Exploring species richness patterns across space and time can help in understanding species distribution and in formulating conservation strategies. Among taxa, amphibians are of utmost importance as they are highly sensitive to environmental changes due to their unique life histories (Zhong et al., 2018). Here, we investigated the spatial and temporal patterns of amphibian species richness on Tianping Mountain in China. Specifically, we established 10 transects at low to high elevations, and sampled amphibians in April, June, August, and October 2017. Our results demonstrated that amphibian species composition and richness varied significantly at both spatial and temporal scales and were associated with gradients of environmental change in microhabitats on Tianping Mountain.
Biodiversity is a hot topic in community ecological research and exhibits a strong relationship with ecosystem functioning (Wang & Brose, 2018; Zhao et al., 2018). Although biodiversity consists of multiple components, species richness is a fundamental measurement underlying various ecological concepts and models (Gotelli & Colwell, 2011). Exploring species richness patterns across space and time can help to understand the distribution of organisms (Fu et al., 2006), reveal the mechanism of species coexistence (Hu et al., 2011), formulate conservation strategies (Olds et al., 2016), and assess environmental changes (Butchart et al., 2010).
Spatial patterns of species richness are usually tested along altitudinal or latitudinal gradients at the local and regional scales, respectively. At the local scale, species richness typically exhibits two types of response along elevational gradients; i.e., it decreases continuously or displays a hump-shaped relationship with increasing elevation (Rahbek, 1995, 2004). These relationships can be explained by environmental factors such as climate, area, and habitat heterogeneity (Cruz-Elizalde et al., 2016; Hernández-Salinas & Ramírez-Bautista, 2012; McCain, 2009; Rahbek, 1995; Sanders et al., 2003) or can be attributed to mid-domain effects, with increasing species distribution overlap towards the center of a bounded geographic domain free of environmental gradients (Colwell et al., 2004; Wu et al., 2013a). At the larger geographic scale (e.g., latitudinal gradients), species richness can increase from the poles to the tropics (e.g., global data of amphibians, birds, and mammals; Marin & Hedges, 2016). This is likely because tropical areas have greater historical lineages and higher environmental heterogeneity than temperate zones (Jansson et al., 2013; Stevens, 2011). In contrast, marine species richness along latitudinal gradients (for both vertebrates and invertebrates, and all species together) is bimodal, with a dip in richness occurring at the equator (Chaudhary et al., 2017). This relationship is strongly determined by temperature, which can influence animal biology and productivity within ecosystems (Chaudhary et al., 2017).
Temporal environmental fluctuations (e.g., seasonality) can also control species composition and richness (Tonkin et al., 2017). The unique assemblages observed at specific times of the year depend on the ecological conditions of each season (Chesson, 2000), and have been well documented in different fauna. For instance, temporal changes in bird species richness can be attributed to seasonal migration (Somveille et al., 2015). In terms of fish communities, species richness and assemblages can vary seasonally due to minimization of interspecific competition for resources (Shimadzu et al., 2013). Therefore, exploring temporal (seasonal) patterns of species richness can help to understand species coexistence and ecosystem stability (Thibaut & Connolly, 2013).
Although numerous studies have examined spatial (e.g., Acharya et al. 2011; Bhattarai et al., 2004; Fu et al., 2007; Gojo Cruz et al., 2018; Kraft et al., 2011; Khatiwada et al., 2019; Wu et al., 2013a, 2013b) and temporal patterns (e.g., Bender et al., 2017; Shimadzu et al., 2013; Tonkin et al., 2017) of species richness in different taxa separately, empirical studies are still needed for simultaneous quantification. This is because both space and time can affect the distribution and activities of species at the local scale. For instance, species belonging to different spatial guilds may exploit different habitat types. Population size may increase at different times as species can adapt to changes in temporal environmental conditions over their life cycle (Shimadzu et al., 2013). This is especially true for amphibians due to their restricted distribution ranges as well as their seasonal migration for spawning and after metamorphosis (Fei et al., 2009). Therefore, in the present study, we focused on the spatial and temporal (seasonal) changes in species richness in amphibians distributed on Tianping Mountain, China. Specifically, we (1) revealed the amphibian species assemblages along elevational gradients, as well as those in different seasons, (2) explored the environmental factors that determined species composition, and (3) tested the spatial and temporal patterns of species richness. Methodological details are provided in the Supplementary Materials.
