Biodivers Conserv DOI 10.1007/s10531-014-0676-8 ORIGINAL PAPER
Environmental determinants of geographic butterfly richness pattern in eastern China Shengbin Chen • Lingfeng Mao • Jinlong Zhang • Kexin Zhou Jixi Gao
•
Received: 13 June 2013 / Revised: 27 February 2014 / Accepted: 7 March 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract A long-standing task for ecologists and biogeographers is to reveal the underlying mechanisms accounting for the geographic pattern of species diversity. The number of hypotheses to explain geographic variation in species diversity has increased dramatically during the past half century. The oldest and the most popular one is environmental determination. However, seasonality, the intra-annual variability in climate variables has been rarely related to species richness. In this study, we assessed the relative importance of three environmental hypotheses: energy, seasonality and heterogeneity in explaining species richness pattern of butterflies in Eastern China. In addition, we also examined how environmental variables affect the relationship between species richness of butterflies and seed plants at geographic scale. All the environmental factors significantly affected butterfly richness, except sampling area and coefficient of variation of mean monthly precipitation. Energy and seasonality hypotheses explained comparable variation in butterfly richness (42.3 vs. 39.3 %), higher than that of heterogeneity hypothesis (25.9 %). Variation partitioning indicated that the independent effect of seasonality was
Communicated by Peter J. T. White. Electronic supplementary material The online version of this article (doi:10.1007/s10531-014-0676-8) contains supplementary material, which is available to authorized users. S. Chen K. Zhou J. Gao (&) Nanjing Institute of Environmental Sciences, Ministry of Environmental Protection, No. 8 Jiangwangmiao Street, Xuanwu District, Nanjing 210042, China e-mail:
[email protected] S. Chen e-mail:
[email protected] L. Mao Department of Plant Biology, Michigan State University, East Lansing, MI 48824, USA J. Zhang Flora Conservation Department, Kadoorie Farm and Botanic Garden, Lam Kam Road, Tai Po, NT, Hong Kong SAR
123
Biodivers Conserv
much lower (0.0 %) than that of energy (5.5 %) and heterogeneity (6.3 %). However, seasonality performed better in explaining butterfly richness in topographically complex areas, reducing spatial autocorrelation in butterfly richness, and more strongly affect the association between butterflies and seed plants. The positive relationship between seed plant richness and butterfly richness was most likely the result of environmental variables (especially seasonality) influencing them in parallel. Insufficient sampling may partly explain the low explanatory power of environmental model (52.1 %) for geographic butterfly richness pattern. Our results have important implications for predicting the response of butterfly diversity to climate change. Keywords Biogeography Butterfly fauna Energy Heterogeneity Seasonality Species richness
Introduction Understanding the underlying mechanisms shaping the geographic pattern of species diversity has been a long-standing task for ecologists and biogeographers. This is increasingly critical not only for biodiversity conservation but also for predicting the response of species diversity to global change (Kerr et al. 2007). The number of hypotheses to explain geographic variation in species diversity has increased dramatically during the past half century (Willig et al. 2003; Field et al. 2009). While different hypotheses refer to ecological, biogeographic, or evolutionary processes, they all predict a prominent role of environmental variables (energy, seasonality and heterogeneity) in explaining geographic species richness patterns (Willig et al. 2003; Field et al. 2009; Tello and Stevens 2010). In addition, plant diversity is often regarded as the direct driver of animal richness, especially for herbivores and frugivores (Kissling et al. 2007; Kissling et al. 2010; Zhang et al. 2013). The roles of energy and heterogeneity in determining geographic species richness patterns have been extensively examined (Evans et al. 2005; Field et al. 2009). Energy can be grouped into two categories, kinetic and potential energy, which potentially control species number in different ways (Clarke and Gaston 2006; Allen et al. 2007). Kinetic energy, which directly reflects fluxes of solar radiation, is often represented by measures of temperature and potential evapotranspiration (Evans et al. 2005; Clarke and Gaston 2006; Allen et al. 2007), while potential energy directly or indirectly relates to the rate of conversion of solar radiation to reduced carbon compounds through photosynthesis, and can be reflected by actual evapotranspiration (Allen et al. 2007). Many studies have found strong positive relationships between species richness of various taxa and energy-related variables (Hawkins et al. 2003; Field et al. 2009). The heterogeneity hypothesis assumes that habitats with complex spatial structure may provide more niches and more diverse ways of exploiting resources, or that they increase speciation rates and thus species richness (Ruggiero and Hawkins 2008). Heterogeneity often represents spatial variation of energy availability at local scales (Tello and Stevens 2010). Seasonality, the intra-annual variability in climate variables (Williams and Middleton 2008; Ackerly et al. 2010; Carrara and Va´zquez 2010), has been rarely related to species richness and the results were controversial (Kerr 1999; Badgley and Fox 2000; Ruggiero and Kitzberger 2004; Qian 2008; Carrara and Va´zquez 2010; Tello et al. 2010).
