1. Introduction

Species turnover between areas, which is commonly called beta diversity or β-diversity, quantifies the change in species composition among areas across space. For a given species diversity in local areas (i.e., α-diversity) within a region, greater beta diversity between local areas would lead to greater species diversity in the region (i.e., γ-diversity). Thus, beta diversity is considered as a link between local (α) and regional (γ) species diversity. Understanding patterns of beta diversity is central to ecology and biogeography (Whittaker, 1977).

Several studies have suggested that beta diversity should be higher for organisms with larger body sizes, which is called the beta diversity−body size (BDBS) hypothesis, because organisms with smaller body sizes are considered to be better dispersers (Drakare et al., 2006) and thus would be expected to have lower rates of species turnover between areas, compared with organisms with larger body sizes. This is because increasing dispersal makes assemblages in different areas more similar to each other and thus reduces species turnover among areas (Drakare et al., 2006). Thus, the BDBS hypothesis depends on the relationship between body size and dispersal ability, which has been frequently highlighted (Fenchel et al., 1997; Finlay et al., 1996). The hypothesis that larger organisms have greater species turnover has been supported by multiple empirical studies. For example, analyzing a dataset including both unicellular (diatoms, desmids, and ciliates) and multicellular (nematodes, bivalves, and polychaetes) across the world, Hillebrand et al. (2001) observed that the similarity of assemblages decreased with increasing geographic distance for all organisms examined in their study, but the rate of this decay of similarity over geographic distance (and thus the increase in species turnover) is much higher for large organisms than for small organisms. Using a dataset including both plants and animals across the world, Drakare et al. (2006) conducted a meta-analysis on the species−area relationship, the slope of which represents species turnover (i.e., steeper slope indicating greater species turnover), and they observed that the slope of the species–area relationship becomes steeper with increasing species’ body size.

In vascular plants, Qian (2009a) observed that beta diversity of pteridophytes is lower than that of seed plants in North America. Because pteridophytes are on average much smaller than seed plants (Flora of North America Editorial Committee, 1993–2024), Qian’s (2009a) observation supports the BDBS hypothesis. However, there is no study examining whether this hypothesis is supported by species within seed plants, which account for over 96% of vascular plant species worldwide (Qian et al., 2022).

Here, taking the advantage of the information available in the plant database of North America (Kartesz, 1999; Qian and Ricklefs, 2007; Qian, 2009a), I test the hypothesis that beta diversity increases with increasing body size in seed plants. Specifically, I compare beta diversity and its relationships with geographic distance and climatic difference between areas among three groups of plants across a gradient of body sizes. In addition to examining the relationship across North America, I also examine the relationship for each of different latitudinal zones, as in Qian and Ricklefs (2007) and Qian (2009a).

2. Material and methods

The geographic area of this study covers the entirety of North America north of Mexico, which included 71 geographic units primarily being states or provinces. They were divided into five latitudinal zones (Zones A through E from south to north; Fig. S1) in order to examine the effect of latitude on beta diversity. Details about the geographic units and latitudinal zones were provided in Qian and Ricklefs (2007) and Qian (2009a). Because of the small number of geographic units in Zone E (located largely in the Arctic), I excluded this zone from analyses at the zonal level, as in Qian and Ricklefs (2007) and Qian (2009a).

Species lists of seed plants for the geographic units were generated based on Kartesz’s (1999) North American plant database, which were supplemented with additional data as described in Qian (2009a). Species exotic to North America were not included. A total of 15,715 species (including hybrids) of seed plants were included, which included 1068 species of trees, 1673 species of non-tree woody plants (hereafter collectively called ‘shrubs’), and 12974 species of herbs, according to Kartesz’s (1999) database.

Beta diversity between pairwised geographic units was defined as Simpson's dissimilarity (β_sim), which is = 1 – Simpson similarity index (Koleff et al., 2003). Simpson similarity index is defined as a/(a + min(b,c)), where a is the number of species shared between two geographic units and b and c are the numbers of species unique to either geographic unit (Simpson, 1960). This beta diversity index reflects true species turnover (Baselga, 2010). I calculated beta diversity for each of the three groups of plants (i.e., trees, shrubs, and herbs).

