Oecologia https://doi.org/10.1007/s00442-021-05011-9 COMMUNITY ECOLOGY – ORIGINAL RESEARCH Temperature and productivity distinctly affect the species richness of ectothermic and endothermic multitrophic guilds along a tropical elevational gradient Chaim J. Lasmar1 · Clarissa Rosa2 · Antônio C. M. Queiroz1 · Cássio A. Nunes3 · Mayara M. G. Imata1 · Guilherme P. Alves1 · Gabriela B. Nascimento1 · Ludson N. Ázara4 · Letícia Vieira5 · Júlio Louzada6 · Rodrigo M. Feitosa7 · Antonio D. Brescovit8 · Marcelo Passamani9 · Carla R. Ribas1 Received: 30 January 2021 / Accepted: 3 August 2021 © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2021 Abstract The diversity of endotherms and ectotherms may be differently affected by ambient temperature and net primary productivity (NPP). Additionally, little is known about how these drivers affect the diversity of guilds of different trophic levels. We assessed the relative role of temperature and NPP in multitrophic guilds of ectothermic (arthropods: ants, ground beetles, spiders, and harvestmen) and endothermic (large mammals) animals along a tropical elevational gradient. We sampled arthropods at eight elevation belts and large mammals at 14 elevation belts in Atlantic rainforest (ranging from 600 to 2450 m.a.s.l.) of Itatiaia National Park, Southeast Brazil. Overall arthropod species richness was more associated with temperature than overall large-mammal species richness, while the latter was more associated with NPP. When separated into trophic guilds, we found that the species richness associated with NPP increased across arthropod trophic levels from herbivores to predators. Conversely, although NPP influenced large-mammal herbivore species richness, its effects did not seem to accumulate across large-mammal trophic levels since the species richness of large-mammal omnivores was more associated with temperature and none of the variables we studied influenced large-mammal predators. We suggest that thermal physiological differences between ectotherms and endotherms are responsible for the way in which arthropods and large mammals interact with or are constrained by the environment. Furthermore, the inconsistency regarding the role of temperature and NPP on species richness across multitrophic guilds of ectotherms and endotherms could indicate that thermal physiological differences might also interfere with energy use and flux in the food web. Keywords Community ecology · Elevational gradient · Species richness · Trophic ecology · Tropical mountain Introduction The evaluation of patterns of species diversity across ecological gradients has been a core focus of ecology studies (MacArthur 1972; Fine 2015). Several hypotheses have been developed to explain how species diversity varies in space based on the influence of historical and ecological factors (Fine 2015). Among them, two main ecological factors, temperature, and net primary productivity (NPP), help to explain how energy in an ecosystem may drive patterns Communicated by Nina Farwig. * Chaim J. Lasmar chaimlasmar@gmail.com Extended author information available on the last page of the article of species richness (Wright 1983; Hawkins et al. 2003a; Evans et al. 2005; Brown 2014). However, their effects are complex: for instance, although both temperature and NPP positively affect the species richness of both endotherms and ectotherms (Hillebrand 2004; Brown 2014), differences in the thermal physiology between these organisms lead to different mechanisms regulating species diversity patterns (Allen et al. 2002; Buckley et al. 2012). Additionally, as both temperature and NPP are related to how energy enters and flows through an ecosystem, their effects vary in relative importance when dealing with diversity patterns among different trophic guilds (Voigt et al. 2003; Welti et al. 2020). The high resource assimilation at high temperatures and the less intense competition caused by the high availability of resources (e.g., high NPP) increase the population sizes of species, which in turn decreases extinction rates 13 Vol.:(0123456789) Oecologia and allow more species to coexist (Evans et al. 2005; Brown 2014). Temperature affects animals through a positive influence on their metabolic rates, up to a point (Gillooly et al. 2001), which also increases speciation rates (Brown 2014). Endotherms are usually bigger than ectotherms, and smaller organisms typically lose too much heat (Buckley et al 2012). This is especially true for ectotherms, such as arthropods, since their body temperature is dependent on environmental temperature (Lessard et al. 2011; Brown 2014). Even in environments with high resource availability, ectotherms are unable to obtain and assimilate energy from the ecosystem at low temperatures (Sanders et al. 2007; Buckley et al. 2012). On the other hand, endotherms acquire resources more independently of external temperatures than ectotherms because they are able to control their body temperature. However, endothermy comes at a high cost (10 times higher than those of ectotherms for energetic maintenance), which results in endotherms obtaining a greater amount of energy from food resources (Buckley et al. 2012). Therefore, the higher consumption of resources possibly brings endotherms faster to the limit of ecosystem resources. Thus, we could expect that physiological constraints caused by temperature would be more strongly associated with diversity patterns of ectotherms than endotherms (Buckely et al. 2012). In contrast, endotherms would be more constrained by NPP effects (Buckely et al. 2012), leading to greater species richness in more productive environments, as reported for large and small mammals along elevational gradients (e.g., McCain et al. 2018; Gebert et al. 2019). Accordingly, both NPP and temperature may also affect how energy flows across trophic guilds in an ecosystem (Evans et al. 2005; Birkhofer and Wolters 2012) and operates among trophic levels (Voigt et al. 2003, 2007; Welti et al. 2020). Through accelerated kinetics and increased metabolic rates, temperature positively affects the assimilation of resources for many trophic levels, from herbivores to top predators (Moorthi et al. 2016; Binkenstein et al. 2018). In more productive environments, the