarity index (Chao et al. 2005; Colwell 2006) on a matrix of 85 species, excluding singletons. This similarity index includes not only species matching, but also similarity of relative abundances. It was calculated using EstimateS, with the upper abundance limit for rare species set to 5 and 200 bootstrap replicates. A value of 0 indicates complete dissimilarity and 1 means identical species composition between two sites. Data analysis Since common patterns for variation with elevation indicated that the response variables are monotonically or unimodally related to elevation, we fitted linear and polynomial equations to climate, vegetation structure, richness, and diversity. We used generalized linear models to test the patterns, and the best model was selected with Akaike’s information criterion using R project software (version 2.5.1 2007) (R Development Core Team 2007). When the response variable was measured in counts (number of taxa); a Poisson distribution was assumed and a log link function was used. In all cases, residual plots were checked to detect departures from model assumptions. Spearman correlation coefficients were calculated for bioclimate variables and basal area, density, Shannon’s Index, and richness. The presence of groups of sites was determined using non-metric multidimensional scaling (NMDS) ordination and cluster analysis in a matrix of the 14 sites, and for the 85 species with two or more individuals. This analysis was carried out using the autopilot mode, slow and thorough, Sørensen distance measurements, 40 runs with real data, and 400 iterations. The cluster analysis was performed using Sørensen (Bray–Curtis) distance measurements with the flexible beta linkage (beta = -0.25) method as 123 Plant Ecol (2013) 214:685–701 recommended by McCune and Grace (2002). Indicator species analysis (ISA) was used as a quantitative, objective criterion to prune the dendrogram resulting from the hierarchical clustering. The cluster step yielding the smallest average p value together with the highest number of significant indicator species provided the basis for choosing the optimum number of groups (McCune and Grace 2002). ISA shows the most representative tree species in each of the groups detected in the cluster analysis and NMDS. ISA yields an indicator value and a statistical significance for this value using a Monte Carlo technique based on 1,000 randomizations. Differences in tree composition among groups of sites were tested with a multiresponse permutation procedure (MRPP; McCune and Grace 2002). The test statistic (T) describes the separation between groups (more negative values reflect stronger separation) and the chance corrected within group agreement (A). When all the species within groups are identical, A reaches its maximum value (A = 1). When the heterogeneity within groups equals the level expected by chance, then A = 0, and when there is more heterogeneity within groups than the level expected by chance, then A \ 0. In MRPP a p value is given for each test group comparison. Classification, ordination, and statistical tests were conducted using PC-ORD software (McCune and Grace 2002). Results Tree species composition Along the 1,400 m of the elevation gradient, a total of 1,639 trees was recorded in 14 sites (1.4 ha). These belong to 128 tree species, 76 genera, and 47 families, including seven morphospecies (see Table 4 in Appendix). The most important families in terms of tree species were Fagaceae (13 species), Fabaceae (7), Lauraceae (8), Rosaceae (6), and Rubiaceae (6). Remarkably, Quercus (12) and Prunus (6) have several species, while most genera have less than three species. The species present in most sites were Oreopanax xalapensis (10 sites), Liquidambar styraciflua (9), Carpinus tropicalis (8), Clethra macrophylla (8), Quercus lancifolia (7), Styrax glabrescens (7), and Turpinia insignis (7). These species were also abundant, accounting for 32 % of the total number of Author's personal copy Plant Ecol (2013) 214:685–701 689 Vegetation structure, richness, and diversity 80 a 76.1 77.5 78.2 70 Model 1 Χ 2 = 9.8, P = 0.002 600 c 156.6 500 400 300 161.4 158.6 Model 2 Χ 2 = 15.8, P = 0.0004 Mean temperature (°C) 200 e 20 b 1200 1000 177.6 180.6 800 185.7 600 400 60 Annual precipitation (mm) Precipitation of warmest quarter (mm) Precipitation seasonality In the region along the elevation gradient, precipitation seasonality (Fig. 1a) increased with elevation, whereas the precipitation of the wettest quarter (Fig. 1b), precipitation of the warmest quarter (Fig. 1c), and total annual precipitation (Fig. 1d) were unimodal with a maximum around 1,800–2,000 m a.s.l. Mean temperature had a decreasing monotonic pattern (Fig. 1e). Along the elevation gradient, basal area, density, and the Shannon Index had a nonsignificant trend (Fig. 2a–c); however, the richness of species, genera and families had a unimodal pattern with a peak around 1,800–2,000 m a.s.l. (Fig. 2d–f). When abiotic and biotic variables were correlated only the precipitation of the wettest quarter was positively correlated with the richness of species, genus and family (q = 0.56, 0.54, and 0.60, respectively, p \ 0.05). The species rarefaction