123
Biodivers Conserv
Seasonality may have two effects on biodiversity, depending on the characteristics of specific organisms and on the scale at which ecological and evolutionary processes happen. On the one hand, seasonality could increase species diversity if species specialize and occupy different temporal niches (Chesson and Huntly 1997). On the other hand, seasonality could lead to low species diversity because species must evolve broad niches to survive in highly seasonal habitats, which may make the community more readily saturated (Tello and Stevens 2010). Our poor understanding on how seasonality controls biodiversity present a major challenge for predicting the future of biodiversity in a changing climate. Therefore, more researches are required to elucidate the effect of seasonality on species diversity of various taxa. Butterflies, one of the best-known invertebrate taxa worldwide, are an important component of biodiversity and ecosystem structure. Climatic conditions strongly constrain the metabolism and behavior of butterflies; therefore, they are particularly suitable for examining species richness-environment relationships (White and Kerr 2007) and predicting potential effect of future climatic change on species distributions (Fleishman 2010). Several studies have found weak direct relationship between butterfly richness and plant richness at different geographical scales (Hawkins and Porter 2003a, b; Mene´ndez et al. 2007; Kitahara et al. 2008). However, these analyses were generally conducted at medium geographic extent and the results were conflicting. So far, almost all geographic scale studies on butterfly diversity have been carried out in Europe or North America. A geographic broadening of investigated regions may benefit the aim of finding general patterns. The butterfly fauna of China, with 12 families, 375 genera and [1300 species, is more diverse than that of Europe or North America (Chou 2000). Most species occur in Eastern China, a topographically complex region that is affected by monsoon climate (Chou 2000). The local butterfly fauna in many areas (such as nature reserves, national parks and counties) of east China has been intensively inventoried during the past decades. In the present study, we used a large dataset on butterfly species richness to test the following two hypotheses: (1) seasonality plays an important role in explaining geographic variation in butterfly richness, together with energy and heterogeneity, and (2) seed plant richness contributes substantially to the geographic pattern of butterfly richness, besides environmental variables.
Materials and methods Species richness data We compiled 58 local butterfly surveys from journal papers and books that contained all of the following information: survey area, topography, location, seed plant richness and butterfly sampling effort (Online Appendix 1). Twenty-seven localities had data on number of specimens, 53 localities had data on number of survey years, and 22 localities had both of them. There was a significant correlation between these two measures of sampling efforts (r = 0.696, p \ 0.001 for ln-transformed variables). Although all localities have been extensively surveyed with the aim of providing complete species lists for butterflies, there was great variation in sampling effort (i.e., 300–5600 specimens, 1–12 years for survey).
123
Biodivers Conserv
Environmental data We used mean annual temperature (TEM), annual precipitation (PRE), annual actual evapotranpiration (AET), and annual potential evapotranspiration (PET) as energy variables. Standard deviation of mean monthly temperature (TEMsd), coefficient of variation of mean monthly precipitation (PREcv), coefficient of variation of mean monthly actual evapotranpiration (AETcv), and coefficient of variation of mean monthly potential evapotranspiration (PETcv) were regarded as variables reflecting seasonality. Coefficient of variation can be misleading for variables including both positive and negative values and meaningful for variables containing only positive values. Since some localities have monthly temperature \0 °C, we used standard deviation for temperature and coefficient of variation for the rest variables. Elevational range (ELER, maximum elevation—minimum elevation) represents heterogeneity because climate usually changes systematically along altitudinal gradients. We also regarded sampling area (AREA) to reflect heterogeneity because larger area may cover more diverse climatic conditions and habitats. The average area of each locality was 426 km2 (±713 SE). Elevational range and sampling area were significantly correlated (r = 0.422, p = 0.003 for ln-transformed variables). Data for temperature