**3. Results**

The natural logarithm of species richness was significantly linear with (*kT*)-1 at the floristic grain. The 95% CI values of all slopes estimated by RMA did not exclude the second primary prediction of MTB, implying that no significant heterogeneity occurred in slopes among taxonomic groups (Fig2; Table2). However, at the nature reserve grain, only two of three large plant groups (angiosperm, Fig3-f and pteridophyte, Fig3-h)showed significantly linear relationship and met the first prediction of MTB, and gymnosperm were not linear and rejected the entire MTB (Fig3-g; Table3). The slope values estimated by RMA regression for all taxonomic groups were significantly exclusive from the second prediction of MTB (Table2).

100 Biodiversity Conservation and Utilization in a Diverse World

nature reserves also were showed in Table 1.

size on the species richness (See Fig 4 and 5).

area were also documented.

**3. Results** 

grain.

**2. Methods** 

Here we aimed to evaluate how the relationship between species richness and temperature predicted by MTB varied with respect to sampling scales, as well as with respect to different plant taxonomic group using an extensive plant data sets including three divisions in vascular plant at two different sample scales including nature reserve grain and floristic

We compiled species richness and other basic characteristics of 11 floristic regions and 270 natural reserves. All of the plant species richness data sets used in our analysis were collected from the previous reports involving eleven floristic regions and 270 nature reserves across the eastern China (Zhao & Fang, 2006, many others; for details see Zhang et al.2011). All species were compiled and classified into three groups (pteridophyte, gymnosperm, and angiosperm) at both floristic and reserve scales (the details see Zhang et al.2011). Here the alien species were excluded from our data analyses and only the native species retained. The areas of nature reserves and floristic regions were respectively range from 0.64 to 6689 and from 52000 to 960000 square kilometers (km2) between 18.4° N and 51.6° N latitude and between 95° E and 130.6° E longitude covering a total terrestrial area of 132,540 km2 (See Fig1 and Table1). The temperature and the size distribution of the 270

The mean annual temperature (MAT), assigned to each nature reserve based on its location and used to analyze the relationship between temperature and species richness, was compiled from a 1971-2000 temperature database of China generated from 722 climate stations across China. Flora's MAT was an average value of all the covered climate stations within each floristic region. Other environmental variables such as geographical range and

Descriptive statistics of plant species richness and environmental variables were produced to interpret the information on the data distributions (Table1). The observed slopes of Intransformed richness versus 1/*kT* relationships and at the floristic and reserve grains across the taxonomic group (three divisions and five families of the angiosperms) were estimated by reduced major axis (RMA) regression (Brown *et al*., 2004; Hawkins *et al*., 2007). The species richness-area relationships were also analyzed to evaluate the effect of the region

The natural logarithm of species richness was significantly linear with (*kT*)-1 at the floristic grain. The 95% CI values of all slopes estimated by RMA did not exclude the second primary prediction of MTB, implying that no significant heterogeneity occurred in slopes among taxonomic groups (Fig2; Table2). However, at the nature reserve grain, only two of three large plant groups (angiosperm, Fig3-f and pteridophyte, Fig3-h)showed significantly linear relationship and met the first prediction of MTB, and gymnosperm were The species-area relationships for all taxonomic divisions at both the floristic and nature reserve special scales indicated that the area size of community have more impact on the species richness for subdivision (e.g. family) than for division (Fig. 4 and 5). Moreover, the observed slope values were close to or encompass (95% CI) the theoretical values predicted by MBT at the spatial scale range of 50- 6698 km2, excluding the size of area class less than 50 km2 (Fig. 6; Table 4).

**Figure 2.** The relationship between natural logarithm of species richness (ln*S*) and inverse temperature (1/k*T*) for seven groups in 11 floristic regions: two divisions (gymnosperm and angiosperm) and five families of angiosperm (Compositae, Poaceae, Rosaceae, Liliaceae and Labiatae).


**Table 2.** Summary of regressions testing Model II (RMA) slopes of richness-temperature relationships for cases with linear relationship between inverse scaled temperature and ln-transformed richness in 11 floristic regions.