analysis was used to compare richness among the sites located at different elevations. Rarefaction curves (76 individuals) separated the forest sites into two groups. A general trend Precipitation of wettest quarter (mm) individuals. In addition, we found a number of locally abundant species that were present in only one (Oreomunnea mexicana) or two (Fagus grandifolia) sites. Model 2 Χ 2 = 10.0, P = 0.007 d 2000 191.1 1500 193 200.5 Model 2 Χ 2 = 14.4, P = 0.0008 1000 1200 1700 2200 2700 Elevation (m a.s.l.) 29.4 27.5 15 26.2 10 1200 Model 1 Χ 2 = 39.3, P < 0.0001 1700 2200 2700 Elevation (m a.s.l.) Fig. 1 Models fitted to abiotic variables along the elevation gradient in the TMCF region of central Veracruz, Mexico. a precipitation seasonality or coefficient of variation of monthly means, b precipitation of the wettest quarter of the year, c precipitation of the warmest quarter of the year, d annual precipitation, and e mean annual temperature. Model 1, y = a ? bx; model 2, y = a ? bx ? cx2; model 3, y = a ? bx ? cx2 ? dx3. V2 and P are the results of the model with the best fit. Numbers are AIC for each model, with the best model in bold (DAIC = 0) 123 Author's personal copy 690 Plant Ecol (2013) 214:685–701 a b 1700 Density (trees/ha) Basal area (m2/ha) 100 132.3 80 130.7 60 131.2 40 n.s. 205 1200 16.8 16.4 Species richness Shannon Index (H') 30 c 3 17.2 2 n.s. 1 25 79 78.1 20 15 82.2 Model 2 Χ 2 = 6.3, P = 0.04 25 e 79 77.4 20 Family richness Genus richness d 10 25 86.9 15 Model 2 Χ 2 = 11.5, P = 0.003 5 1200 205.9 n.s 700 20 10 203.3 1700 2200 2700 Elevation (m a.s.l.) f 75 20 73.9 82.3 15 10 Model 2 Χ 2 = 10.8, P = 0.005 5 1200 1700 2200 2700 Elevation (m a.s.l.) Fig. 2 Models fitted to biotic variables along the elevation gradient in the TMCF region of central Veracruz, Mexico. a basal area, b density, c Shannon Index, d species richness, e genus richness, and f family richness. Model 1, y = a ? bx; model 2, y = a ? bx ? cx2; model 3, y = a ? bx ? cx2 ? dx3. V2 and P are the results of the model with the best fit. Numbers are AIC for each model, with the best model in bold (DAIC = 0) indicated that sites located at higher elevations (sites 8, 10, 11, 12, and 13) were richer than those at lower elevations (sites 1, 3, 4, and 7). However, sites 2 and 5 were in the former group, and site 9 was in the latter (Fig. 3). The site (14) at the highest elevation had the lowest rarefied richness. Beta diversity calculated with the Chao-Jaccard Index varies between 0 and 0.72 (Table 2) and also suggests two trends along the elevation gradient. Sites (1–8, except site 6) located at lower elevations were consistently more similar among themselves than those at higher elevations. High turnover in species composition was observed among the sites located at the highest elevation (sites 9–13), except the Fagus forests (10 and 11, 0.40), and the highest elevation forests (0.42). However, species turnover was close to 100 % between pairs of sites at extreme opposite elevations on the gradient (Table 2). 123 Ordination and classification For the NMDS ordination, the greatest reduction in ‘‘stress’’ was achieved with a two-dimensional solution (final stress = 9.58, final instability \0.000001, Author's personal copy Plant Ecol (2013) 214:685–701 691 25 Richness 20 15 10 5 12 5 6 7 1 11 2 13 4 14 10 8 9 3 100 120 140 160 0 20 0 40 60 80 180 Number of individuals Fig. 3 Rarefaction curves for number of tree species in 14 forest sites located along the elevation gradient in central Veracruz, Mexico. The vertical line indicates common abundance level at which richness among sites was compared. Dashed lines are sites at [1,600 m a.s.l.; solid lines are sites at \1,600 m a.s.l Table 2 Chao-Jaccard Index between pairs of forest sites located along the elevation gradient in central Veracruz, Mexico 1 2 3 4 5 6 2 3 4 5 6 7 8 9 10 11 12 13 14 0.45 0.57 0.68 0.57 0.15 0.28 0.26 0.03 0.05 0.13 0.06 0.01 0.01 0.65 0.66 0.40 0.11 0.35 0.21 0.02 0.04 0.09 0.05 0.01 0 0.72 0.63 0.28 0.37 0.34 0.10 0.12 0.10 0.14 0.01 0 0.51 0.14 0.42 0.30 0.03 0.07 0.13 0.05 0 0 0.20 0.44 0.32 0.07 0.13 0.20 0.12 0.02 0.01 0.10 0.03 0.02 0.03 0.17 0.04 0.01 0.01 0.31 0.06 0.25 0.11 0.07 0.01 0.04 0.13 0.10 0.07 0.34 0.01 0.02 0.06 0.03 0.16 0.02 0.00 0.40 0.18 0.06 0.08 7 8 9 10 11 12 13 0.16 0.06 0.12 0.07 0.19 0.42 Jaccard values vary from 0—which represents complete dissimilarity to 1—which represents complete similarity in tree species composition. Values in bold type indicate less beta diversity between sites 44 iterations). The proportions of variance represented by Axis 1 and Axis 2 were 0.31 and 0.48, respectively. The Monte Carlo test indicated that the extracted axes were significantly different from those expected by chance (p = 0.02). The NMDS ordination strongly suggests that there is a sharp division between upper and lower montane forest sites. The NMDS clearly separated sites located at lower elevations from those located at high elevations along Axis 2 (Fig. 4). Furthermore, the high elevation forests were separated along Axis 1 into: site 9 located to the left of Axis 1 and dominated by Oreomunnea mexicana, sites located at the highest elevations (sites 12–14) and the Fagus