Previous studies have shown that beta diversity in plants is strongly associated with geographic and climatic distances between areas (Qian et al., 2020; Zhou et al., 2023). Accordingly, I related beta diversity to geographic and climatic distances. Geographic distance between two geographic units was the Euclidean distance between the centroids of the geographic units. With regard to climatic distance, because previous studies have shown that mean annual temperature and annual precipitation are the most important climatic factors determining large-scale distributions of plants, I calculated climatic distance based on these two climatic variables between each pair of geographic units. Specifically, I first used z-score to normalize each of the two climatic variables to have a mean of zero and a SD of one, and then calculated the Euclidean distance of the two climatic variables between each pair of geographic units. I used the same data of the two climatic variables as in Qian and Ricklefs (2007) for the 71 geographic units.

For each group of plants in each latitudinal zone, I conducted ordinary least squares regressions to assess latitudinal variation in the slope of the relationship between beta diversity and geographic distance (i.e., distance decay); I conducted a series of variation partitioning (Legendre and Legendre, 2012) to assess the variance in beta diversity explained jointly by geographic and climatic distances, independently by geographic distance, and independently by climatic distance. All statistical analyses were carried out using SYSTAT v.10 (Wilkinson et al., 1992).

3. Results

Across the entire study area (North America north of Mexico), the average beta diversity of the 2485 pairs of the 71 geographic units was 0.547 (±0.270 SD), 0.640 (±0.264 SD) and 0.568 (±0.234 SD) for trees, shrubs and herbs, respectively. Pearson's correlation coefficient between beta diversity and geographic distance was 0.686, 0.703 and 0.725 for trees, shrubs and herbs, respectively; and Pearson's correlation coefficient between beta diversity and climatic distance was 0.604, 0.653 and 0.704 for trees, shrubs and herbs, respectively. In other words, the correlation of beta diversity with either distance was strongest in herbs and weakest in trees among the three groups of plants.

When the four latitudinal zones were considered separately, the average of beta diversity within each latitudinal zone was highest in shrubs and lowest in trees (Fig. 1), which is consistent with the aforementioned result for the entire study area. For each of the three plant groups, the average beta diversity within a latitudinal zone decreased with increasing latitude (Fig. 1).

When beta diversity was related to geographic distance for geographic units within each latitudinal zone, shrubs had the strongest relationship (steepest slope) between beta diversity and geographic distance in three of the four latitudinal zones (i.e., Zones A through C), and trees had the strongest relationship in the other zone (Fig. 2). For each of the three plant groups, the slope of the relationship between beta diversity and geographic distance decreased strongly and monotonically with increasing latitude, ranging from 0.31 to 0.14, from 0.32 to 0.12, and from 0.28 to 0.09, respectively, for trees, shrubs, and herbs (Fig. 2).

Across North America, geographic and climatic distances together explained 52.1%, 56.9% and 62.9% of the variation in beta diversity in trees, shrubs and herbs, respectively. The variation explained jointly by geographic and climatic distances was much larger than that explained independently by either geographic or climatic distance, and the variation explained independently by geographic distance was greater than that explained independently by climatic distance in all the three groups of plants (Fig. 3).

4. Discussion 4.1. Relationship between body size and beta diversity

The present study finds that the average beta diversity across North America is highest in shrubs and lowest in tress. This finding holds true when different latitudinal zones were considered separately. The slope of the relationship between beta diversity and either geographic or climatic distance is steepest in herbs and least steep in trees across North America. Of the four latitudinal zones, trees have the steepest slope of the relationship between beta diversity and geographic distance only in the northernmost zone. Thus, these findings are in general contrary to the hypothesis that beta diversity increases with increasing body size in seed plants, although our finding that shrubs have higher beta diversity than herbs is consistent with the BDBS hypothesis.