greater abundance of animals in low trophic levels (e.g., herbivores) is caused by the high energy input in the ecosystem; these abundant low trophic level animals serve as resources for animals at higher trophic levels (e.g., predators), resulting in an increase in the abundance of these animals (Evans et al. 2005). In this bottom-up effect, high animal abundance at all trophic levels decreases extinction rates and allows for more diversity in the whole food chain (Turney and Buddle 2016). Although predators also control the abundance and diversity of lower trophic levels (Terborgh 2015), they are more sensitive to the energy input and flux in the food chain (Voigt et al. 2007; Turney and Buddle 2016; Brose et al. 2017). The energy lost from one trophic level to another results in the increase of resource limits, in which predators are the ultimate energy receptors in the ecosystem. Previous studies have reported 13 that arthropods at higher trophic levels are more sensitive to changes in NPP than species at lower trophic levels (Kaspari 2001; Haddad et al. 2009). Thus, even if temperature is the major driver of the diversity of ectotherms, we expect an increase in the influence of NPP across trophic levels not only for endotherms but also for ectotherms. Although few studies have investigated the influence of both temperature and NPP on multitrophic guilds of animals (e.g., Xu et al. 2018; Welti et al. 2020), no studies to date have evaluated whether the effects of temperature and NPP across trophic guilds are consistent among ectothermic and endothermic animals. Searching for similar patterns across the trophic levels of both endothermic and ectothermic animals might be helpful in explaining the influence of environmental drivers on biodiversity and energy flux and ultimately on ecosystem functioning (Turney and Buddle 2016). In this context, tropical elevational gradients are a great lab for testing the differences in the influence of temperature and NPP on biodiversity. Elevational gradients present a large variation in environmental factors on a relatively small spatial scale, wherein animal diversity is constrained by this environmental variation but not by dispersal limitations (Körner 2007; Sundqvist et al. 2013). Many ecological drivers are directly linked to elevation besides temperature, such as land area, atmospheric pressure, and clear-sky turbidity (Körner 2007). However, temperature and NPP are known as the main drivers for tropical elevational fauna (Peters et al 2016; Gebert et al. 2019). In addition, because tropical animals have evolved in a more climatically stable environment, they typically have narrow thermal tolerances (Janzen 1967). This results in a high species turnover in tropical mountains due to ecological filters imposed by temperature variation (Janzen 1967; Rahbek et al. 2019), which is advantageous for disentangling temperature from any other effects of drivers that are typically not strongly correlated to elevation, such as NPP. Understanding how temperature and NPP constrain species richness in both ectotherms and endotherms as well as across different trophic guilds is essential for predicting how biodiversity and ecosystem functioning will respond to a rapidly changing world. Here, we aimed to assess the relative role of temperature and NPP in multitrophic guilds of ectothermic (arthropods: ants, ground beetles, spiders, and harvestmen) and endothermic (large mammals) animals along a tropical elevational gradient. Although we expect that both temperature and NPP will positively influence both endotherms and ectotherms, we specifically tested whether the relative importance of NPP concerning temperature is higher for endotherms than for ectotherms. Additionally, regardless of the most important environmental driver for these patterns, we also tested whether there was an increase in the association of NPP with species richness across trophic levels for both ectotherms (arthropods) and endotherms (large mammals). Oecologia Materials and methods Study area We conducted our sampling at Itatiaia National Park (INP) in Southeast Brazil (between 22° 16′–22° 28′ S and 44° 34′–44° 42′ W). INP was established in 1937 and is the oldest national park in Brazil in the Atlantic rainforest biome. The park is part of the Itatiaia Massif, which is within the Mantiqueira Mountain Range. The protected area starts at 600 m of elevation and peaks at 2878 m. Several vegetation types occur along INP elevational gradient varying from forests to grasslands, including lower montane forest (at 0–500 m.a.s.l.), montane forest (500–1500 m.a.s.l.), and upper montane forest (1500–2000 m.a.s.l.). High-altitude grasslands (2000–2500 m.a.s.l.) and natural remnants of upper montane forest (Safford 1999) comprise the highest part of the INP. As vegetation structure, in terms of forest or grassland, affects the observed diversity patterns along elevational gradients (Lasmar et al. 2020), we sampled both arthropods and mammals only in forest habitats. Sampling of arthropods and classification into trophic guilds We sampled four groups of arthropods to represent ectothermic animals: ants (Insecta: Hymenoptera: Formicidae), ground beetles (Insecta: Coleoptera: Carabidae), spiders (Arachnida: Araneae), and harvestmen (Arachnida: Opiliones). All samplings took place during the rainy season between March and April 2015, and we chose eight elevation belts (at 600, 848, 1134, 1515, 1810, 2000, 2200, and 2457 m.a.s.l; Fig. 1a) spatially separated by at least 1.4 km. To sample arthropods, we set a 200 m transect in each elevation belt containing ten sampling points each, which were 20 m apart from each other. At each sampling point, we set a 1.5 m × 1.5 m grid of four unbaited pitfall traps composed of one trap at each grid vertex (Fig. 1b). The pitfall traps consisted of recipients (11 cm in diameter and 11 cm in depth) filled with a solution of water, salt, and liquid soap. They operated for 48 h in the field, as this period is known to be enough to sample the majority of arthropod species on the ground (Engelbrecht 2013; Lasmar et al. 2017). We started the samplings instantly after pitfall installation, as we considered the digging-in effect (i.e., high captures