grandifolia forests (sites 10 and 11) in the center of Axis 1. Also site 6, a lower elevation forest, which lacks regionally common species but has regionally uncommon species, was located at the right end of Axis 1 (Fig. 4). The presence of recognizable groups was confirmed by the cluster analysis. The dendrogram from 123 Author's personal copy 692 Plant Ecol (2013) 214:685–701 13 14 cle qcr prs NMDS 2 prr dri syg tuo cls 12 ore 9 8 10 wei val 7 qsa liq pod fag 11 ter pam orx 5 clm qla 3 car 2 tui qxa 4 1 cor plo 6 upper lower Fagus Oreomunnea limestone NMDS 1 Fig. 4 Nonmetric multidimensional scaling (NMDS) for the 14 forest sites along the elevation gradient in central Veracruz, Mexico. Numbers are forest sites, acronyms are species with maximum indicator values C0.50 according to ISA; car, Carpinus tropicalis; clm, Clethra macrophylla; cls, Clethra schlechtendalii; cle, Cleyera integrifolia; cor, Cornus excelsa; dri, Drimys granadensis; fag, Fagus grandifolia var. mexicana; liq, Liquidambar styraciflua; ore, Oreomunnea mexicana; orx, Oreopanax xalapensis; pam, Persea americana; plo, Persea longipes; pod, Podocarpus matudai; prr, Prunus rhamnoides; prs, Prunus samydoides; qcr, Quercus corrugata; qla, Quercus lancifolia; qsa, Quercus sartorii; qxa, Quercus xalapensis; syg, Symplocos longipes; ter, Ternstroemia sylvatica; tui, Turpinia insignis; tuo, Turpinia occidentalis; val, Vaccinium leucanthum; wei, Weinmannia pinnata the cluster analysis of the sites was pruned at three groups (using ISA as the quantitative, objective criterion), excluding sites 6, 9, and 13 (Fig. 5). The first group included sites located at low elevations (sites1–8, except site 6), a second group included Fagus forests (sites 10 and 11), and the third one included sites located at the highest elevation on the gradient (sites 12 and 14, Fig. 5). This level of grouping provided a good compromise between loss of information (about 50 % retained) and groups of sites. Clustering the sites into the three groups provided the maximum separation between groups, and the heterogeneity within groups tends to equal what one would expect by chance (MRPP, T = -4.22, A = 0.19, p = 0.0002). The ISA identified ten species as strong indicators of the groups (p \ 0.05, Table 3). Discussion Elevation is a good proxy for several climate variables. Along elevation gradients, precipitation-related variables and mean temperature have been identified 123 Fig. 5 Cluster analysis dendrogram of the sites located between 1,250 and 2,550 m a.s.l. in central Veracruz, Mexico, using Sørenson distance and flexible beta (-0.25) linkage method, cut at 50 % of the information remaining scale as the important factors controlling forest type distribution on mountains and appear to define ecological elevation turnover points or critical elevations (Holdridge et al. 1971; Grubb 1977; Kitayama 1992; LópezMata et al. 2012; Salas-Morales and Meave 2012). In Mexican forests, Salas-Morales and Meave (2012) indicated that temperature may be a critical factor involved in a decrease in species richness at 1,800 m a.s.l. The upper limit of the lower montane zone has been correlated with an abrupt increase in surplus water and may be correlated with the altitude of cloud formation (Grubb 1977; Kitayama 1992). In addition, López-Mata et al. (2012) reported that species richness in the TMCF is explained by correlated variables of rainfall in the humid months of the year, seasonal rainfall, annual evapotranspiration and elevation. Factors related to elevation changes and that are likely controlled by seasonality in precipitation are important determinants of changes in vegetation structure, richness, and diversity, even across a relatively short TMCF gradient. However, on our TMCF gradient, precipitation, and temperature were not correlated with vegetation structure. There was a nonsignificant trend in density and basal area to increase with elevation, as reported for La Chinantla, Oaxaca, Mexico (Rincón 2007), in contrast to reported trends Author's personal copy Plant Ecol (2013) 214:685–701 693 Table 3 Three groups of forest communities determined by indicator species analysis Species Maximum indicator value P Carpinus tropicalis 0.61 0.152 Clethra macrophylla 0.87 0.004 Liquidambar styraciflua 0.86 0.009 Quercus lancifolia 1.00 0.004 Quercus sartorii 0.61 0.233 Quercus xalapensis 0.57 0.284 Turpinia insignis 1.00 0.004 Chiococca pachyphylla 0.50 0.371 Citharexylum ligustrinum 0.50 0.371 Cleyera integrifolia 0.67 0.109 Cornus excelsa 0.50 0.371 Drimys granadensis 0.50 0.355 Fagus grandifolia 1.00 0.035 Oreopanax xalapensis 0.66 0.113 Persea americana 1.00 0.035 Podocarpus matudai 0.50 0.371 Prunus brachybotrya 0.50 0.371 Prunus samydoides 0.50 0.355 Rhamnus capreifolia 0.50 0.371 Ternstroemia sylvatica 0.96 0.026 Vaccinium leucanthum 1.00 0.035 Viburnum tiliifolium 0.50 0.355 Weinmannia pinnata 0.76 0.114 Clethra schlechtendalii 0.50 0.397 Cupressus lusitanica 0.50 0.343 Phyllonoma laticuspis 0.50 0.397 Pinus ayacahuite 0.50 0.343 Pinus patula 0.50 0.343 Prunus rhamnoides 0.97 0.043 Quercus corrugata 0.95 0.044 Quercus glabrescens 0.50 0.397 Sambucus nigra var. canadensis 0.50 0.343 Symplocos longipes 0.50 0.397 