and precipitation were obtained from WorldClim (Hijmans et al. 2005). Data for the evapotranspiration variables stemmed from the Global Evapotranspiration and Water Balance Data Sets developed by Ahn and Tateishi (1994) and Tateishi and Ahn (1996). Climatic data were assembled for the 58 localities according to their geographical midpoints. Statistical analysis First, we tested the normality of all variables, and ln-transformed butterfly richness (BR), seed plant richness (SPR), elevational range, sampling area, number of specimens, and number of survey years to obtain normal distributions. We used multiple ordinary least squares regression with linear and quadratic terms to examine the effects of each putative predictor and each set of variables (energy, seasonality, and heterogeneity) on butterfly richness. We used adjusted R2 to estimate explained variance. We examined the statistical significance of the correlations and regressions based on corrected degrees of freedom, calculated using Dutilleul’s (1993) modified t test. There were strong correlations between putative predictors (Table S1). To resolve this problem before modeling analysis, we eliminated the putative predictor with the largest variance inflation factor each time until all the selected predictors’ variance inflation factor \5 (Quinn and Keough 2002). We repeated this procedure separately for the three sets of variables, i.e. energy variables (TEM, PRE, AET and PET), seasonality variables (TEMsd, PREcv, AETcv and PETcv), and heterogeneity variables (ELER and AREA). For each set of variables, we used Akaike’s information criterion (AIC) to rank all possible models and selected the model with the minimum AIC as the best one. We used variation partitioning approach (Heikkinen et al. 2005) to decompose the variation in butterfly richness among the three sets of predictors (energy, seasonality and heterogeneity). Here, variation partitioning lead to eight fractions: pure effect of energy, seasonality, and heterogeneity; combined variation due to joint effects of energy and seasonality, energy and heterogeneity, seasonality and heterogeneity, combined effects of the three groups of explanatory variables, and finally unexplained variation. We used ‘‘permutation of the residuals’’ approach to test the significance of the independent effect of energy, seasonality and heterogeneity (Legendre and Legendre 1998).
123
Biodivers Conserv
To examine whether seed plant richness provide additional explanation for butterfly richness, besides environmental variables, we repeated the above regression analyses for seed plant richness, using energy, seasonality and heterogeneity variables as putative predictors. Then, we performed residual regression to examine the relationship between butterfly richness and seed plant richness when the effects of environmental factors were gradually removed. Spatial autocorrelation in macroecological species richness data may result in inflation of type I error and bias the estimation of coefficients in statistical modeling (Diniz-Filho et al. 2003; Bini et al. 2009). We used Moran’s I of butterfly richness and residuals of ordinary least squares regression models at different distances to test whether spatial autocorrelation was present. The significance of Moran’s I was determined by permutation approach. Moran’s I coefficient indicated that residuals of energy and heterogeneity models had significant spatial autocorrelation at least at the shortest distance (Table S2). Therefore, we used simultaneous autoregressive models to repeat multiple regression analyses, and compared the results from both non-spatial ordinary least squares regression and simultaneous autoregressive models (Table S3). To determine whether sampling effort affect local butterfly richness, we regressed the residual butterfly richness of environmental model against ln-transformed number of specimens and survey years. The data analyses were performed in R (R Development Core Team 2009) and ‘Spatial Analysis in Macroecology’ (SAM) (Rangel et al. 2010).