in the early periods of pitfall operation) negligible (Lasmar et al. 2017). We also extracted 1 m2 of forest leaf litter leaf using Fisher’s (1998) “miniWinkler” (Fig. 1b). For the Winkler samples, arthropods were extracted for 72 h in the miniWinkler sacks. Combining both pitfall traps and Winkler samples allowed us to obtain the best representation of arthropod diversity (Sabu et al. 2011). We sorted the ants by genera following Baccaro et al.’s (2015) auxiliary keys. The ants in each genus were then sorted into morphospecies and identified at the species level whenever possible by comparison with identified species in the ant reference collection of the Laboratório de Ecologia de Formigas at the Universidade Federal de Lavras (UFLA), Lavras, Brazil. Ant species identification was confirmed by taxonomists at the Universidade Federal do Paraná (UFPR) in Curitiba, Brazil (Dr. Alexandre Casadei-Ferreira, Dr. Thiago R. da Silva, and Weslly Franco). Voucher specimens were deposited at UFLA and at the Entomological Collection Padre Jesus Santiago Moure of UFPR (DZUP) in Curitiba, Brazil. Carabid ground beetles were identified by auxiliary keys (Straneo and Ball 1989), and the morphospecies were identified to the lowest level possible. The specimens were deposited in the Entomological Collection of the Laboratório de Ecologia in the Departamento de Ciências Florestais at UFLA. For harvestmen and spiders, we considered only adult individuals. We sorted harvestmen based on the analysis of the specimen exomorphology using relevant taxonomic literature (Pinto-da-Rocha and Giribet 2007) and by comparison against specimen in the type material. The specimens were deposited in the Collection of Arachnida, Myriapoda, and Onychophora at Museu Nacional at Universidade Federal do Rio de Janeiro (MNRJ). Unfortunately, all the collected harvestmen were lost in a fire in 2018 along with the bulk of the arachnological collection of MNRJ (Kury et al. 2018). The spiders were identified and deposited at the Arachnological Collection of the Laboratório de Coleções Zoológicas at the Instituto Butantan. The spider specimens were determined using the key of Brescovit et al. (2002) and the auxiliary bibliography of the World Spider Catalog (2021). We classified arthropods into three trophic guilds according to their ecological role in the ecosystem following previous studies (see Basset et al. 2012; Kitching et al. 2020). Thus, arthropods were assigned to one of the following trophic guilds: (1) arthropod herbivores, which feed predominantly on plant resources or through symbiosis with fungi that grow in plant resources; (2) arthropod omnivores, which feed on a wide range of substances, such as plant resources, feces, and dead and living organisms; and (3) arthropod predators, which feed predominantly on other living animals. In short, herbivore and omnivore arthropods consisted solely of ants and arthropod predators of ants, ground beetles, harvestmen and spiders. Each arthropod species and its corresponding trophic guild are shown in Table S1 of the Online Resource. 13 Oecologia (a) (b) (c) Fig. 1 a Elevational gradient sampled in Itatiaia National Park in Southeast Brazil. Black stars represent the arthropod (ants, ground beetles, harvestmen, and spiders) sampling sites while black circles represent the large-mammal sampling sites. b Scheme of the arthro- pod sampling transects and sampling points using pitfall traps and Winkler extractor. c Scheme of the large mammals sampling with camera traps Sampling of large mammals and classification into trophic guilds elevation belts (871, 889, 953, 1100, 1175, 1258, 1316, 1422, 1472, 1672, 1813, 1875, 2000, 2246 m.a.s.l.; Fig. 1a), all in forest habitats that were spatially separated by at least 730 m. In each elevation belt, we installed one motion-activated camera trap (Bushnell HD, © Bushnell We sampled large mammals in the rainy season between January and April of 2014 and 2015. We chose 14 13 Oecologia Outdoor Products, California, USA) that remained in operation for 80 consecutive days with a total sample effort of 1,120 trap/nights (Fig. 1c). We installed camera traps in places commonly used by mammals based on the presence of animal signs and trails (Srbek-Araujo and Chiarello 2013). We used no bait to avoid artificial animal attraction and to maintain the premise of equal catchability. We programmed camera traps to take three photos in 30-s intervals when triggered by an animal. We identified large mammals through photographic records, and we considered as independent observations the records that presented a time lapse of > 1 h between them (SrbekAraujo and Chiarello 2013). We classified mammals into three trophic guilds based on their role in the ecosystem (see Srbek-Araujo and Kierulff 2016): (1) herbivores, (2) omnivores, and (3) predators. Each large-mammal species and its corresponding trophic guild are shown in Table S2 in Online Resource. Measuring temperature and net primary productivity To extract net primary productivity (NPP) for each elevation belt for elevation gradients of both arthropods and large mammals, we assessed the Land Processes Distributed Active Archive Center website from NASA (https:// lpdaac.usgs.gov) to extract MODIS TERRA and AQUA NPP (MOD17A3HGF and MYD17A3HGF, https:// lpdaac.usgs.gov/products/mod17a3hgfv006) at 500 m of resolution. As NPP is an annual product and we collected our samples in 2014 and 2015, we extracted NPP values for the year that we collected the samples and for the corresponding previous year because we always sampled at the beginning of the given year; therefore, our response variable might also be influenced by the NPP of the previous year. We calculated the average values between MODIS TERRA and AQUA NPP and for two years (i.e., the previous year and the year we collected the samples) to obtain one value for each elevation belt. To extract temperature data, we obtained the mean annual temperature from the CHELSA database (Karger et al. 2017) with a 1 km resolution. CHELSA mean annual temperature corresponds to average values from monthly temperature for the years 1979–2013. Mean annual temperature was highly correlated with elevation in both arthropod (Pearson r = − 0.99) and large-mammal