Turpinia occidentalis 0.50 0.397 Group 1 Lower montane forest Group 2 Forests with Fagus Group 3 Upper montane forest Species, maximum indicator values C0.50, and P-values were calculated using a Monte Carlo permutation test for each species. Observed maximum indicator values varied between 0 and 1 about decreasing (Tang and Ohsawa 1997; Homeier et al. 2010) or increasing basal area in other regions (Kitayama 1992; Lieberman et al. 1996; Vázquez and Givnish 1998; Aiba and Kitayama 1999). Although there was no trend detected for vegetation structure, consistent unimodal patterns were found for species, genus, and family richness and these were related to precipitation seasonality. The results indicate that the richness of tree species, genus and family increased between 1,250 and 2,000 m a.s.l., and then decreased toward the site located at 2,550 m a.s.l., whereas rainfall seasonality increased with elevation. The trend of increasing richness with elevation generally refers to studies that include a partial elevation gradient, but the decrease that follows may indicate that these TMCF sites represent the top of a hump-shaped distribution (Toledo 2013). Therefore, scale is something relevant to consider when comparing studies and to infer how structure and richness change in the elevation range. A diversity peak at an intermediate elevation along elevation gradients may be a common, and perhaps even the general pattern (Lomolino 2001; Rahbek 2005). Several studies have found this unimodal pattern in species richness with high values at midelevations (Tang and Ohsawa 1997; Vázquez and Givnish 1998; Lomolino 2001; Oommen and Shanker 2005; Sánchez-González and LópezMata 2005; Grytnes and Beaman 2006; Wang et al. 2007). Other studies have reported a decrease in woody species richness with elevation (Gentry 1988; Kitayama 1992; Lieberman et al. 1996; Kappelle and Zamora 1995; Aiba and Kitayama 1999; Grytnes and Beaman 2006; Behera and Kushwaha 2007; Rincón 2007; Homeier et al. 2010; Salas-Morales and Meave 2012). In some cases, other patterns have been detected. For instance, the elevation patterns of species richness on Mount Kinabalu, Borneo were reported as hump-shaped for all species, fern species and epiphytic species (Grytnes and Beaman 2006). However, the species richness of trees showed a monotonically decreasing trend from the lowest elevations to the summit (Aiba and Kitayama 1999). In Mexico, in the Sierra de Manantlan, Jalisco, the richness of tree species apparently did not vary with elevation (Vázquez and Givnish 1998), and in the Eastern Himalaya, the two maxima in species number that have been reported correspond to the transition zones between the two forest types (Behera and Kushwaha 2007). 123 Author's personal copy 694 Beta diversity increased along the elevation gradient. The turnover of species was lower between sites located at lower elevations, but increased with elevation. Species replacement was complete between sites located at the extremes of the gradient. Similarly, beta diversity increased steadily and species replacement was nearly 100 % along the elevation gradient in Sierra Nevada, Mexico (Sánchez-González and López-Mata 2005). This complete dissimilarity supports the trend of different associations located in the lower and upper parts of the studied gradient. There is convincing evidence of floristic patterns across our study area. Two groups of forests defined by elevation and tree species composition were detected in this study. These results, supported by the cluster analysis and NMDS, indicate that groups of sites can reasonably be interpreted as lower and upper montane forest sites, occurring from 1,250 to 1,630 and from 1,800 to 2,550 m a.s.l., respectively. In addition, subgroups of forest sites can be differentiated with the combined results of the rarefaction, beta diversity and ordination analyses. Bach and Gradstein (2011) reported that for local purposes, cluster analysis and structure-based vegetation analysis appear to be suitable tools for a preliminary approach to detecting elevation belts. We found that basal area and tree density did not identify elevation belts; rather, species composition was more useful. The elevation patterns in species, genus, and family richness were unimodal and peaked at almost the same elevations, therefore, allowing distinct associations can be recognized. The lower elevation (1,200–1,650 m a.s.l.) forests were less diverse than those at upper elevations (except for a forest located at the highest elevation in the gradient), and among them there was a high degree of similarity in species composition and thus lower beta diversity. The indicator species were consistently the most common trees in the region, such as Clethra macrophylla, Liquidambar styraciflua, and Turpinia insignis, along with some species of Quercus. The upper montane forests are more diverse than the lower forests, and species common in the lower montane forests were less abundant or not found at the upper elevations. In the upper montane region, a group of forests had Prunus rhamnoides and Quercus corrugata as indicator species, and species such as Cleyera theaeoides, Ternstroemia sylvatica and