Results Butterfly richness in eastern China showed evident latitudinal gradient (Fig. S1), averaged 141 and varied from 37 to 376 (Fig. 1). All environmental factors, except sampling area and coefficient of variation of mean monthly precipitation, significantly affected butterfly richness in single-predictor models, but their effects differed in some degree. Butterfly richness generally increased with energy and heterogeneity variables, while decreased with seasonality variables (Table 1). Multiple ordinary least squares regression analyses showed that, energy model (PET ? PET2) explained 42.3 % of the variation in butterfly richness, slightly higher than that of seasonality model (TEMsd) (39.3 %), and much higher than that of heterogeneity model (ELER) (25.9 %). The environmental model combining all the selected variables explained 52.1 % of the variation in butterfly richness (Table 2). Simultaneous autoregressive models yielded similar results to multiple ordinary least squares regression models; therefore, the effects of spatial autocorrelation on the results of our analyses are negligible (Table 2, Table S3). Variance partitioning analysis showed that energy hypothesis had significant independent effect (5.5 %), much higher than that of seasonality (0.0 %). Heterogeneity had the highest independent effect (6.3 %) among these three sets of variables. There were significant positive relationships between elevational range and residual butterfly richness of energy and seasonality models (Fig. 3). Correlograms of Moran’s I of butterfly richness, and residuals from energy, seasonality and heterogeneity models indicated that energy and seasonality removed most positive spatial autocorrelation at short distances and negative spatial autocorrelation at long distances, while heterogeneity reduced positive spatial autocorrelation at intermediate distances but left considerable spatial autocorrelation at short and long distances (Fig. 4a; Table S2). Spatial autocorrelation in residuals of environmental model at each distance was
123
Biodivers Conserv
Fig. 1 Spatial locations of 58 butterfly faunas analyzed in this study
not significant (p [ 0.05; Fig. 4b; Table S2), suggesting that spatial autocorrelation in model residuals was negligible. There was a highly significant correlation between butterfly richness and seed plant richness (Fig. 5a). However, the correlation between residuals of these two variables decreased at different degree when the influences of environment (energy, seasonality and heterogeneity) were removed (Fig. 5b, c, d). When all environmental effects were partitioned out, there was no significant correlation between the residuals of species richness of
123
Biodivers Conserv Table 1 Regression coefficients from single predictor ordinary least squares regression Variable
Linear
Quadratic
TEM
(?)0.285*
0.321*
PET
(?)0.333*
0.423*
PRE
(?)0.191*
0.223*
AET
(?)0.321*
0.346*
TEMsd
(-)0.393*
0.404*
PETcv
(-)0.346*
0.368*
PREcv
(-)0.145
0.146
AETcv
(-)0.211*
0.227*
AREA
(?)0.032
0.024
ELER
(?)0.259**
0.265**
Energy variables
Seasonality variables
Heterogeneity variables
AREA, ELER and Species richness was ln-transformed. p-values were calculated after accounting for spatial autocorrelation using Dutilleul (1993)’s method TEM mean annual temperature, PRE annual precipitation, AET annual actual evapotranpiration, PET annual potential evapotranspiration, TEMsd standard deviation of mean monthly temperature, PREcv coefficient of variation of mean monthly precipitation, AETcv coefficient of variation of mean monthly actual evapotranpiration, PETcv coefficient of variation of mean monthly potential evapotranspiration, ELER elevational range, AREA sampling area ** p \ 0.01; * p \ 0.05
butterflies and seed plants (Fig. 5e). There were significant linear correlations between lntransformed number of specimens or survey years and residual butterfly richness left by environmental model (Fig. 6).
Discussion Using inventory data on 58 local butterfly fauna in Eastern China, we demonstrated that there are strong relationships between butterfly richness and environmental factors (energy, seasonality, and heterogeneity) and seed plant richness. By using variation partitioning method to explicitly determine the unique and joint effects of each set of environmental variables, and using residual regression to examine the association between butterfly richness and seed plant richness, we found that (1) the variation explained by seasonality in butterfly richness was comparable to that of energy, much higher than that of heterogeneity and (2) butterfly richness showed significant correlation with seed plant richness, due to their similar responses to environmental variables. Seasonality Our analyses showed that the explanatory power of seasonality for butterfly richness was just slightly lower than energy, but much higher than heterogeneity (Table 2). This partly
123
123 0.259 0.521
TEMsd (-0.636)
ELER (0.499)
PET (1.725), PET2 (-1.278), TEMsd (-0.079), ELER (0.441)
Seasonality
Heterogeneity
Environment
Df. degree of freedom corrected by Dutilleul (1993)’s method
Abbreviations of environmental variables are the same as in Table 1
The standardized regression coefficients are written in brackets
0.423
PET (2.946), PET2 (-2.360)
Energy 0.393
R2adj
Variables in model
Predictor set
15.9
18.6
37.9
20.3
F
12.4
23.5
10.1
14.5
Df.