gradients (Pearson r = − 0.98). This is in accord with Barry (2008), who states that it is well established that temperature decreases 0.6 °C per 100 m elevation gain. Both the NPP and temperature data were extracted with ArcGIS 10.1 software, and we ensured that all elevation belts were not in the same grid cell. Data analyses We checked the sample coverage for each of our response variables by estimating asymptotic species richness from frequency data (i.e., frequency of occurrence in sampling points per transect) using a nonparametric species richness estimator (Chao 1; Colwell and Coddington 1994; Gotelli and Colwell 2011). We estimated arthropod diversity through frequency data because an ant worker, as a eusocial organism, cannot be considered an individual (Gotelli et al. 2011), which led us to standardize this for all arthropods. We also estimated large-mammal diversity through frequency data since we only counted photographs that were registered on camera traps. We checked for correlation between observed species richness and Chao 1‐estimated diversity. If we found high correlations (r ≥ 0.7), we only performed the analyses with observed values of species richness. If those parameters were not strongly correlated (r < 0.7), we performed analyses with both observed species richness and Chao 1-estimated diversity. We verified whether observed species richness followed a linear or unimodal function in response to elevation for the total species richness of arthropods (n = 8) and large mammals (n = 14) as well as for each of the three trophic guilds of arthropods and mammals separately. To do so, we performed a theoretical approach following the second-order Akaike information criterion corrected for small samples (AICc— Burnham and Anderson 2002). To perform the model selection, we built, for each observed species richness (response variables), three models that presented elevation as a linear function, elevation2, to represent a unimodal function and the null model as explanatory variables. We used the function ‘aictab’ from the AICcmodavg package in R (Mazerolle 2020) to test which function best explained our response variables. The explanatory variables with the lowest AICc indicated the function (linear or unimodal) that best fit our response variable pattern along the elevational gradient. To test the influence of temperature and NPP on the species richness of ectotherms (all arthropods) and endotherms (all large mammals) as well as on the different trophic guilds, we first verified the correlation between our explanatory variables, NPP and temperature. For both gradients, NPP and temperature were not strongly correlated (arthropod gradient, Spearman r = 0.63; large-mammal gradient, Spearman r = 0.68). Thus, to test whether temperature was more important than NPP in driving the species richness of ectotherms and whether NPP was more important than temperature in driving the species richness of endotherms, we compared the responses of overall arthropod species richness (n = 8) and overall species richness of large mammals (n = 14) by constructing generalized linear models (GLMs) with temperature and NPP as explanatory variables. We simplified both models by performing a backward approach, consisting 13 Oecologia of removing nonsignificant explanatory variables from the models until only the significant variables were present in the model. To verify that the collinearity between temperature and NPP was not a problem for fitting and interpreting our model, we calculated the variance inflation factor (VIF) for each model using the vif function in the ‘car’ package in R (Fox and Weisberg 2019). If the VIF values were lower than 5, the collinearity did not require attention, but it did require attention for VIF values greater than 5. We used the “Quasipoisson” family for all our models, and we verified which of those significant variables better explained our response variables by assessing the 95% confidence intervals of their proportion of variation explained (R2) in the regression models through a resample from 1000 bootstrap samples using the ‘boot.ci()’ function from the ‘boot’ package (Canty and Ripley 2012). To test whether the influence of NPP on species richness increases across trophic levels in both ectotherms (arthropods) and endotherms (large mammals), we constructed GLMs for each trophic guild as a response variable and performed a backward approach with elevation and NPP as explanatory variables. We extracted the variance inflator factor (VIF) for each model. We also compared the 1000 bootstrapped R2 values from significant explanatory variables from GLMs of trophic guilds of arthropods and large mammals. When performing the GLMs, we assessed the residual to obtain the adequacy of the error distribution (Crawley 2002). We also performed variance partitioning analyses for all models (i.e., overall arthropods and large mammals, as well as for all trophic guilds), to verify the unique and shared proportion of variance explained by temperature and NPP. We performed this using the ‘var.part’ function in the ‘vegan’ package in R (Oksanen et al. 2020). We confirmed the adequacy of the error distribution family for each model as well as VIF values lower than 5 (see Table S5 in the Online Resource). We performed all analyses in R version 4.0.3 (R Development Core Team 2020). Results In the arthropod elevational gradient, we identified a total of 286 species from 969 ant occurrences, 322 ground beetles, 179 harvestmen (145 adults and 34 juveniles), and 1229 spiders (686 adults and 543 juveniles). Per trophic guild, we identified seven species of herbivores (seven ant species), 96 species of omnivores (96 ant species), and 183 predators (72 ant species, 16 harvestmen species, 23 ground beetle species, and 72 spider species). In the large-mammal gradient, we recorded 21 species from 383 occurrences. One of these species was Sus scrofa, a nonnative animal of the American continent that was introduced in the highest part of the 13 Mantiqueira Mountain Range; therefore, we removed S. scrofa from the analyses since all diversity patterns changed when including this introduced species (See Table S3 and S4 in Online Resource). We recorded five species of