Species Distribution Patterns, Species-Area and Species-Temperature Relationships in Eastern Asian Plants 103

**Figure 4.** Species richness-area relationships for the seven plant groups in floristic regions: two divisions (gymnosperm and angiosperm) and five families of angiosperm (Compositae, Poaceae,

**Figure 5.** Species richness-area relationships for the eight plant groups in 270 nature reserves: three divisions (Pteidophyte, Gymnosperm and Angiosperm) and five families of angiosperm (Compositae,

Rosaceae, Liliaceae and Labiatae).

Poaceae, Rosaceae, Liliaceae and Labiatae).

**Figure 3.** The relationship between natural logarithm of species richness (ln*S*) and inverse temperature (1/k*T*) for eight plant groups in 270 nature reserve: three divisions (Pteidophyte, Gymnosperm and Angiosperm) and five families of angiosperm (Compositae, Poaceae, Rosaceae, Liliaceae and Labiatae).


**Table 3.** Summary of regressions testing Model II (RMA) slopes of richness-temperature relationships for cases with linear relationship between inverse rescaled temperature and ln-transformed richness in nature reserve.

**Figure 3.** The relationship between natural logarithm of species richness (ln*S*) and inverse temperature (1/k*T*) for eight plant groups in 270 nature reserve: three divisions (Pteidophyte, Gymnosperm and Angiosperm) and five families of angiosperm (Compositae, Poaceae, Rosaceae, Liliaceae and Labiatae).

Group Figure N R2 P RMA slope(95%CI) Compositae 3-a 71 0.05 0.08 0.73(0.56– 0.91) Poaceae 3-b 70 0.00 0.58 -0.71(-0.89– -0.54) Rosaceae 3-c 71 0.05 0.05 0.96(0.73– 1.18) Liliaceae 3-d 49 0.02 0.36 0.81(0.58– 1.05) Labiatae 3-e 56 0.01 0.60 -0.81(-1.03– -0.59) Angiosperm 3-f 255 0.05 0.09 -0.90(-1.01– -0.79) Gymnosperm 3-g 234 0.00 0.326 -0.91(-1.03– -0.79) Pteridophyte 3-h 189 0.21 <0.001 **-1.55(-1.75– -1.35)**  Vascular plant 3-i 193 0.09 <0.001 **-0.82(-0.93– -0.71)** 

**Table 3.** Summary of regressions testing Model II (RMA) slopes of richness-temperature relationships for cases with linear relationship between inverse rescaled temperature and ln-transformed richness in

nature reserve.

**Figure 4.** Species richness-area relationships for the seven plant groups in floristic regions: two divisions (gymnosperm and angiosperm) and five families of angiosperm (Compositae, Poaceae, Rosaceae, Liliaceae and Labiatae).

**Figure 5.** Species richness-area relationships for the eight plant groups in 270 nature reserves: three divisions (Pteidophyte, Gymnosperm and Angiosperm) and five families of angiosperm (Compositae, Poaceae, Rosaceae, Liliaceae and Labiatae).

Species Distribution Patterns, Species-Area and Species-Temperature Relationships in Eastern Asian Plants 105

eastern Asia (Wang *et al*., 2009). However, when we analyzed these data sets at the floristic regions ranging from 52000 km2 to 960000 km2, not only this linear relationship was observed, but also the slopes is highly in agreement with the theoretical values of MTB (Allen *et al*. 2002; Brown *et al*. 2004). Therefore, the plant species richness patterns predicted by MTB apparently depended on the grain size (Ellison, 2007). This scenario may be due to the fact that the number of species at the large scale overwhelmed the number of species at the relative small sample scale (e.g. nature reserve). However our analysis of species richness-area relationships showed no significant relations at floristic grain (Fig 4). The adjacent nature reserves frequently have the similar annual temperature, but the other environmental factors (i.e. water, elevation and nutrition) may exhibit a lot of variations between them that can also strongly influence the local plant species richness (Storch *et al*., 2007). The large-scale (floristic region) patterns are not simply explicable in terms of knowledge of small-scale (nature reserve) processes (Storch and Gaston, 2004). On the contrary, despite the habitat heterogeneity including annual temperature is large between plant flora, it is usually overwhelmed within plant flora because of the enormous sample