59.9
81.7
\0.001 \0.001
68.9 68.4
\0.001
AIC
\0.001
P
Table 2 Summary of the multiple ordinary least squares regression models evaluating the effect of energy, seasonality and heterogeneity on butterfly richness
Biodivers Conserv
Biodivers Conserv Fig. 2 Results of variation partitioning for butterfly richness in terms of the proportion of variation explained (%). Variation of species richness is explained by three sets of variables: seasonality, energy, and heterogeneity, and the unexplained variation; a, b, and, c are unique effects of seasonality, energy and heterogeneity, respectively; while d, e, f and g indicates their joint effects
Seasonality a = 0.0
d = 2.5
e = 19.7 g = 17.1 Energy b = 5.5
Heterogeneity f = 0.0
c = 6.3
undetermined variation 47.9
indicated that seasonality played an important role in explaining geographic variation in butterfly richness, together with energy and heterogeneity. Studies that considered seasonality as a predictor of species richness yielded confounding results. Some studies found seasonality to be good predictors of species richness (e.g. mammals: Andrews and O’Brien 2000; Badgley and Fox 2000; Tello and Stevens 2010; Carrara and Va´zquez 2010; birds: Qian 2008; Williams and Middleton 2008; Qian et al. 2009; Carrara and Va´zquez 2010), while others found species richness was not strongly influenced by seasonality (e.g. insects: Kerr 1999; Schuldt and Assmann 2009; herptiles: Qian et al. 2007). These controversial results could potentially be explained by the discrepancy in their dependency on external energy supply and ability to maintain body temperature (i.e., endotherms vs. ectotherms). Since most terrestrial ectotherms depend heavily on external energy input to maintain their metabolism, the bottleneck for their survival and persistence may be availability, other than temporal variation of energy. However, for endotherms, which can maintain thermal homeostasis, the bottleneck may convert from energy to seasonality. Consistent with this, the independent effect of seasonality is null for butterfly richness in our study (Fig. 2), and seasonality variables were not selected in the best fitting environmental model for amphibians and reptiles in China (Qian et al. 2007) and for carabid beetles in the western Palaearctic area (Schuldt and Assmann 2009). However, we must notice that energy and seasonality variables were calculated from the same basic data in our study. Therefore, it was not surprising that energy and seasonality would overlap in their capacity to explain butterfly richness (Fig. 2). There were also other evidences highlighting the importance of seasonality for explaining geographic butterfly richness pattern. First, elevatioanl range showed higher positive correlation with the residual butterfly richness of energy model than that of seasonality model (Fig. 3). This indicated that although both sets of variables have comparable explanatory powers, seasonality performed better in explaining spatial variation in butterfly richness in topographically complex areas than energy did. Second, Moran’s I correlograms showed that seasonality did better in reducing spatial autocorrelation in residuals of butterfly richness at the shortest distance than energy and heterogeneity did (Fig. 4a; Table S2). Third, when the influence of seasonality was removed, the correlation between residual richness of butterfly and seed plant was weaker than that of energy and heterogeneity models (Fig. 5). This indicated that the role of seasonality in shaping the
123
Biodivers Conserv 1.5
a
1.0
Residual-BR
Fig. 3 The relationships between ln-transformed elevational range and residual butterfly richness (BR) of energy model (a), and seasonality model (b)
r 2 = 0.228 p < 0.001
0.5 0.0 -0.5 -1.0 -1.5
b
Residual-BR
1.0
r 2 = 0.075 p = 0.022
0.5
0.0
-0.5
-1.0 5
6
7
8
9
Ln-transformed elevarional range
association between butterfly richness and seed plant richness was more important than that of energy and heterogeneity. Seed plant richness There was significant correlation between butterfly species richness and seed plant richness when the influence of environmental variables was included. However, when we removed the impacts of environmental variables by fitting all environmental variables, the residual richness of butterfly and seed plants showed no significant correlation (Fig. 5). This indicated that there was no direct association between butterfly richness and seed plant richness, and the correlation between them simply stemmed from their similar responses to environmental variables (Hawkins and Porter 2003b). The may be two reasons accounting for the indirect association between butterfly richness and seed plant richness. First, the geographic species richness pattern is primarily determined by few common species (Lennon et al. 2004), and for butterflies, the distribution of habitat generalists was more likely influenced by climate variables than that of specialists (Mene´ndez et al. 2007). Second, only a small fraction of seed plants can be used by butterflies as nectars for adults and hostplants for larvae (Chou 2000). That is, richness of total plant richness was a rough agency of resource availability. Therefore, for butterflies, it is more likely to find direct association between the diversity of specialists and that of nectars or host plants (Kitahara et al. 2008).