largemammal herbivores, six large-mammal omnivores and eight large-mammal predators. The list of all recorded species as well as their trophic guilds can be found in Table S1 and S2 in Online Resource. We found strong correlations between observed species richness and Chao 1-estimated diversity for all arthropods combined (r = 0.99), arthropod herbivores (r = 0.99), arthropod omnivores (r = 0.98), and arthropod predators (r = 0.99). We also found strong correlations between observed species richness and Chao 1-estimated diversity for total large mammals (r = 0.96), large-mammal herbivores (r = 1), large-mammal omnivores (r = 0.96) and large-mammal predators (r = 0.99). Therefore, we used only observed species richness for both arthropods and largemammal parameters for all analyses. Elevational gradient patterns Overall, arthropods, arthropod omnivores, and arthropod predators better followed a unimodal pattern along elevation, while arthropod herbivores linearly declined with elevation (Fig. 2a–d and Table S3 in Online Resource). We found that the unimodal function best explained the elevational gradient for species richness of overall large mammals, largemammal herbivores and omnivores (Fig. 2e–h and Table S3 in Online Resource). However, neither the linear nor unimodal function explained the elevational gradient for largemammal predators. Effects of temperature and net primary productivity on all arthropods and all large mammals We found that overall arthropod species richness was positively related to temperature and to NPP (Table 1 and Fig. 3a, b). On the other hand, overall species richness of large mammals was positively related to NPP but not to temperature (Table 1, Fig. 3c, d). As expected, our bootstrap R2 analyses shows that temperature had the greatest explanatory power for overall arthropod species richness, while NPP was the major ecological driver influencing the species richness of all large mammals (Fig. 4a). The variance partitioning showed that the majority of the overall explained variance was shared between temperature and NPP (Fig. S1a and b in the Online Resource). However, we also observed that the unique explained variance of temperature was higher for overall arthropods species richness (42% of the overall explained variance) in comparison to large mammals (31% of the overall explained variance). On the other hand, the shared contribution, which also accounts for NPP effects, was Oecologia 140 4 (a) 60 (b) Species richness 120 80 (c) 3 100 60 40 80 2 40 30 60 20 40 1 20 10 20 0 0 600 1000 1400 1800 2200 0 600 Elevation (m.a.s.l.) 14 1000 1400 1800 2200 0 600 4 1400 1800 2200 600 1000 Elevation (m.a.s.l.) Elevation (m.a.s.l.) (e) 1000 4 (f) 1400 1800 2200 Elevation (m.a.s.l.) 6 (g) 12 Species richness (d) 50 (h) 5 3 10 3 4 8 2 2 1 1 0 0 3 6 2 4 1 2 0 900 1200 1500 1800 2100 900 Elevation (m.a.s.l.) 1200 1500 1800 2100 0 900 Elevation (m.a.s.l.) 1200 1500 1800 2100 900 Elevation (m.a.s.l.) Fig. 2 Elevational species richness patterns in Itatiaia National Park, Southeast Brazil. Graphs show the linear or unimodal relationship of elevation and the observed species richness for a all arthropods, b arthropod herbivores, c arthropod omnivores, d arthropod predators, e all large mammals, f large-mammal herbivores, g large-mammal 1200 1500 1800 2100 Elevation (m.a.s.l.) omnivores, and h large-mammal predators. Points show the observed species richness according to elevation, and the lines are model predictions of the lowest AICc function (i.e., linear or unimodal). In graph h, the null model best represents the relationship between large-mammal predators and elevation Table 1 Generalized linear models (GLMs) for two elevational gradients of arthropods and large mammals in Itatiaia National Park in Southeast Brazil Dependent Independent df Pseudo R2 Standardized rc 2.5% CI 97.5% CI F p All arthropods Net primary productivity Temperature Net primary productivity Temperature Net primary productivity Temperature Net primary productivity Temperature Net primary productivity Temperature Net primary productivity Temperature Net primary productivity Temperature Net primary productivity Temperature 5 0.88 6 0.77 5 0.84 6 0.75 12 0.30 11 0.53 12 0.36 – – 0.326 0.325 Excluded 1.393 0.409 0.532 0.411 Excluded 0.449 Excluded 0.295 0.631 Excluded 0.367 Excluded Excluded 0.086 0.143 – 0.733 − 0.025 0.238 0.208 – 0.061 – − 0.578 0.095 – 0.114 – – 0.581 0.506 – 2.230 0.923 0.829 0.637 – 1.015 – 1.645 1.245 – 0.634 – – 25.83 12.17 3.21 20.65 14.12 12.49 17.46 5.20 5.51 1.96 9.57 5.45 3.01 8.31 2.84 < 0.01 0.004 0.017 0.132 0.003 0.013 0.016 0.005 0.071 0.036 0.188 0.010 0.039 0.111 0.013 0.117 0.930 Arthropod herbivores Arthropod omnivores Arthropod predators All large mammals Large-mammal herbivores Large-mammal omnivores Large-mammal predators The GLMs were separately built for the observed species richness of all arthropods, large mammals and their trophic guilds as the dependent variables, and we used temperature and NPP as independent variables. We simplified all the GLMs by performing a backward approach. We reported the GLM degrees of freedom (d.f.) and Pseudo R2 for the final models and the standardized regression coefficients (rc) and their confidence intervals (CI) only for independent variables that were present in the final models. We also reported statistical test values (F) and p values for the excluded independent variables from the final (excluded) model and for the remaining independent variables in the final model. Bold values indicate significant relationships (p ≤ 0.05) 13 Oecologia All large mammals All arthropods 140 (a) 120 100 100 80 80 60 60 40 40 20 0 Species richness 120 12 14 16 18 14 (b) 12 Species richness 140 14 (c) 12 10 10 8 8 6 6 4 4 20 2 2 0 0 20 12000 Mean annual temperature (°C) 14000 16000 0 12 18000 14 Arthropod herbivores 4 (e) 3 3 2 2 1 1 0 14 16 18 20 Mean annual temperature (°C) 12000 14000 3 3 2 2 1 1 16000 Net primary producitivity (g_C/m²/year) 60 (i) 14 40 30 30 20 20 10 10 0 18000 (h) 18 12000 14000 16000 18000 Net primary producitivity (g_C/m²/year) Large-mammal omnivores 4 (j) 4 (k) 14 16 18 20 Mean annual temperature (°C) 3 3 2 2 1 1 0 0 12 12000 14000 16000 (l) 0 12 18000 14 (m) 60 60 40 40 20 20 0 14 16 18 20 Mean