INTRODUCTION
Describing the pattern and variations in spatial pattern of biodiversity and revealing its underlying mechanisms remain a central focus in ecology. With the popularization of the concept of biodiversity (e.g., species richness), related research is developing rapidly (Wilson & Ehrlich, 1991). But less attention has been paid to the ultimate cause of species richness: species range size (e.g., geographic distribution, latitudinal range, elevational range). The degree of overlap of the range size forms the species richness pattern we have observed, and the higher the degree of overlap of species range in a certain area, the higher the species richness. Moreover, species range size can reflect the niche of species (Pfenninger, Nowak, & Magnin, 2007). It is also correlated with the formulation in species protection strategies (Herzog, Oswaldo, Embert, Caballero, & Quiroga, 2012), especially the conservation of faunas (Kevin J Gaston, 1996) and the assessment of the probability of extinction (Price, Helbig, & Richman, 1997). However, the underlying mechanism shaping species range size is unclear (Kevin J Gaston, 1998; Webb & Gaston, 2003).
Most studies on species range size focused on examining the positive relationship between communities range size (species mean range size) and elevation (or latitude), which is called Rapoport’s rule (Rapoport, 1975, 1982; Stevens, 1989, 1992; Böhm et al., 2017; Colwell & Hurtt, 1994; Kevin J Gaston, 1999; K J Gaston, Blackburn, & Spicer, 1998; Hausdorf, 2006; Mccain & Knight, 2013; Rohde, Heap, & Heap, 1993; Zhou et al., 2019). However, few studies have explored the drivers of species range size (but see J.-Y. Kim et al., 2019; Letcher & Harvey, 1994; Luo et al., 2011; Stevens, 1989, 1992). Species range size could be related to different factors such as climate, ambient energy, and habitat heterogeneity which referred to the climate variability hypothesis, the ambient energy hypothesis, and the habitat heterogeneity hypothesis. The climate variability hypothesis states that species living in areas with larger variations in climatic conditions (e.g., temperature) tend to have wider tolerances; that is, climate stability is negatively correlated with species range size (Stevens, 1989, 1992). The ambient energy hypothesis predicts that environmental energy (e.g., productivity) determines the degree of overlap of species range size (Kerr & Packer, 1997). For example, species’ interaction is more frequent in the tropical regions, as the tropical regions receive more environmental energy than the temperate regions (environmental energy is negatively related to latitude), which can support finer niche separation and the coexist of more species, leading to smaller species range size (Brown, 1981). The habitat heterogeneity hypothesis, which is an extension or supplement of the ambient energy hypothesis, states that the greater heterogeneity of the spatial and topographical habitat structure allows for finer subdivision of restricted energy, suggesting that the higher the habitat heterogeneity, the smaller the species range size could be (species range size, to some extent, is a surrogate of species niche) (J.-Y. Kim et al., 2019). However, the influence of habitat heterogeneity on the species range size may be closely related to the combination of habitat types (Hu et al., 2017). The fit of these hypotheses above to different taxonomic groups and gradients is still uncertain and needs to be further tested in more biogeographical regions.
Local species richness can be strongly impacted by the proximity of the species’ range margins of potentially interacting species. The higher the degree of the proximity of the species’ range margins of potentially interacting species (which can be defined as inflow intensity, J.-Y. Kim et al., 2019) can increase the species richness. This phenomenon was called rescue effect (Stevens, 1992). Moreover, different shapes of richness patterns (e.g., monotone decreasing and unimodal pattern) may be influenced by kinds of rescue effects (i.e., Steven’s rescue effect, alternative rescue effect, and non-directional rescue effect). The premise of Steven’s rescue effect (or Rapoport’s rescue effect) is that if species in high elevation (or latitude) region have a larger range size (the distribution of species range size follows the Rapoport’s rule), species overlap in low elevation (or latitude) region will increases (species from higher elevation or latitude tend to inflow to low elevation or latitude) (Stevens, 1992). While alternative rescue effect proposes that decreasing species richness along elevation or latitude could result from differential upper limits of species with source populations below mid-point of the gradient even if the species have a larger range size in the low elevation or latitude (Almeida-Neto, Machado, Pinto-da-Rocha, & Giaretta, 2006). Besides, non-directional rescue effect (Almeida-Neto et al., 2006) argued that species from both ends of a gradient (if species from both high and low elevation have larger range sizes) would lead to an increase of species richness in the middle elevation and resulting in a unimodal species richness pattern.
Mountain ecosystems are hot spots in biological diversity (Lomolino, 2001). The environmental conditions (e.g., climate, habitat, and productivity) vary remarkably along elevations over small spatial extents, making elevational gradient an outstanding natural laboratory for studies in ecology and biogeography. Here, we used a comprehensive dataset of breeding birds collected from 2018 to 2019 along an elevational gradient in the Eastern Himalayas of China. Our objectives were to a) assess how species mean elevational range size changes along elevation; b) explore the drivers influencing species mean elevational range size; c) explain the relationship between species range size and species richness (rescue effect) to expand our understanding of underlying mechanisms of community structure and species richness pattern.