This study aimed to evaluate the hypothesis that postulates an effect of environmental variables as promoters of ecological changes which impact on the coexistence of morphologically similar and syntopic frugivorous species. What can generate competitive exclusion when resources are limited or, on the contrary, can promote greater tolerance between species in those environments with greater diversity and availability of resources. Specifically, we expected that life zones with more tropical climates would promote greater coexistence of Sturnira species, expressed as greater equity in assemblages, by containing a greater diversity of plant species. Furthermore, we expected that in temperate and cold regions assemblages would exhibit low equity with dominance of one specie. On the other hand, we expected that early successional-stage forests would promote greater species coexistence by increasing the diversity of early-successional-stage plants such as Solanum, which fruits that are the core of the diet of these species (Fleming 1986; Sánchez et al. 2012b; Saldaña-Vázquez et al. 2013; Sánchez and Giannini 2018). Surprisingly, our results did not support both predictions. Contrary to the first prediction of our hypotheses, we found high evenness across all life zones analyzed in both tropical and temperate forests. Being the "cool temperate moist forest" life zone the one with the greatest evenness (0.92) and richness of species presented with four coexisting species on average, which represents the highest average of Sturnira species for our sample.
Traditionally, attempts have been made to explain the patterns of regional diversity and abundance of bats based on environmental variables such as productivity, temperature, climatic seasonality, latitude, altitude, geographic area, and environmental energy (Lyons and Willig 1999; Willig et al. 2003; Willig and Bloch 2006; Tello and Stevens 2010; Stevens 2011; Sánchez and Giannini 2014) or by conservation of the ecological niche over time (Ramos Pereira and Palmeirim 2013). According to these works, these factors would go down the scale, directly affecting the local diversity and complexity of the assemblages (Tello and Stevens 2010). For its part, from a biogeographical approach it is expected that the local composition of species depends on the regional set of species, which is determined by large-scale biogeographical processes. In this way, the specific composition of an assemblage or assemblage may result from the biogeographical patterns that act on the processes of speciation, extinction, and dispersion, rather than the result of a local ecological limitation (Wiens and Donoghue 2004; Wiens 2012).
An example of this is the uplift of the Andes and its role in the diversification of the Amazonian biota (Hoorn et al. 2010; Wiens et al. 2012) or the Mid-Miocene Climatic Optimum and its impact on temperatures. and rainfall on a global scale, which strongly altered the diversity of the Neotropical flora (Hoorn et al. 2010). In our case, the coexistence of species of the genus Sturnira in a particular region or life zone may also be the result of sympatric speciation events or post-speciation dispersal (Halas et al. 2005). This could explain our results and the high diversity we detected in the “cool temperate moist forest” life zone. In fact, biogeographical studies of the genus Sturnira indicate that the genus diversified in the northern region of the Andes where the “cool temperate moist forest” life zone occurs (Velazco and Patterson 2013). In addition, the colonization of lowlands in the Pacific and Mesoamerica occurred secondarily and later. Interestingly, most of the extant lineages arose recently during the Late Pliocene and Pleistocene (3.5 MY) with in situ speciation events (in the Andes) and with up to 40 dispersal and recolonization events (post-speciation dispersal), denoting a strong biogeographical effect on the current distribution of the species, evidencing the strong historical component of the evolution of the genus that explains the richness and co-occurrence of the species in the current assemblages (Velazco and Patterson 2013). This does not exclude the importance of proximal or local factors such as the availability and diversity of resources, which are decisive in the maintenance of the assemblages.
Although life zones and disturbance type had no effect on the evenness of Sturnira species, some intrinsic factors that were not evaluated in this study could explain the high evenness observed in subtropical and temperate regions. For example, the dietary nucleus of the Sturnira genus is made up of plants of the Solanum genus, which is typical of successional stages of secondary vegetation and has periods of extended fruiting (Fleming 1986; Sánchez et al. 2012b). This could generate a prolonged availability of resources that could promote the high equity that we observe. In addition, Solanum is a genus of plants whose center of diversity and endemism is found in the Andes (Echeverría-Londoño et al. 2020), where there is a marked and strong climatic seasonality of temperature and rainfall (Garreaud 2009). For example, Solanum and Cestrum (Solanaceae) were the most diverse and dominant genera in the understory of the evergreen cloud forest from Argentina (Giannini 1999). These plant species dominances influence the diet composition of Sturnira in the highest and coldest sectors of the mountain wet forest (Sánchez and Dos Santos 2015). Interestingly, the life zones "Polar desert, Subtropical dry forest, Cool temperate moist forest and Tropical dry forest" were the ones that presented the highest evenness Sturnira values, which could suggest a dominance of Solanum in these bat-plant interaction systems and explain the equity values we find. Hence, further studies involving diet sampling and availability of resources in these life zones are necessary to test this assumption.
Other intrinsic factors that were not evaluated are the foraging and roost behavior of the species of the genus Sturnira. These factors, together with the high availability of food, have been proposed as variables that explain the coexistence of fruit bat species. For example, the high availability of pioneer plants of the genus Piper, and differential foraging behavior (a species emerges from roosts earlier) and diurnal roosting (a species perches closer to resources) are factors that promote coexistence in syntopic species of the genus Carollia in montane forests of Ecuador (Bonaccorso et al. 2007). The seasonal variation in the abundances of each species of frugivore that makes up the assemblages is also a factor that must be considered, since small changes in the dominance of the species throughout a year may be key to allowing the co-occurrence of similar species in the assemblages. It has been detected changes in the dominance of Sturnira lilium and S. erythromos between samples from the same site and throughout a seasonal period in the montane forests of Argentina (Sánchez et al. 2012a). Similarly, it has been detected marked monthly and seasonal differences in the dominance of the trophic guilds of phyllostomid bats in the Atlantic Forest of Brazil, showing that Neotropical bat assemblages are organized in a very dynamic way (Mello 2009).
In conclusion, the determining ecological implications in the coexistence of the sympatric species of the genus Sturnira on a large scale evaluated in this work, continue to be a mystery, and show the complex and multivariate set of factors that could be promoting the high evenness that we observe in the different life zones and habitats. However, it is important to deepen future studies in different perspectives such as the physiology of metabolism and the ecology of migration, considering the wide range of altitudinal ranges that these species present to minimize competition for the availability of food resources. On the other hand, climatic seasonality, and its association with plant genera with its hotspot in the Andes could be related to the high evenness of the species of the genus Sturnira and its response in the different life zones evaluated by minimizing temporal overlap.