annual temperature (°C) 12000 14000 12000 14000 16000 18000 Net primary producitivity (g_C/m²/year) 6 (0) 5 5 4 4 3 3 2 2 1 1 0 0 12 18 Large-mammal predators 6 (n) Species richness 80 16 Mean annual temperature (°C) Net primary producitivity (g_C/m²/year) Arthropod predators Species richness 16000 50 40 80 16 Mean annual temperature (°C) Species richness Species richness 50 14000 0 12 18000 Arthropod omnivores 60 12000 Net primary producitivity (g_C/m²/year) 4 (g) 0 0 12 18 Large-mammal herbivores 4 (f) Species richness Species richness 4 16 Mean annual temperature (°C) Net primary producitivity (g_C/m²/year) (d) 16000 18000 Net primary producitivity (g_C/m²/year) (p) 0 12 14 16 18 Mean annual temperature (°C) 12000 14000 16000 18000 Net primary producitivity (g_C/m²/year) Fig. 3 Relationship between temperature and net primary productivity with arthropods and large mammals along a tropical elevational gradient at Itatiaia National Park, Southeast Brazil. Graphs show the comparison of the relationship of temperature and net primary productivity with the observed species richness of all arthropods (a, b), all large mammals (c, d), arthropod herbivores (e, f), large-mammal herbivores (g, h), arthropod omnivores (i, j), large-mammal omni- vores (k, l), arthropod predators (m, n), and large-mammal predators (o, p). Note that in c, d, k, and l, we only presented graphs without the nonnative species S. scrofa. Points show the observed species richness according to a given value of temperature or net primary productivity, and the lines are model predictions of significant relationships (e.g., p ≤ 0.05) higher for species richness of large mammals (69% of the overall explained variance) than for arthropods (57% of the overall explained variance). This was in accordance with our bootstraps R 2 analyses and indicated that the effect of temperature is strongly associated with arthropods while the NPP effect was more strongly associated with large mammals. 13 Oecologia Fig. 4 Comparison of the proportion of variance explained (R2) by the significant effects of temperature and net primary productivity (NPP) from the GLM regressions. Comparisons of temperature and NPP effects on observed species richness were made for a all arthropods and for all largemammal species richness and for b arthropods and large mammals separated in trophic guilds (herbivores, omnivores, and predators). White boxplots represent temperature effects, while blue boxplots represent NPP effects. Vertical dashed lines are bootstrapped confidence intervals constructed by 1000 bootstrap samples with replacement, and black dots represent data outliers. Notch areas on boxplots mark the 95% confidence intervals of the median value (shown as black horizontal lines) Temperature effect 1.0 Net primary productivity effect (a) 0.8 0.6 R² 0.4 0.2 0.0 All arthropods 1.0 All arthropods All large mammals (b) 0.8 0.6 R² 0.4 0.2 0.0 Arthropod Arthropod Arthropod Arthropod L. mammal L. mammal L. mammal herbivores omnivores omnivores predators herbivores herbivores omnivores Effects of temperature and net primary productivity on multitrophic guilds of arthropods and large mammals When verifying the influence of temperature and NPP across trophic guilds, we found that the species richness of arthropod herbivores was positively related to temperature, but we detected no relationship with NPP (Table 1, Fig. 3e, f). For the species richness of arthropod omnivores, we found a positive relationship with both temperature and NPP (Table 1, Fig. 3i, j). For the species richness of arthropod predators, we found a positive relationship with NPP, but we detected no relationship with temperature (Table 1, Fig. 3m, n). For large-mammal herbivores, we found a positive relationship with both temperature and NPP, while for largemammal omnivores, we found a positive relationship with temperature, but we detected no relationship with NPP (Table 1, Fig. 3g, h, k, l). We found that large-mammal predators were not related to temperature or NPP (Table 1, Fig. 3o, p). Comparing the bootstrap R2 values from the GLMs, we found an increase in NPP associated with an increase in the trophic guild level of arthropods, as we expected (Fig. 4b). However, distinctly from what we expected, NPP association with species richness did not increase across largemammal trophic guilds, which indicated a more complex pattern (Fig. 4b). The variance partitioning analysis also showed that increasing arthropods trophic levels increased NPP association (Fig S1c, e, g in Online Resource). From arthropod herbivores to predators, the shared explained variance increased from 0 to 65% while the unique explained variance of NPP increased from 8 to 24% in relation to the overall explained variance. However, we observed no evidence that increasing trophic levels also increased NPP association for large-mammal trophic guilds (Fig. S1d, f, h in Online Resource). Therefore, bootstrap analysis was also in accordance with variance partition analysis, showing 13 Oecologia that increasing trophic levels of ectotherms (arthropods) also increased NPP association while endotherms (large mammals) presented complex patterns about NPP association across trophic levels. Discussion As expected, temperature was strongly associated with arthropod diversity, while NPP was associated with largemammal diversity. However, distinctly from the expected results, we found no consistency of stronger NPP association with species richness of higher trophic levels between arthropods and large mammals. The importance of NPP association increased across trophic levels of arthropods, while for large-mammal trophic guilds, we did not observe clear patterns. Our results indicate that differences in animal thermal physiology might affect the way diversity is constrained by environmental temperature and NPP, as reviewed in Buckley et al. (2012). In addition, our findings also indicate that thermal physiology probably affects the energy flux in the food web, which we think is a step forward in understanding the differences in the responses of ectotherms and endotherms to the environment. Effects of temperature and net primary productivity on all arthropods and all large mammals Our results are in accordance with