In total, 15 species belonging to seven families were recorded during the four sampling sessions (Supplementary Table S1). Results showed that amphibian assemblages changed along the elevational gradients. Overall, the dominant species recorded at the low, mid-, and high elevation transects included Amolops ricketti and Odorrana schmackeri, Paramegophrys liui, and Paa boulengeri and Leptobrachium boringii, respectively (Figure 1A). The amphibian assemblages also exhibited strong temporal shifts during the four seasons. The dominant species in April, June, August, and October were Paramegophrys liui and L. boringii, Megophrys sangzhiensis and Pseudohynobius flavomaculatus, O. margaretae and O. schmackeri, and Paa boulengeri and A. ricketti, respectively (Figure 1B).
Spatial and temporal patterns of amphibian species richness on Tianping Mountain.
The redundancy analysis (RDA) results were significant when testing the effects of environmental factors on amphibian species composition (P=0.005). The first two axes explained 31.73% of the variation (20.13% and 11.60%, respectively). Among the 16 environmental factors, air humidity, water temperature, altitude, tree number, canopy density, shrub coverage, fallen leaf coverage, fallen leaf depth, and water conductivity had significant effects on species composition (P<0.05). Juvenile and adult O. schmackeri, adult A. chunganensis, juvenile and adult A. ricketti, and adult Fejervarya multistriata were positively associated with water temperature and water conductivity, but were negatively associated with altitude, tree number, canopy density, and fallen leaf depth. In contrast, juvenile Bufo gargarizans, adult L. boringii, and adult Paramegophrys liui were positively associated with altitude, tree number, canopy density, and fallen leaf depth, but were negatively correlated with water temperature and water conductivity. In addition, adult Feirana quadranus, adult Pseudorana sangzhiensis, adult O. margaretae, juvenile and adult Paa boulengeri, adult M. sangzhiensis, and adult Rhacophorus chenfui were positively associated with shrub coverage and air humidity. In contrast, adult Hyla gongshanensis and juvenile Pseudohynobius flavomaculatus were negatively associated with shrub coverage and air humidity (Figure 1C).
Overall, total amphibian species richness increased significantly along elevational gradients when incorporating data from all four sampling sessions (R2=0.53, P=0.010; Figure 1D), indicating that more species were detected at high elevation sites. However, the response of species richness to elevational gradients differed each month. Specifically, there were no significant changes in April and August (R2=–0.12, P=0.844; R2=–0.06, P=0.519), but a significant increase and U-shaped relationship were observed in June and October, respectively (R2=0.38, P=0.034; R2=0.70, linear term: P=0.009, quadratic term: P=0.014; Figure 1E). At the temporal scale, total species richness varied significantly among the different months (Friedman test; χ2=16.129, df=3, P=0.001). Specifically, species richness in April was significantly lower than that in August (W=17, P=0.011), but was significantly higher than that in October (W=76, P=0.044). Species richness in October was significantly lower than that in June (W=93, P=0.001) and August (W=97, P<0.001). However, there was no significant difference in species richness between April and June (W=24.5, P=0.052) or between June and August (W=43, P=0.612; Figure 1F).
Overall, our results indicated that amphibian species assemblages varied along elevational gradients. Specifically, A. ricketti and O. schmackeri were the dominant species at the low elevation transects. However, these species, as well as F. multistriata and H. gongshanensis, were restricted to transects below 800 m a.s.l. (i.e., were not recorded at mid- or high elevations). Results also showed that Paramegophrys liui was dominant at the mid-elevation transects, whereas Paa boulengeri and L. boringii were dominant at the high elevation transects. In addition, we found one high elevation-restricted species (Pseudorana sangzhiensis). These results support previous claims that elevations associated with specific habitat conditions (e.g., temperature and vegetation variation) are important for determining amphibian species distribution (Khatiwada et al., 2019; Meza-Joya & Torres, 2016) and provide specific spatial niches for species. We also found that temporal niche utilization differed among the amphibians, with dominant species changing from Paramegophrys liui and L. boringii in April, to Pseudohynobius flavomaculatus in June, O. schmackeri in August, and Paa boulengeri and A. ricketti in October. Some species were also seasonally restricted. For instance, Paramegophrys liui was not detected after June, O. schmackeri and M. sangzhiensis were only active in June and August, and Pseudorana sangzhiensis was only recorded in August. These results could be attributed to breeding times and temperature adaptations (Fei et al., 2009; Snyder & Weathers, 1975). Therefore, our results are in agreement with previous studies, which state that Paramegophrys liui and L. boringii breed in April–July and February–April, respectively (Fei et al., 2009, 2012), whereas O. schmackeri and M. sangzhiensis usually breed in June–August (Fei et al., 2009, 2012). We also sampled several species in all seasons (e.g., Paa boulengeri and A. ricketti), which may be related to their broader temperature adaptation range (Fei et al., 2009, 2012).