For the purpose of evaluating the MBT's robustness, Hawkins *et al*. (2007a) show the relationship between the inverse of temperature and the natural log of richness in terrestrial ectotherms (including amphibians, reptiles), invertebrates, mammals and plant around the world. However, in their plant data sets, detailed taxonomic unit (e g, pteridophyte, gymnosperm and family unit) were not contained. In their 46 data sets, 14 had no significant relationship; 9 of the remaining 32 were linear, meeting the first prediction of the MBT, but the slope values against its second prediction. So, they contended that it was important to use appropriate taxonomic ranges for accepting or refusing the prediction of MBT (see also,

Our results clearly showed that the significant taxonomic dependence in the nature reserve data sets. Pteridophyte unit which potentially supported the first prediction of MTB dominantly differs from the other groups in particular. Pteridophytes have a reproductive strategy based on the high dispersibility of spores, and have a strong moisture dependence of the sexual reproduction (Pausas & Sáez, 2000; Lehmann, *et al*., 2002; Castán & Vetaas, 2005). Thus, the life history and growth cycle for pteridophytes are probably more directly and tightly linked to abiotic factors than many other groups of plants because of the lack of co-evolved relationships with animal vectors (Barrington, 1993; Lwanga *et al*., 1998; Pausas & Sáez, 2000; Castán & Vetaas, 2005). So our results possibly support the perspective that the ability of MBT to predict richness patterns will also depend on dispersal ability (Latimer,

The plant species were not subdivided into division group to test the slopes converge around the predicted value -0.65 by MTB (Allen *et al*. 2002; Brown *et al*. 2004). Whereas the significant heterogeneity of slopes were observed at both floristic region and reserve scale among the different taxonomic groups as the most recently reported by Hawkins *et al*. (2007a,b) and Wang *et al*.(2009), indicating that the plant groups may hold variable

scale (Field *et al*., 2009).

Ellison, 2007).

2007).

**Figure 6.** Temperature–richness relationships for the vascular plant group along an area classes. Note that the classification of area is following: a)<50, b)50-100, c)100-200, d)200-400, e)400-800, f)800-1600, g)>1600.


**Table 4.** Summary of regressions testing Model II (RMA) slopes of richness-temperature relationships for vascular plant group with linear relationship between inverse rescaled temperature and lntransformed richness along an area classes.

#### **4. Discussion**

Hawkins' *et al*. (2007a) suggested that the relationship of logarithm transformed species richness and inverse temperature was nonlinear through analyzing the datasets of Chinese angiosperm taken from nature reserves with a range of area from 100 km2 to 247 km2. Here we similarly failed to observe significantly linear relationships between them at the nature reserve grain with the regions ranging from 0.64 km2 to 6689 km2, excepting for two large groups (angiosperm and pteridophyte). Moreover, almost all slope values were exclusive from the predictive range of MTB (Table 2) as the pattern of tree species distribution in

g)>1600.

transformed richness along an area classes.

**4. Discussion** 

**Figure 6.** Temperature–richness relationships for the vascular plant group along an area classes. Note that the classification of area is following: a)<50, b)50-100, c)100-200, d)200-400, e)400-800, f)800-1600,

Area classes(km2) Figure N R2 P RMA slope(95%CI) <50 7-a 32 0.08 0.12 1.04(0.67-1.42) 50-100 7-b 25 0.04 0.32 -0.72(-1.02– -0.41) 100-200 7-c 44 0.25 <0.001 **-0.63(-0.80– -0.46)**  200-400 7-d 33 0.06 0.19 -1.24(-1.69– -0.80) 400-800 7-e 22 0.36 0.003 **-0.60(-0.82– -0.37)**  800-1600 7-f 12 0.25 0.08 -0.96(-1.55– -0.38) >1600 7-g 16 0.18 0.10 -0.84(-1.27– -0.40)

**Table 4.** Summary of regressions testing Model II (RMA) slopes of richness-temperature relationships for vascular plant group with linear relationship between inverse rescaled temperature and ln-