Weinmannia intermedia were well represented in upper 123 Plant Ecol (2013) 214:685–701 sites. Part of the upper montane forests included the monodominant Fagus grandifolia stands located at 1,800–1,900 m a.s.l. Fagus is a locally dominant tree, but geographically it is a rare species in the TMCF of Mexico. Fagus forests are characterized by particular microenvironmental conditions that make them different from the others (Williams-Linera et al. 2003; Téllez-Valdés et al. 2006). The high heterogeneity in the ecoregion is exemplified by three sites that were not clustered, and were identified as composed of forests with distinctive characteristics. Two sites were part of the upper montane region: one with the evident presence of three species of pines resulting from proximity to the conifer forest ecotone, and another with Oreomunnea mexicana. This forest had the highest basal area values, and Oreomunnea trees accounted for 50 % of the dominance in the community. The precipitation recorded in this stand is among the highest for the region, though not as high as the annual precipitation reported for another Oreomunnea forest in Oaxaca (5,000–6,000 mm), where it was the absolute dominant element in the upper tree layer (Rzedowski and Palacios-Chávez 1977). The third site is located in the lower montane region, indicating that this region is also heterogeneous. This forest grows on a limestone outcrop at the top of a hill, and physiognomically resembles a dwarf forest with more epiphytes and a different tree species composition than the other forests located at similar elevations. We found regionally rare species (Cercis canadensis, Clusia guatemalensis and Quercus pinnativenulosa) at this site, which was notably lacking the species that are abundant in the neighboring lower forests (e.g., Liquidambar styraciflua). Some studies have shown that limestone karst substrates support different community types with distinctive species composition and vegetation structure (Rivera et al. 2000; Aukema et al. 2007). Conclusions Our results indicate that the variation in precipitation with increasing elevation does not clearly result in a directional change in vegetation structure, however, changes in the richness of taxa and indicator species were directional. Also, the results of this study represent notable progress regarding the question of Author's personal copy Plant Ecol (2013) 214:685–701 695 recent disturbance, biogeographic affinities, and the phylogenetic distribution of the taxa. Meanwhile, the montane forest can continue to be considered very heterogeneous and high in beta diversity. whether there are limits between forest types and whether these can be described in terms of their vegetation structure and tree species composition. The lower montane forest of central Veracruz is a distinct vegetation type as it is less diverse and has common and widely distributed tree species, whereas the upper montane forest is more diverse, includes Fagus and Oreomunnea forests, and seems to be the true cloud forest (sensu Cházaro 1992; Bruijnzeel et al. 2010). Although it is possible to distinguish variations and natural types of vegetation at both elevations, the presence of other communities demonstrate the high local heterogeneity. In the region, changes in richness and diversity were observed, but it is likely that other factors should be incorporated to define limits. The trends reported here need to be examined in light of other factors besides climate, such as historical and Acknowledgments The authors grateful to Susana Valencia Avalos for the valuable help with Quercus taxonomy, and to Francisco Lorea in identifying the specimens of Lauraceae. The authors thank Libertad Sanchez for providing access to the unpublished data from the El Mirador site. The authors also thank the owners of all the study sites for allowing us to conduct this research and for protecting their forests. Appendix See Table 4. Table 4 List of tree species in forest sites located between 1,250 and 2,550 m a.s.l. in central Veracruz, Mexico Species Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Saurauia leucocarpa Schltdl. – – – – – – 15 6 – 1 – – – – Saurauia pedunculata Hook. 1 4 – – – – – – – – 1 – – – Sambucus nigra var. canadensis (L.) B. L. Turner – – – – – – – – – – – – 1 2 Viburnum tiliifolium (Oerst.) Hemsl. – – – – – – – – 1 – – 1 10 4 14 12 3 – 34 6 7 4 – – – – – 4 – – – – – – – – – – – – 3 – – 2 2 4 – – – – 4 – – – Ilex sp. 1 – – – – – – – – – – – – – 1 Ilex sp. 2 – – – – – – – – – – – 1 – – Oreopanax liebmannii Marchand – – – – – 2 – – – – – – – – Oreopanax xalapensis (Kunth) Decne. and Planch. – – – 1 5 1 2 1 3 5 2 1 1 – – – – – – – 1 – – – – – Koanophyllon pittieri (Klatt.) R.M. King and H. Rob. – 2 1 – 1 1 – – – – – – – – Telanthophora grandifolia (Less.) H. Rob. and Brettel – – – – – 1 – – – – – – – – Actinidiaceae Adoxaceae – Altingiaceae Liquidambar styraciflua L. Annonaceae Annona cherimola Mill. Aquifoliaceae Ilex discolor var. tolucana (Hemsl.) Edwin ex Linares Araliaceae Asteraceae Eupatorium sp. 123 Author's personal copy 696 Plant Ecol (2013) 214:685–701 Table 4 continued Species Site number 1 Verbesina sp. 2 3 4 5 6 7 8 9 10 11 12 13 14 – 1 – – – – – – – – – – – – – – – – – – – – – – 1 – – Betulaceae Alnus acuminate Kunth – Carpinus tropicalis (Donn.Sm.) Lundell 36 12 31 29 18 12 – 3 – – 15 – – Ostrya virginiana (Mill.) K. Koch – – – – 5 – – – – – – – – – – – – – – 1 – – – – – – – – 1 1 – – – – – – – – – – – Brunelliaceae Brunellia mexicana Standl. Cannabaceae Trema micrantha (L.) Blume Celastraceae Euonymus mexicanus Benth. – – – – – – – – – – 1 – – – Wimmeria concolor Schltdl. and Cham. – – – – 5 – – – – – – – – – Morphospecies 1 – – – – – – – – – 1 – – – – – – – – – – – 40 – – – 1 – – Clethra alcoceri Greenm. – – – – – – – – – – 4 – – 2 Clethra macrophylla M.Martens and Galeotti 5 3 12 4 5 – 16 3 – 2 – – – – Clethra schlechtendalii Briq. – – – – – – – – 3 – – 7 – – – – – – – 1 – – – – – – – – – – – – – 4 – – – – 2 – – – – – – – – – 1 – – 1 3 1 1 – – – – – – – – – – – – – 1 4 – – – – – – – – – 1 – 1 – – – – – – – – – – – – – – 1 – – – – – – – – – 1 5 3 – – – – – 1 – – – – 6 13 – – 14 – – Chloranthaceae Hedyosmum mexicanum C. Cordem Clethraceae Clusiaceae Clusia guatemalensis Hemsl. Cornaceae Cornus excelsa Kunth Cunoniaceae Weinmannia pinnata L. Cupressaceae Cupressus lusitanica Mill. Dipentodontaceae Perrottetia ovata Hemsl. Ericacae Gaultheria acuminata Schltdl. and Cham. Vaccinium leucanthum Schltdl. Euphorbiaceae Alchornea latifolia Sw. Bernardia macrocarpa A. Cerv. and Flores Olv. – – – – – – – – 1 – – – – – Cnidoscolus multilobus (Pax) I.M. Johnst. – 3 – – – – – – – – – – – – Gymnanthes longipes Mull. Arg. – – – 1 – – – – – 4 – – – Fabaceae Acacia pennatula (Schltdl. and Cham.) Benth. – 1 – – – – – – – – – – – – Cercis canadensis L. – – – – – 1 – – – – – – – – Cojoba arborea (L.) Britton and Rose – – – 5 3 1 – 4 – – – – Inga sp. 1 – – – – – – – – – – – 1 – – – – – – – – – – – – 1 – – Inga sp. 2 123 Author's personal copy Plant Ecol (2013) 214:685–701 697 Table 4 continued Species Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Leucaena leucocephala (Lam.) de Wit – 3 – – – – – – – – – – – – Lonchocarpus guatemalensis Lundell – – – – 1 – – – – – – – – – Fagus grandifolia var. mexicana (Martı́nez) Little – – – – – – – – – 18 34 – – – Quercus acherdophylla Trel. – – – – – – – – – – – – – 1 – Fagaceae Quercus acutifolia Née 2 – – – – – 2 – – – – – – Quercus corrugata Hook. 1 – – – – – – – – – 1 12 – 13 Quercus cortesii Liebm. – – – – – – – 17 3 – – – – – Quercus crassifolia Bonpl. – – – – – – – – – – – – 5 – Quercus delgadoana S. Valencia, Nixon and L.M. Kelly – – – – – – 10 – – 14 6 9 – 13 Quercus germana Schltdl. and Cham. 13 5 – 11 – – – – – – – – – – Quercus glabrescens Benth. – – – – – – – – – – – 1 12 – Quercus lancifolia Schltdl. and Cham. Quercus pinnativenulosa C.H. Mull. – 2 – – 31 – – – – – – – – – Quercus sartorii Liebm. 4 – – 4 4 – 2 7 – – – 1 – – Quercus xalapensis Bonpl. – 22 6 14 – – 6 – – – – – – – – – – – – – – – 115 – – – – – Cinamommum effusum (Meisn.) Kosterm. 5 4 2 – 3 – – – – – – 2 1 – Beilschmiedia mexicana (Mez) Kosterm. – – – 3 – – – – – – – – – – Litsea glaucescens Kunth Ocotea disjuncta Lorea-Hern. – – – – – – – – – – – – – – – – – 1 – – – – – – 1 – – – Ocotea effusa (Meissn.) Hemsl. – – – – – – – – – – – 1 – – Juglandaceae Oreomunnea mexicana (Standl.) J.F. Leroy Lauraceae Ocotea psychotrioides Kunth. – 3 – – 4 – – 2 – – – – – – Persea americana Mill. – – – – – 1 – – – 1 2 – – – Persea longipes (Schltdl.) Meisn. – – – – – 4 – – – – – – – – – – – – 4 – – – 1 1 1 2 – – Hampea integerrima Schltdl. – – – – – 1 – – – – – – – – Heliocarpus donnellsmithii Rose ex Don. Sm. – – – 1 – – – – – 2 – – – – Magnoliaceae Magnolia schiedeana Schltdl. Malvaceae Melastomataceae Conostegia arborea (Schltdl.) Steud. – – – – – – – 2 – – – – – – Miconia glaberrima (Schltdl.) Naudin – – 4 – 4 – – 2 – 7 – 6 – – Miconia mexicana (Bonpl.) Naudin – – – – – – 3 – – – – – – – Miconia lonchophylla Naudin Miconia oligotricha (DC.) Naudin – – – – – – – – – – – – – – – – 1 – – – – – – – – 1 – – – – – – – – 2 – – – – – – – – – – – – – – – 1 – – – – – Monimiaceae Mollinedia viridiflora Tul. Moraceae Trophis cf. cuspidata Lundell 123 Author's personal copy 698 Plant Ecol (2013) 214:685–701 Table 4 continued Species Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 – 4 5 1 1 1 – – – – – 4 – – Eugenia mexicana Steud. 5 – – – – – – – – – – – – – Eugenia xalapensis (Kunth) DC. 2 – – – – – – – – – – – – – Primulaceae Myrsine coriacea (Sw.) R.Br. ex Roem. and Schult. Myrtaceae Pentaphylacaceae Cleyera integrifolia (Benth.) Choisy – – – – – – – – 3 2 2 2 10 – Ternstroemia sylvatica Schltdl. and Cham. – – – – 2 – – – – 5 7 – – – – – – – – – – – 3 – – 3 – – Phyllonomaceae Phyllonoma laticuspis (Turcz.) Engl. Pinaceae Pinus ayacahuite Ehrenb. ex Schltdl. – – – – – – – – – – – – 36 4 Pinus patula Schiede ex Schltdl. and Cham. – – – – – – – – – – – – 5 83 Pinus pseudostrobus Lindl. – – – – – – – – – – – – 5 – – – – – – – – – – – 5 – – – Podocarpaceae Podocarpus matudae Lundell Rhamnaceae Rhamnus capreifolia Schltdl. – – – – – – – – – – 3 – – – Rhamnus longistyla C.B. Wolf – – – – – – – – – 1 – – – – Rhamnus macvaughii L.A.Johnst. and M.C. Johnst. – – – – – – – – – – – – 1 – Prunus brachybotrya Zucc. Prunus rhamnoides Koehne – – – – – – – – – – – – – – – 2 – – – – 2 – – 13 1 5 – – Prunus samydoides