Previous studies have indicated that Paramegophrys liui and L. boringii usually require high elevation habitats with broad vegetation coverage (Fei et al., 2009, 2012), which provide good breeding sites. This may explain why their abundance was positively correlated with environmental factors such as canopy density, fallen leaf depth, and tree number. However, their abundance was negatively correlated with water temperature as their breeding season occurs from March to June under relatively low temperatures (Fei et al., 2012; Yu & Lu, 2010). Juvenile B. gargarizans exhibited similar patterns as P. liui and L. boringii, which was probably because they recently underwent metamorphosis from water bodies. However, A. chunganensis, A. ricketti, and O. schmackeri were negatively associated with altitude, as these species prefer large, strong-flowing streams with low vegetation cover, which were more common at the lower elevation transects with higher temperatures (Wu et al., 2015; Xiong et al., 2005). The observations for F. multistriata at low elevations could be attributed to their high abundance in and preference for lowland paddy fields close to large streams. Unsurprisingly, F. quadranus, Pseudorana sangzhiensis, O. margaretae, Paa boulengeri, M. sangzhiensis, and R. chenfui were positively associated with shrub coverage and air humidity, as these conditions provide diverse food items such as insects, snails, and crabs (Huang et al., 2011; Yuan & Wen, 1990). In contrast, H. gongshanensis and juvenile Pseudohynobius flavomaculatus were negatively correlated with air humidity and shrub coverage as they inhabit paddy fields close to streams (Liao & Lu, 2010) and small ponds covered by rocks (Fei et al., 2009), respectively.
In contrast with many earlier studies (e.g., Hu et al., 2011; Khatiwada et al., 2019), total amphibian species richness increased continuously along the elevational gradients, with more species detected at higher elevations. This observation is interesting as most previous studies suggest that amphibian species richness should exhibit a hump-shaped response to elevation due to mid-domain effects (e.g., amphibians in Hengduan Mountain, China; Fu et al., 2006; amphibians in tropical Andes; Meza-Joya & Torres, 2016) or decrease continuously with elevation (e.g., eastern Nepal Himalaya; Khatiwada et al., 2019). This is because high elevations (e.g., >3 000 m on Hengduan Mountains and eastern Nepal Himalaya; Hu et al., 2011; Khatiwada et al., 2019) usually correspond to low temperatures, with fewer amphibian species able to survive in cold regions (Funk et al., 2012). Our conflicting results could be attributed to the limited elevation of transects selected on Tianping Mountain (<1 500 m), with temperature not a primary limiting factor of amphibian species richness at the spatial scale. In addition, our results could also be attributed to the larger heterogeneity of habitats at higher elevations on Tianping Mountain (Cao et al., 1997), which could support more amphibian species (Hernández-Salinas & Ramírez-Bautista, 2012; Meza-Joya & Torres, 2016). It was not surprising that more species were observed in June and August as these months occur in the active season (warmer and wetter climate conditions) for most amphibian species, with their populations increasing in suitable habitats (Bickford et al., 2010).
In conclusion, the present study demonstrated the spatial and temporal patterns of amphibian species richness on Tianping Mountain in Hunan Province, China. Future studies should focus on more facets of biodiversity to better understand the roles of spatial and temporal variation in community assembly processes of mountain-dwelling amphibians. As the spatial and temporal niches of amphibian species were different, specific conservation strategies should be implemented. Furthermore, we confirmed that amphibian species occurrence was strongly determined by biotic and abiotic features of their microhabitats, which mediated species composition along the elevational transects. Because microhabitats are easily affected by human disturbance, long-term monitoring should be conducted to investigate the relationship between amphibian diversity and environmental change.