123
Biodivers Conserv 1.0
a
Moran's I
0.5
0.0
Butterfly richness Residuals of energy model Residuals of seasonality model Residuals of heterogeneity model
-0.5
1.0
b
Moran's I
0.5
Butterfly richness Residuals of environment model
0.0
-0.5
-1.0 500
1000
1500
2000
2500
Distance (km) Fig. 4 Moran’s I correlograms of butterfly richness and residuals. a butterfly richness and residuals of energy model, seasonality model, and heterogeneity model; b butterfly richness and residual of environmental model
Energy and heterogeneity Multiple ordinary least squares regression and variation partitioning showed that, energy was the most important factor explaining the variance in butterfly richness, in accordance with previous analyses (Hawkins and Porter 2003a, b; Hawkins et al. 2003; White and Kerr 2007). Energy-related variables could influence butterfly richness through several physiological, behavioral and evolutionary processes. Higher temperature not only allows butterflies to invest more time in foraging (Turner et al. 1987; Boggs and Murphy 1997), but also influences egg hatching, growth rate, development and survival (Gotthardt et al. 2000; Bale et al. 2002) which may result in more generations during one growth season. Although the explanatory power of heterogeneity model was lower than those of energy and seasonality, it had the largest unique component (6.3 %) in variation partitioning (Fig. 2). This indicated that heterogeneity was a complimentary factor of climatic variables. This has also been shown by previous studies in which heterogeneity variables were frequently selected in best fitting environmental models for butterfly richness (Kerr et al. 2001; Hawkins and Porter 2003a, b; White and Kerr 2007). The highest pure effect of
123
Biodivers Conserv 6.5
a
6.0
BR
5.5 5.0 4.5 4.0
r2 = 0.378 p < 0.001
3.5 3.0 6.0
6.5
7.0
7.5
8.0
8.5
SPR 1.0
1.5
0.5
c
b
0.0 -0.5 2
r = 0.176 p < 0.001
-1.0 -1.5 -1.0
-0.5
0.0
0.5
Residual-BR
Residual-BR
1.0
b 0.5 0.0
r2 = 0.115 p < 0.005
-0.5 -1.0 -1.0
1.0
-0.5
1.5
0.5 0.0 2
r = 0.181 p < 0.001
-1.0 -1.5 -1.0
-0.5
0.0
Residual-SPR
0.5
1.0
Residual-BR
Residual-BR
d
-0.5
0.5
1.0
Residual-SPR
Residual-SPR 1.5 1.0
0.0
1.0
r2 = 0.011 p = 0.186
e
0.5 0.0 -0.5 -1.0 -1.0
-0.5
0.0
0.5
1.0
Residual-SPR
Fig. 5 Relationships between butterfly species richness (BR) and seed plant richness (SPR). Influence of environmental variables was included or removed by fitting environmental variables prior to drawing the scatter plots. a Influence of environmental variables included; b Influence of energy removed; c Influence of seasonality removed; d Influence of heterogeneity removed; e Influence of energy, seasonality, and heterogeneity removed
heterogeneity in shaping the geographic butterfly richness pattern may be due to strong niche separation along altitudinal gradients in local butterfly faunas, irrespective of climatic conditions (Chou 2000). Sampling effort The insufficient sampling may partly account for the low fraction of variance explained by environmental variables in our analysis (but see Kerr et al. 2001; Hawkins and Porter 2003a, b). The number of specimen is the best measure of sampling effort (Willott 2001). The number of survey years is a rough estimation of sampling effort. The significant linear correlation between number of specimen or survey years and residuals left by
123
Biodivers Conserv 0.8
a
0.6 0.4
Residual-BR
Fig. 6 Relationships between residual butterfly species richness (BR) of environmental model and a ln-transformed number of specimen; b ln-transformed number of survey years
0.2 0.0 -0.2 -0.4
r 2 = 0.387 p < 0.001
-0.6 -0.8 5
6
7
8
9
10
Ln-transformed number of specimen 0.8
b 0.6
Residual-BR
0.4 0.2 0.0 -0.2 -0.4 2
r = 0.273 p < 0.005
-0.6 -0.8 0
1
2
3
Ln-transformed number of survey years
environmental model (Fig. 6a, b) indicated that, some local butterfly fauna are underestimated (Chou 2000). However, these factors that reduced the explanatory power of environmental model may not bias the analytical results, because spatial richness pattern is mainly determined by common species (Lennon et al. 2004), which are seldom overlooked during field surveys (Gaston et al. 1995). Acknowledgments We thank Jan Beck of Universita¨t Basel, and Hong Qian of Illinois State Museum for comments on previous versions of the manuscript. We thank three anonymous reviewers for helpful suggestions. S. Chen thanks his new son, Jiayou Chen, for his encouragement. Financial support from the National Key Technology R&D Program (2012BAC01B08) and the Special Public Science and Technology Research Program for Environmental Protection (201209027) was also acknowledged.