Schltdl. – – – – – – – – – 3 – – – – Prunus tetradenia Koehne – Rosaceae – – – – – – – 4 – – – – – Prunus sp. 1 – – – – – – – – – – – 1 – Prunus sp. 2 – – – – – – – – – – – 1 – – – – 1 5 32 – – – 2 – – – Rubiaceae Arachnothryx capitellata (Hemsl.) Borhidi – Chiococca pachyphylla Wernham – – – – – – – – – – 8 – – – Deppea grandiflora Schltdl. – 2 – – – 1 – – – – – – – – Palicourea padifolia (Willd. ex Roem. and Schult.) C.M. Taylor and Lorence – – – 1 1 – 3 – – – – – – – Psychotria galeottiana (M. Martens) C.M. Taylor and Lorence – – – – – – – 1 2 – – – – – Morphospecies 2 – – – – – – 1 – – – – – – – – – Rutaceae Zanthoxylum aff. flavum Vahl. – – – – 1 – – – – – – – Zanthoxylum melanostictum Schltdl. and Cham. – – 7 – 1 – – 3 9 – – 5 – Zanthoxylum sp. 1 – – – – – – – 2 – – – – – Zanthoxylum sp. 2 – – – – – – – – 2 – – – – Sabiaceae 123 Author's personal copy Plant Ecol (2013) 214:685–701 699 Table 4 continued Species Site number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Meliosma alba (Schltdl.) Walp. – 1 – – – – – – – – – – – – Meliosma dentata (Liebm.) Urb. – – – – – – – – – – – – 3 – Solanaceae Solanum nigricans M. Martens and Galeotti – – – – – – – 1 – 1 – – Witheringia sp. – – – – – – – – 1 – – – Morphospecies 3 – – – – – – – – – – – – 1 Morphospecies 4 – – – – – – – – – – – 1 – Staphyleaceae Turpinia insignis (Kunth) Tul. 19 15 15 6 13 – 17 3 – – – – – – Turpinia occidentalis (Sw.) G. Don – – – – – – – 1 – – 3 – – – – 3 2 9 – – 5 6 – – 2 3 – – – – – – – – – – 1 – – – – – 1 – – 1 – – 9 – 2 2 1 – – – – – – – – – – – – – – – 5 Styracaceae Styrax glabrescens Benth. Symplocaceae Symplocos limoncillo Bonpl. Symplocos longipes Lundell Taxaceae Taxus globosa Schltdl. Verbenaceae Citharexylum ligustrinum Van Houtte – – – – – – – – – – 3 – – – Citharexylum mocinoi D. Don. – 6 – – – – – – – – – – – – – – – – – – – – – 2 – – – – Morphospecies 5 1 – – – – – – – – – – – – – Morphospecies 6 – 1 – – – – – – – – – – – – Morphospecies 7 – – – 1 – – – – – – – – – – Morphospecies 8 – – – 1 – – – – – – – – – – Morphospecies 9 – – – – – – – 1 – – – – – – Morphospecies 10 – – – – – – – – 1 – – – – – Morphospecies 11 Number of species – 16 – 23 – 15 – 16 – 23 – 19 – 18 – 21 – 24 – 20 – 25 2 28 – 19 – 16 Number of individuals per 0.1 ha 109 106 112 117 99 84 148 126 180 75 129 103 92 134 Winteraceae Drimys granadensis L.f. Family unknown Sites range from the lowest to highest elevation; see characteristics in Table 1. Values are number of trees C5-cm dbh counted in 0.1 ha at each site References Aiba S, Kitayama K (1999) Structure, composition and species diversity in an altitude-substrate matrix of rain forest tree communities on Mount Kinabalu, Borneo. Plant Ecol 140:139–157 Aukema JE, Carlo TA, Collazo JA (2007) Landscape assessment of tree communities in the northern karst region of Puerto Rico. Plant Ecol 189:101–115 Bach K, Gradstein R (2011) A comparison of six methods to detect altitudinal belts of vegetation in tropical mountains. Ecotropica 17:1–13 Behera MD, Kushwaha SPS (2007) An analysis of altitudinal behavior of tree species in Subansiri district, Eastern Himalaya. Biodivers Conserv 16:1851–1865 Bruijnzeel LA, Scatena FN, Hamilton LS (2010) Tropical montane cloud forests. Science for conservation and management. Cambridge University Press, Cambridge 123 Author's personal copy 700 Challenger A (1998) Utilización y conservación de los ecosistemas terrestres de México. Pasado, presente y futuro. CONABIO, Instituto de Biologı́a, Sierra Madre, México Challenger A, Soberón J (2008) Los ecosistemas terrestres. In: Sarukhán J et al (eds) Capital natural de Mexico, Conocimiento actual de la biodiversidad, vol 3. CONABIO, Mexico, pp 87–108 Chao A, Chazdon RL, Colwell RK, Shen T (2005) A new statistical approach for assessing similarity of species composition with incidence and abundance data. Ecol Lett 8:148–159 Cházaro M (1992) Exploraciones botánicas en Veracruz y estados circunvecinos. I. Pisos altitudinales de vegetación en el centro de Veracruz y zonas limı́trofes con Puebla. La Ciencia y el Hombre 10:67–115 Colwell RK (2006) EstimateS: statistical estimation of species richness and shared species from samples. Version 8.0.0. User’s guide and application published at http://purl.oclc. org/estimates Gentry AH (1988) Changes in plant community diversity and floristic composition on environmental and geographical gradients. Ann Mo Bot Gard 75:1–34 Gillespie TW, Lipkin B, Sullivan L, Benowitz DR, Pau S, Keppel G (2012) The rarest and least protected forests in biodiversity hotspots. Biodivers Conserv 21:3597–3611 González-Espinosa M, Meave JA, Lorea-Hernández F, IbarraManrı́quez G, Newton AC (2011) The red list of Mexican cloud forest trees. Fauna & Flora International, Cambridge Grubb PJ (1977) Control of forest growth and distribution on wet tropical mountains: with special reference to mineral nutrition. Annu Rev Ecol Syst 8:83–107 Grytnes JA, Beaman JH (2006) Elevational species richness patterns for vascular plants on Mount Kinabalu, Borneo. J Biogeogr 33:1838–1849 Hemp A (2010) Altitudinal zonation and diversity patterns in the forests of Mount Kilimanjaro, Tanzania. In: Bruijnzeel LA, Scatena FN, Hamilton LS (eds) Tropical montane