studies that have shown a decline in arthropod species richness with increasing elevation due to temperature decrease (e.g., Peters et al. 2016) and those that have reported a positive association of NPP with large mammals in an African elevational gradient (e.g., Gebert et al. 2019). NPP was also positively associated with arthropods, indicating that resource availability is important for increasing arthropod species richness, possibly by reducing species competition and local extinction (Evans et al. 2005). However, as arthropod diversity was more strongly associated with temperature, we suggest that temperature limits arthropod species richness through physiological constraints and that it seems to impose greater constraints on species richness of arthropods than NPP. As arthropods are dependent on external temperatures, low temperatures decrease their foraging activity (Cerdá et al. 1998; Lasmar et al. 2021), which influences their population sizes at high elevations, resulting in a decrease in arthropod species richness (Sanders et al. 2007; Lasmar et al. 2020). On the other hand, endotherms are able to harvest food resources even in cold temperatures at the expense of a high amount of food resources to maintain their vital functions (Buckley et al. 2012). Therefore, resource availability is a key driver that limits the number of coexisting species of large mammals in the tropics (Gebert et al. 2019). Considering our limited 13 sample size (arthropods n = 8; large mammals n = 14), we should be cautious about generalizing our results, and more studies in other tropical regions could be helpful in consolidating our findings and inferences. This is especially true when considering that we evaluated only one group of endotherms (i.e., only large mammals), and, therefore, the causes of the stronger positive association of NPP on large mammals could be due to both their endothermy and to other specific traits of large mammals. However, the stronger association between NPP concerning temperature on other endotherms was also supported by Buckley et al. (2012) as well as in other case studies (e.g., birds, Hawkins et al. 2003b; Acharya et al. 2011; bats and rodents, Owen 1990; McCain et al. 2018). In this sense, differences in thermal physiology between ectotherms and endotherms may be responsible for the way arthropods and large mammals interact with or are constrained by the environmental temperature and NPP. Effects of temperature and net primary productivity on multitrophic guilds of arthropods and large mammals For arthropod trophic guilds, temperature is strongly associated with the diversity of lower trophic levels, while NPP is strongly associated with the diversity of higher trophic levels. In addition to the fact that accelerated metabolic rates at high temperatures increase arthropod foraging activity (Cerdá et al. 1998; Lasmar et al. 2021) and, consequently, energy use from the ecosystem (Twomey et al. 2012), primary consumers are extremely sensitive to changes in plant phenology, which are temperature dependent (Thackeray et al. 2016). However, the greater association of arthropod herbivores with temperature in this study could also occur because most of the species are represented by fungus-garden ants, which are extremely influenced by temperature (Bollazzi and Roces 2002). Thus, because the species richness of arthropod herbivores is underrepresented in our data, the generalization of our results to other important arthropod herbivores (e.g., Hemiptera, Lepidoptera, Orthoptera) needs to be tested. Nevertheless, we consider our representation of arthropod omnivores and predators to be highly satisfactory and enough to confirm our prediction that the association of NPP with species richness increases across arthropod trophic levels. This is in accordance with previous studies that found that the abundance and species richness of arthropod predators are more constrained by resource availability than primary consumers (e.g., Kaspari 2001; Xu et al. 2018). Abundance across trophic levels accumulates along the NPP gradient (Oksanen et al. 1981; Kaspari 2001), which makes predators more sensitive to the energy input in the ecosystem than organisms at lower trophic levels (Turney and Buddle 2016; Brose et al. 2017). The higher the trophic level, the more energy is needed to maintain viable population sizes Oecologia and then high species richness (Turney and Buddle 2016). Our results are partially following Mayr et al. (2020), who found that Hymenoptera at the low trophic level are mainly influenced by temperature while higher trophic levels are influenced by both temperature and resource availability. Our limited sampling size may have caused the lack of statistical support for the influence of temperature on the species richness of predators in this study, even though it is clear that the association of species richness with NPP increases from lower to higher trophic levels. Therefore, even though temperature is one of the major constraints on arthropod diversity (Gillooly et al. 2001; Lessard et al. 2011), we suggest that the sensitivity to energy input in the ecosystem increases across arthropod trophic levels, which positively affects species richness. However, this seems not to be the rule among ectotherms and endotherms since the trophic guilds of large mammals followed distinct patterns. The species richness of large-mammal trophic guilds presented a complex pattern concerning temperature and NPP. NPP was only associated with the species richness of large-mammal herbivores. As endotherms may use the energy from the ecosystem more independently of climate than ectotherms (Buckley et al. 2012), the availability of resources could constrain primary consumers’ diversity of large mammals (Beck 2006). However, this effect of the energy input in the ecosystem on species richness seems not to be strong when trophic levels increase. This is because although NPP influenced large-mammal herbivores, the effect of temperature seemed to be stronger and the diversity of large-mammal omnivores was only associated with temperature, while no variables were associated with largemammal predators. NPP may present stronger effects on communities with more phylogenetic