It is not clear why beta diversity in trees is lower than that in shrubs and herbs in North America. One possibility may be that many tree species are dispersed by wind. For example, many gymnosperm trees (e.g., species in the genus Pinus) and angiosperm trees (e.g., species in the genera Acer, Platanus and Ulmus) have winged seeds that help them travel long distances in the wind, reducing beta diversity.

Previous studies also showed mixed results on whether beta diversity increases with increasing body size. For example, Qian (2009a) found that seed plants have higher beta diversity than pteridophytes in North America. Because all pteridophyte species in North America are herbaceous whereas seed plants include not only herbaceous plants but also trees and shrubs, the average body size of pteridophytes would be smaller than that of seed plants. Furthermore, pteridophytes are better dispersers than seed plants because the propagule size of pteridophytes is much smaller than that of seed plants—pteridophyte propagules are spores (usually < 0.1 mm in length; Adsersen 1995) and are capable of dispersing thousands of kilometers by wind (Wolf et al., 2001) whereas seed plant propagules are seeds or fruits usually being larger than pteridophyte spores by several orders of magnitude (Adsersen, 1995). Accordingly, the finding of Qian (2009a) supports the BDBS hypothesis. Hillebrand et al. (2001) observed that compared to multicellular taxa, unicellular taxa have lower slopes of the species−area relationships and the species composition of unicellular taxa is less influenced by geographic distance, which both indicate that the beta diversity is lower in unicellular taxa than in multicellular taxa. Because body size in unicellular taxa is on average smaller than that in multicellular taxa (Hillebrand et al., 2001), the finding of Hillebrand et al. (2001) also supports the BDBS hypothesis. In a meta-analysis using 794 species−area relationships involving various groups of organisms, Drakare et al. (2006) found that the slope of the species−area relationship is steeper in organisms with larger body size. Because a steeper slope of the species−area relationship represents a greater beta diversity (Drakare et al., 2006), the finding of Drakare et al. (2006) also supports the BDBS hypothesis.

Qian (2009b) found that the beta diversity of birds and mammals is lower than that of reptiles and amphibians at a global scale, and explained this pattern as birds and mammals have better dispersal ability than reptiles and amphibians. However, the order of the beta diversity of these four groups of terrestrial vertebrates, which is reptiles > amphibians > mommals > birds (Qian 2009b), does not match the order of their mean body sizes, which is mammals > birds > reptiles > amphibians (based on the body size data analyzed in Guo et al., 2026). In a meta-analysis based on the results of 99 studies including both animals and plants (e.g., vertebrates, insects, bryophytes, and vascular plants), Soininen et al. (2018) showed a negative relationship between species turnover and body size. Taken together, in general, these findings do not support the BDBS hypothesis.

The relationship between the beta diversity and body size is built on the hypothesis that organisms with smaller body sizes have better dispersal abilities and thus are expected to be distributed more widely, which reduce species turnover, compared to organisms with larger body sizes. However, several studies have shown that organisms with better dispersal ability do not have lower beta diversity. Furthermore, based on comparisons among birds, bats, non-flying mammals, reptiles and amphibians (Calderón-Patrón et al., 2013) and between non-flying and flying organisms (Harrison et al., 1992), several studies at regional scales did not find an effect of dispersal ability on beta diversity. This suggests that other factors may play a stronger role than dispersal ability in driving species turnover across space.

4.2. Relationships of beta diversity with geographic and climatic distances

Qian (2009a) showed that in both pteridophytes and seed plants in North America, beta diversity decreases and the slope of the relationship between beta diversity and geographic distance decreases with increasing latitude. The present study extended Qian (2009a) by examining these relationships separately for trees, shrubs, and herbs, and found the same relationships as in Qian (2009a). Specifically, in each of the three plant groups, with increasing latitude, beta diversity decreases and the slope of the relationship between beta diversity and geographic distance also decreases. Whether these general relationships hold true for plants in other continental regions or for other groups of organisms within and beyond North America needs to be tested in future studies.