Hawkins' *et al*. (2007a) suggested that the relationship of logarithm transformed species richness and inverse temperature was nonlinear through analyzing the datasets of Chinese angiosperm taken from nature reserves with a range of area from 100 km2 to 247 km2. Here we similarly failed to observe significantly linear relationships between them at the nature reserve grain with the regions ranging from 0.64 km2 to 6689 km2, excepting for two large groups (angiosperm and pteridophyte). Moreover, almost all slope values were exclusive from the predictive range of MTB (Table 2) as the pattern of tree species distribution in eastern Asia (Wang *et al*., 2009). However, when we analyzed these data sets at the floristic regions ranging from 52000 km2 to 960000 km2, not only this linear relationship was observed, but also the slopes is highly in agreement with the theoretical values of MTB (Allen *et al*. 2002; Brown *et al*. 2004). Therefore, the plant species richness patterns predicted by MTB apparently depended on the grain size (Ellison, 2007). This scenario may be due to the fact that the number of species at the large scale overwhelmed the number of species at the relative small sample scale (e.g. nature reserve). However our analysis of species richness-area relationships showed no significant relations at floristic grain (Fig 4). The adjacent nature reserves frequently have the similar annual temperature, but the other environmental factors (i.e. water, elevation and nutrition) may exhibit a lot of variations between them that can also strongly influence the local plant species richness (Storch *et al*., 2007). The large-scale (floristic region) patterns are not simply explicable in terms of knowledge of small-scale (nature reserve) processes (Storch and Gaston, 2004). On the contrary, despite the habitat heterogeneity including annual temperature is large between plant flora, it is usually overwhelmed within plant flora because of the enormous sample scale (Field *et al*., 2009).

For the purpose of evaluating the MBT's robustness, Hawkins *et al*. (2007a) show the relationship between the inverse of temperature and the natural log of richness in terrestrial ectotherms (including amphibians, reptiles), invertebrates, mammals and plant around the world. However, in their plant data sets, detailed taxonomic unit (e g, pteridophyte, gymnosperm and family unit) were not contained. In their 46 data sets, 14 had no significant relationship; 9 of the remaining 32 were linear, meeting the first prediction of the MBT, but the slope values against its second prediction. So, they contended that it was important to use appropriate taxonomic ranges for accepting or refusing the prediction of MBT (see also, Ellison, 2007).

Our results clearly showed that the significant taxonomic dependence in the nature reserve data sets. Pteridophyte unit which potentially supported the first prediction of MTB dominantly differs from the other groups in particular. Pteridophytes have a reproductive strategy based on the high dispersibility of spores, and have a strong moisture dependence of the sexual reproduction (Pausas & Sáez, 2000; Lehmann, *et al*., 2002; Castán & Vetaas, 2005). Thus, the life history and growth cycle for pteridophytes are probably more directly and tightly linked to abiotic factors than many other groups of plants because of the lack of co-evolved relationships with animal vectors (Barrington, 1993; Lwanga *et al*., 1998; Pausas & Sáez, 2000; Castán & Vetaas, 2005). So our results possibly support the perspective that the ability of MBT to predict richness patterns will also depend on dispersal ability (Latimer, 2007).

The plant species were not subdivided into division group to test the slopes converge around the predicted value -0.65 by MTB (Allen *et al*. 2002; Brown *et al*. 2004). Whereas the significant heterogeneity of slopes were observed at both floristic region and reserve scale among the different taxonomic groups as the most recently reported by Hawkins *et al*. (2007a,b) and Wang *et al*.(2009), indicating that the plant groups may hold variable activation energies rather than an invariant value. Our more recently research showed that validity of the MTB lies on if the area size of the community has no significant effect on species richness (Zhang et al. 2011). Therefore we believe that the slope value for each taxonomic group should be co-influenced by the restriction of distribution range, the area size of sampling community and other abiotic factors, as well as the inherent activation energy differences.