References Ackerly DD, Loarie DR, Cornwell WK, Weiss SB, Hamilton H, Branciforte R, Kraft NJB (2010) The geography of climate change: implications for conservation biogeography. Divers Distrib 16:476–487
123
Biodivers Conserv Ahn CH, Tateishi R (1994) Development of a global 30-minute grid potential evapotranspiration data set. ISPRS J Photogramm Remote Sens 33:12–21 Allen AP, Gillooly JF, Brown JH (2007) Recasting the species-energy hypothesis: the different roles of kinetic and potential energy in regulating biodiversity. In: Storch D, Marquet PA, Brown JH (eds) Scaling Biodiversity. Cambridge University Press, Cambridge Andrews P, O’Brien E (2000) Climate, vegetation, and predictable gradients in mammal species richness in southern Africa. J Zool 251:205–231 Badgley C, Fox DL (2000) Ecological biogeography of North American mammals: species density and ecological structure in relation to environmental gradients. J Biogeogr 27:1437–1467 Bale JS, Masters GJ, Hodkinson ID, Awmack C, Bezemer TM, Brown VK, Butterfield J, Buse A, Coulson JC, Farrar J, Good JEG, Harrington R, Hartley S, Jones TH, Lindroth RL, Press MC, Symrnioudis I, Watt AD, Whittaker JB (2002) Herbivory in global climate change research: direct effects of rising temperature on insect herbivores. Glob Change Biol 8:1–16 Bini LM, Diniz-Filho JAF, Rangel TFLVB, Akre TSB, Albaladejo RG, Albuquerque FS, Aparicio A, Arau´jo MB, Baselga A, Beck J, Bellocq MI, Bo¨hning-Gaese K, Borges PAV, Castro-Parga I, Chey VK, Chown SL, De Marco P Jr, Dobkin DS, Ferrer-Casta´n D, Field R, Filloy J, Fleishman E, Go´mez JF, Hortal J, Iverson JB, Kerr JT, Kissling WD, Kitching IJ, Leo´n-Corte´s JL, Lobo JM, Montoya D, ´ , Pausas JG, Qian H, Rahbek C, Morales-Castilla I, Moreno JC, Oberdorff T, Olalla-Ta0 rraga MA ´ , Rueda M, Ruggiero A, Sackmann P, Sanders NJ, Terribile LC, Vetaas OR, Hawkins Rodrı´guez MA BA (2009) Coefficient shifts in geographical ecology: an empirical evaluation of spatial and nonspatial regression. Ecography 32:193–204 Boggs CL, Murphy DD (1997) Community composition in mountain ecosystems: climatic determinants of montane butterfly distributions. Glob Ecol Biogeogr Lett 6:39–48 Carrara R, Va´zquez DP (2010) The species-energy theory: a role for energy variability. Ecography 33:942–948 Chesson P, Huntly N (1997) The roles of harsh and fluctuating conditions in the dynamics of ecological communities. Am Nat 150:519–553 Chou I (2000) Monographia Rhopalocerorum Sinensium, 2nd edn. Henan Science and Technology Press, Zhengzhou Clarke A, Gaston KJ (2006) Climate, energy and diversity. Proc Roy Soc B 273:2257–2266 Diniz-Filho JAF, Bini LM, Hawkins BA (2003) Spatial autocorrelation and red herrings in geographical ecology. Glob Ecol Biogeogr 12:53–64 Dutilleul P (1993) Modifying the t test for assessing the correlation between two spatial processes. Biometrics 49:305–314 Evans KL, Warren PH, Gaston KJ (2005) Species-energy relationships at the macroecological scale: a review of the mechanisms. Biol Rev 80:1–25 Field R, Hawkins BA, Cornell HV, Currie DJ, Diniz-Filho JAF, Gue´gan JF, Kaufman DM, Kerr JT, Mittelbach GG, Oberdorff T, O’Brien EM, Turner JRG (2009) Spatial species-richness gradients across scales: a meta-analysis. J Biogeogr 36:132–147 Fleishman E (2010) Understanding species richness gradients informs projected responses to climate change. J Biogeogr 37:1175–1176 Gaston KJ, Blackburn TM, Loder N (1995) Which species are described first?: the case of North American butterflies. Biodivers Conserv 4:119–127 Gotthardt K, Nylin S, Wiklund C (2000) Mating opportunity and the evolution of sex-specific mortality rates in a butterfly. Oecologia 122:36–43 Hawkins BA, Porter EE (2003a) Water-energy balance and the geographic pattern of species richness of western Palearctic butterflies. Ecol Entomol 28:678–686 Hawkins BA, Porter EE (2003b) Does herbivore diversity depend on plant diversity? the case of California butterflies. Amer Naturalist 161:40–49 Hawkins BA, Field R, Cornell HV, Currie DJ, Guegan JF, Kaufman D, Kerr JT, Mittelbach G, Oberdorff T, O’Brien EM, Porter EE, Turner JRG (2003) Energy, water, and broad-scale geographic patterns of species richness. Ecology 84:3105–3117 Heikkinen RK, Luoto M, Kuussaari M, Po¨yry J (2005) New insights into butterfly environment relationships using partitioning methods. Proc R Soc B 272:2203–2210 Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A (2005) Very high resolution interpolated climate surfaces for global land areas. Int J Climato 25:1965–1978 Kerr JT (1999) Weak links: ‘‘Rapoport’s rule’’ and large-scale species richness patterns. Global Ecol Biogeogr 8:47–54 Kerr JT, Southwood TRE, Cihlar J (2001) Remotely sensed habitat diversity predicts butterfly species richness and community similarity in Canada. Proc Nat Acad Sci USA 98:11365–11370