cloud forests, Science for conservation and management. Cambridge University Press, Cambridge, pp 134–141 Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. Int J Climatol 25:1965–1978 Holdridge LR, Grenke WC, Hatheway WH, Liang T, Tosi JA (1971) Forest environments in tropical life zones: a pilot study. Pergamon Press, Oxford Homeier J, Breckle SW, Günter S, Rollenbeck RT, Leuschner C (2010) Tree diversity, forest structure and productivity along altitudinal and topographical gradients in a speciesrich Ecuadorian montane rain forest. Biotropica 42:140–148 Kappelle M, Zamora N (1995) Changes in woody species richness along altitudinal gradient in Talamanca montane Quercus forest, Costa Rica. In: Churchill SP, Balslev H, Forero E, Lutein JL (eds) Biodiversity and conservation of neotropical montane forest. The New York Botanical Garden, New York, pp 135–148 Kitayama K (1992) An altitudinal transect study of the vegetation on Mount Kinabalu, Borneo. Vegetatio 102:149–171 Lieberman D, Lieberman M, Peralta R, Hartshorn GS (1996) Tropical forest structure and composition on a large-scale altitudinal gradient in Costa Rica. J Ecol 84:137–152 123 Plant Ecol (2013) 214:685–701 Lomolino MV (2001) Elevation gradients of species-density: historical and prospective views. Glob Ecol Biogeogr 10:3–13 López-Mata L, Villaseñor JL, Cruz-Cárdenas G, Ortiz E, OrtizSolorio C (2012) Predictores ambientales de la riqueza de especies de plantas del bosque húmedo de montaña de México. Bot Sci 90:27–36 Luna I, Almeida L, Llorente J (1989) Florı́stica y aspectos fitogeográficos del bosque mesófilo de montaña de las cañadas de Ocuilan, estados de Morelos y México. An Inst Biol Ser Bot 59:63–87 Martin PH, Fahey TJ, Sherman RE (2010) Vegetation zonation in a neotropical montane forest: environment, disturbance and ecotones. Biotropica 43:533–543 McCune B, Grace JB (2002) Analysis of ecological communities. MjM Software, Gleneden Beach Oommen MA, Shanker K (2005) Elevational species richness patterns emerge from multiple local mechanisms in Himalayan woody plants. Ecology 86:3039–3047 R (2007) R, A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna at http://www.R-project.org Rahbek C (2005) The role of spatial scale and the perception of large-scale species richness patterns. Ecol Lett 8:224–239 Reich RM, Bonham CD, Aguirre-Bravo C, Cházaro-Basañez M (2010) Patterns of tree species richness in Jalisco, Mexico: relation to topography, climate and forest structure. Plant Ecol 210:67–84 Rincón A (2007) Estructura y composición florı́stica de los bosques tropicales húmedos de montaña de Santa Cruz Tepetotutla, Oaxaca, México. Bachelor of Science thesis, Facultad de Ciencias, Universidad Nacional Autónoma de México, Mexico Rivera LW, Zimmerman JK, Aide TM (2000) Forest recovery in abandoned agricultural lands in a karst region of the Dominican Republic. Plant Ecol 148:115–125 Rojas-Soto O, Sosa V, Ornelas JF (2012) Forecasting cloud forest in eastern and southern Mexico: conservation insights under future climate change scenarios. Biodivers Conserv 21:2671–2690 Rzedowski J (1978) Vegetación de México. Editorial Limusa, Mexico Rzedowski J (1996) Análisis preliminar de la flora vascular de los bosques mesófilos de montaña de México. Acta Bot Mex 35:25–44 Rzedowski J, Palacios-Chávez R (1977) El bosque de Engelhardtia (Oreomunnea) mexicana en la región de La Chinantla (Oaxaca, México). Una reliquia del Cenozoico. Bol Soc Bot Mex 36:93–127 Salas-Morales SH, Meave JA (2012) Elevational patterns in the vascular flora of a highly diverse region in southern Mexico. Plant Ecol 213:1209–1220 Sánchez-González A, López-Mata L (2005) Plant species richness and diversity along an altitudinal gradient in the Sierra Nevada, Mexico. Divers Distrib 11:567–575 SAS (2005) JMP user’s guide. SAS Institute, Cary Tang CQ, Ohsawa M (1997) Zonal transition of evergreen, deciduous, and coniferous forests along the altitudinal gradient on a humid subtropical mountain, Mt. Emei, Sichuan. China Plant Ecol 133:63–78 Téllez-Valdés O, Dávila-Aranda P, Lira-Saade E (2006) The effects of climate change on the long-term conservation of Author's personal copy Plant Ecol (2013) 214:685–701 Fagus grandifolia var. mexicana, an important species of the Cloud Forest in Eastern Mexico. Biodivers Conserv 15:1095–1107 Toledo M (2013) Diversidad y estructura de la vegetación arbórea a lo largo del gradient altitudinal del Cofre de Perote, Veracruz. MSc thesis, Instituto de Ecologı́a, Universidad Nacional Autónoma de México, Mexico Vázquez JA, Givnish TJ (1998) Altitudinal gradients in tropical forest composition, structure, and diversity in the Sierra de Manantlán. J Ecol 86:999–1020 Wang ZH, Tang ZY, Fang JY (2007) Altitudinal patterns of seed plant richness in the Gaoligong Mountains, south-east Tibet, China. Divers Distrib 13:845–854 701 Williams-Linera G (2007) El bosque de niebla del centro de Veracruz. Ecologı́a, historia y destino en tiempos de fragmentación y cambio climático. CONABIO—Instituto de Ecologı́a, A.C., Mexico Williams-Linera G, Rowden A, Newton AC (2003) Distribution and stand characteristics of relict populations of Mexican beech (Fagus grandifolia var. mexicana). Biol Conserv 109:27–36 123