I find that beta diversity is associated with geographic distance more strongly than with climatic distance, regardless of whether tress, shrubs, or herbs are considered. This finding is consistent with the findings of many previous studies covering broad spatial extents. For example, Qian et al. (2020) found that in angiosperms in China the variation in species turnover that is explained independently by geographic distance is greater than that explained independently by climatic distance. Similarly, Qian et al. (2024) found that in Africa the variation in species turnover of angiosperms that is explained independently by geographic distance is greater than that explained independently by climatic distance. Mechanisms causing the difference between the roles of geographic and climatic distances driving species turnover likely differ among taxa and regions. In North America, the continental ice sheets covered the northern part of the continent during the Last Glacial Maximum, therefore, nearly all species currently present in the northern half of the continent were dispersed into the region from the southern part of the continent and the unglaciated areas of the northwesternmost part of the continent after the retreat of the ice sheets. Northern distribution limits of some species are not in equilibrium with current climate (Johnstone and Chapin, 2003). Furthermore, while this study included the key climatic factors shaping plant distributions (i.e., mean annual temperature and annual precipitation), other climatic factors may explain additional variation in beta diversion in seed plants. These factors may have, to some degree, caused weaker association of beta diversity with climatic distance than with geographic distance in this study.

4.3. Limitations

I acknowledge several limitations in this study. First, because data for body size across a continuous body size gradient (such as plant height) are not available, this study used the three growth forms of plants (i.e., tree, shrub, and herb) to differentiate plant body sizes. This coarse classification may mask within-group differences in dispersal capacity, niche breadth, or beta diversity, undermining the rigor of the body size gradient test, although it remains true that the average body size of trees is largest and that of herbs is smallest. Second, this study reported patterns of beta diversity among the three categories of body size without exploring underlying drivers, due to lack of data. For example, this study reported that tree beta diversity is lower than that of shrubs and herbs. Although we know that some traits (e.g., generation times, animal-mediated seed dispersal, niche breadth), historical processes (e.g., post-glacial colonization dynamics), and contemporary ecological interactions can modulate the relationship between body size and species turnover (Ricklefs, 2004; Rosenzweig, 1995), it remains unknown how they independently and jointly drive beta diversity. If wind-dispersal seeds is dominant in trees while animal-dispersed seeds are dominant in shrubs, this may cause wind-dispersed trees to exhibit lower turnover than animal-dispersed shrubs. Third, the BDBS hypothesis relies on the assumption that body size correlates with dispersal capacity, which is determined by seed size, seed weight, and seed dispersal mode (e.g., wind, animal, gravity). Because data on seeds for most species of this study are lacking, I am not able to determine the relationship between seed size and beta diversity. Nevertheless, previous studies showed a positive relationship between seed size and plant size (Kidson and Westoby, 2000), suggesting that body size may be used as a surrogate of seed size to some degree. Future studies on the relationship between dispersal capacity and beta diversity should take these limitations into account when relevant data are available.

5. Conclusions

The present study is the first using seed plants to test the hypothesis that organisms with larger body sizes have higher beta diversity (species turnover), by examining the relationship in three groups of seed plants with different body sizes (trees > shrubs > herbs). While beta diversity in shrubs is higher than that in herbs, supporting the hypothesis, beta diversity in trees is lower than that in shrubs and herbs, contrary to the hypothesis. These findings hold true regardless of North America north of Mexico is considered as a whole or different latitudinal zones are considered separately. Thus, the overall finding of this study only partially supports the hypothesis. This study also shows that in all the three groups of plants, beta diversity decreases with increasing latitude, and is more strongly related with geographic distance than with climatic distance.

Acknowledgements

I thank anonymous reviewers for their helpful comments.

Data accessibility statement

Data used in this study were published and cited in the manuscript.

CRediT authorship contribution statement

Hong Qian: Conceptualization, Data curation, Writing − original draft, Writing − review & editing, Investigation, Formal analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pld.2026.03.007.