123
Biodivers Conserv Kerr JT, Kharouba HM, Currie DJ (2007) The macroecological contribution to global change solutions. Science 316:1581–1584 Kissling WD, Rahbek C, Bo¨hning-Gaese K (2007) Food plant diversity as broad-scale determinant of avian frugivore richness. Proc R Soc B 274:799–808 Kissling WD, Field R, Korntheuer H, Heyder U, Bo¨hning-Gaese K (2010) Woody plants and the prediction of climate-change impacts on bird diversity. Proc R Soc B 365:2035–2045 Kitahara M, Yumoto M, Kobayashi T (2008) Relationship of butterfly diversity with nectar plant species richness in and around the Aokigahara primary woodland of Mount Fuji, central Japan. Biodivers Conserv 17:2713–2734 Legendre P, Legendre L (1998) Numerical Ecology, 2nd, English edn. Elsevier Science BV, Amsterdam Lennon JJ, Koleff P, Greenwood JJD, Gaston KJ (2004) Contribution of rarity and commonness to patterns of species richness. Ecol Lett 7:81–87 Mene´ndez R, Gonza´lez-Megı´as A, Collingham Y, Fox R, Roy DB, Ohlemu¨ller R, Thomas CD (2007) Direct and indirect effects of climate and habitat factors on butterfly diversity. Ecology 88:605–611 Qian H (2008) Effects of historical and contemporary factors on global patterns in avian species richness. J Biogeogr 35:1362–1373 Qian H, Wang X, Wang S, Li Y (2007) Environmental determinants of amphibian and reptile species richness in China. Ecography 30:471–482 Qian H, Wang S, Li Y, Wang X (2009) Breeding bird diversity in relation to environmental gradients in China. Acta Oecol 35:819–823 Quinn GP, Keough MJ (2002) Experimental design and data analysis for biologists. Cambridge University Press, Cambridge R Development Core Team (2009) R: A language and environment for statistical computing, version 2.12.2. R Foundation for Statistical Computing (online). Available from: http://www.R-project.org Rangel TF, Diniz-Filho JAF, Bini LM (2010) SAM: a comprehensive application for Spatial Analysis in Macroecology. Ecography 33:46–50 Ruggiero A, Hawkins BA (2008) Why do mountains support so many species of birds? Ecography 31:306–315 Ruggiero A, Kitzberger T (2004) Environmental correlates of mammal species richness in South America: effects of spatial structure, taxonomy and geographic range. Ecography 27:401–416 Schuldt A, Assmann T (2009) Environmental and historical effects on richness and endemism patterns of carabid beetles in the western Palaearctic. Ecography 32:705–714 Tateishi R, Ahn CH (1996) Mapping evapotranspiration and water balance for global land surfaces. ISPRS J Photogram Remote Sens 51:209–215 Tello JS, Stevens RD (2010) Multiple environmental determinants of regional species richness and effects of geographic range size. Ecography 33:796–808 Turner JRG, Gatehouse CM, Corey CA (1987) Does solar energy control organic diversity? Butterflies moths and the British climate. Oikos 48:195–205 White PJT, Kerr JT (2007) Human impacts on environment-diversity relationships: evidence for biotic homogenization from butterfly species richness patterns. Global Ecol Biogeogr 16:290–299 Williams SE, Middleton J (2008) Climatic seasonality, resource bottlenecks, and abundance of rainforest birds: implications for global climate change. Divers Distrib 14:69–77 Willig MR, Kaufman DM, Stevens RD (2003) Latitudinal gradients of biodiversity: pattern, process, scale, and synthesis. Annu Rev Ecol Evol Syst 34:273–309 Willott SJ (2001) Species accumulation curves and the measure of sampling effort. J Appl Ecol 38:484–486 Zhang J, Kissling WD, He F (2013) Local forest structure, climate and human disturbance determine regional distribution of boreal bird species richness in Alberta, Canada. J Biogeogr 40:1131–1142
123