Species Distribution Patterns, Species-Area and Species-Temperature Relationships in Eastern Asian Plants 107

Allen, A. P., Gillooly, J. F., Savage, V. M., Brown, J. H. 2006 Kinetic effects of temperature on rates of genetic divergence and speciation. *Proc. Natl Acad. Sci. USA* 103, 9130–9135. Brown, J. H., Gillooly, J. F., Allen, A. P., Savage, V. M., West, G. B. 2004 Toward a metabolic

Cardinale, B. J., Hillebrand, H., Harpole, W. S., Gross, K., Ptacnik, R. (2009) Separating the influence of resource 'availability' from resource 'imbalance' on productivity–diversity

Cassemiro, F. A. S., Barreto, B. S., Rangel, T. F. L. V. B., Diniz-Filho, J. A. F. 2007 Nonstationarity, diversity gradients and the metabolic theory of ecology. *Global Ecol.* 

Colwell, R. K., Lees, D. C. (2000) The mid-domain effect: geometric constraints on the

Deng, J.M., Li, T., Wang, G.X., Liu, J., Zhao, C.M., Ji, M.F., Zhang, Q., Liu, J.Q. 2008 Tradeoffs between the metabolic rate and population density of plants. *Plos One*, 3 (3), 1799. Deng, J. M., Wang, G. X., Morris, E. C., Wei, X. P., Li, D. X., Chen, B. M., Zhao, C. M., Liu, J., Wang, Y. 2006 Plant mass–density relationship along a moisture gradient in north-west

Ellison, A. M. (2007) Metabolic theory and patterns of species richness. *Ecology*, 88, 1889. Enquist, B. J., Brown, J. H., West, G. B. 1998 Allometric scaling of plant energetics and

Evans, K. L., Gaston, K. J. (2005) Can the evolutionary–rates hypothesis explain species–

Evans, K. L., Warren, P. H., Gaston, K. J. (2005) Species–energy relationships at the macroecological scale: a review of the mechanisms. *Biological Reviews*, 80, 1–25. Field, R., Hawkins, A. B., Cornell, H. V., Currie, D. J., Diniz-Filho, J. A. F., Guégan, J. F., Kaufman, D. M., Kerr, J. T., Mittelbach, G. C., Oberdorff, T., O'Brien, E. M. and Turner, J. R. G. (2009) Spatial species-richness gradients across scales: a meta-analysis. *Journal of* 

Gillooly, J. F. & Allen, A. P. 2007 Linking global patterns in biodiversity to evolutionary

Hawkins, B. A., Albuquerque, F. S., Araújo, M. B., Beck, J., Bini, L. M., Cabrero-Sańudo, F. J., Castro-Parga, I., Diniz-Filho, J. A. F., Ferrer-Castán, D., Field, R., *et al*. 2007a A global evaluation of metabolic theory as an explanation for terrestrial species richness

Hawkins, B.A., Diniz-Filho, J. A. F., Bini, L. M., Araújo, M. B., Field, R., Horta,l J., Kerr, J. T., Rahbek, C., Rodríguez, M. Á., Sanders, N. J. (2007b) Metabolic theory and diversity

Hawkins, B. A., Field, R., Cornell, H. V., Currie, D. J., Guegan, J. F., Kaufman, D. M., Kerr, J. T., Mittelbach, G. G., Oberdorff, T., O'Brien, E. M., Porter, E. E., Turner, J. R. G. (2003) Energy, water, and broad-scale geographic patterns of species richness. *Ecology*, 84,

Hunt, G., Cronin, T. M., Roy, K. 2005 Species–energy relationship in the deep see: a test

geography of species richness. *Trends in Ecology and Evolution*, 15, 70–76.

theory of ecology. *Ecology* 85, 1771–1789.

relationships. *Ecology letters*, 12, 475–487.

China. *Journal of Ecology* 94, 953–958.

population density. *Nature* 395,163–165.

*Biogeography*, 36, 132–147.

gradients. *Ecology* 88, 1877–1888.

3105–3117.

energy relationships? *Functional Ecology*, 19, 899–915.

Gaston, K. J. (2000) Global patterns in biodiversity. *Nature*, 405, 220–227.

gradients: where do we go from here? *Ecology*, 88, 1898–1902.

using the Quaternary fossil record. *Ecol Lett* 8, 739–747.

dynamics using metabolic theory. *Ecology* 88, 1890–1894.

*Biogeogr* 16, 820–822.