diversity than on communities with narrowly defined clades, because the latter typically use a smaller fraction of NPP in comparison to a phylogenetically broader community (Peters et al. 2020). Considering that we had a narrower number of clades distributed among large-mammal trophic guilds, the correlation of NPP with their true food resources might be less strong. This may explain why we found an effect of NPP on overall large mammals (i.e., a more broad phylogenetic community), a weak effect of NPP for large-mammal herbivores and no effect of NPP for other trophic groups. On the other hand, temperature is known to also increase trophic interactions and resource assimilation (Fine 2015; Peters et al. 2020). For large-mammal omnivores, we should consider that these animals feed on plant resources, such as fruits and seeds, and on other animals, including many arthropods (e.g., PinedaMunoz and Alroy 2014; Galetti et al. 2015), all extremely influenced by temperature (Peters et al. 2016). Therefore, we suggest resource availability as an important driver of omnivore species richness. However, this may be not simply dependent on the energy input provided by producers but also on other factors that influence the food resources obtained from both producers and other prey at lower trophic levels. In conclusion, it is possible that trophic interactions, which are positively influenced by temperature (Fine 2015) could outweigh the effects of environmental NPP on species richness at low and intermediate trophic levels (e.g., large-mammal herbivores and omnivores). The dependence of trophic interactions could also explain the lack of association of large-mammal predators with environmental temperature and NPP. Due to the lack of a relationship between large-mammal predators and our variables, we have two not mutually exclusive hypotheses. First, large-mammal predators occur naturally at low densities in the environment and forage for large distances (Smith et al. 2017); in addition, human-related actions reduce or dislocate individuals from their original areas (Rogala et al. 2011). Despite our study area being a protected area, it historically suffered from deforestation, poaching and criminal fires. Therefore, human-related variables that influence carnivores may not have been considered in this study, and we may need biogeographical scales to better understand patterns of large-mammal predators. Second, the prey of predator animals could also obscure their association with temperature and NPP. This is because largemammal predators are strongly influenced by their prey abundance (Smith et al. 2017), in which at least their species richness in this study was associated with both temperature and NPP. In addition, Sus scrofa is known to influence both the spatial and temporal patterns of native mammals (Hegel et al. 2019; Galetti et al. 2015), such as the prey of predators, and may also be a new source of prey (Hegel et al. 2018). Both could affect the presence of large mammals along the elevational gradient. Conclusion We report here that the species richness of ectothermic arthropods is more associated with temperature, while that of endothermic large mammals is associated with productivity (NPP) along a tropical elevational gradient. Most of the comparisons reporting that the relative importance of temperature and NPP varies between ectotherms and endotherms were assessed between vertebrates (reviewed in Buckley et al. 2012). Here, we provide evidence that those differences in the role of NPP and temperature in driving species richness also apply when comparing ectothermic invertebrates with endothermic vertebrates. Additionally, we found no consistency regarding the role of temperature and NPP on the species richness structure across multitrophic guilds of ectotherms and endotherms. The association of species richness with NPP increased across trophic levels for arthropods; however, although NPP is important 13 Oecologia for large-mammal herbivores, its influence does not seem to accumulate across trophic levels. Large-mammal trophic guilds seem to be more indirectly dependent on temperature via other trophic interactions than simply on environmental NPP. Based on that, it is possible that besides global temperature change might strongly affect arthropods, it might also affect large-mammal trophic guilds in the tropics. Here, we also suggest that thermal physiological differences might also interfere with energy use and flux in food webs, which differently reflect species richness across trophic levels between ectotherms and endotherms. Several studies propose several mechanisms explaining how species abundance, biomass and richness are distributed across trophic levels in the food web (e.g., Oksanen et al. 1981; Turney and Buddle 2016; Wang and Brose 2018). Therefore, it would be useful for future studies to consider thermal physiological differences in animals when assessing the distinct facets of trophic guilds, such as the pyramid shape of abundance, biomass, and species richness. Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00442-021-05011-9. Acknowledgements This work was funded by Fundação de Amparo à Pesquisa de Minas Gerais (FAPEMIG, Grant PPM 00243/14), and Tropical Conservation Act/Fundo Brasileiro da Biodiversidade (TFCA/ FUNBIO). We thank Marcell K. Peters and an anonymous referee for their helpful comments. We are thankful to the staff of National Itatiaia Park, especially to Leo Nascimento and Marcelo Motta, who permitted the sampling in the park area. We are also indebted to Maria Regina de Souza, Tobias R. Silva, Luiza Santiago, Edson Guilherme de Souza, Daniel Q. Domingos, Ernesto O. Canedo–Júnior, Graziele Santiago, Luana Zurlo, Fernando H. Puertas, Thamíris C. K. de Abreu, and José Cristiano for their help with logistics and fieldwork. Thanks to Mariana Rabelo, Ícaro Carvalho, and Felipe Lopes for helping with laboratory work. We are grateful to the staff at the Laboratório de Sistemática de Formigas da Universidade Federal do Paraná for confirming ant identification, especially to Alexandre Ferreira. CJL and CAN received a postdoctoral fellowship from PNPD/CAPES. ACMQ received a postdoctoral fellowship from CEMIG—Companhia Energética de Minas Gerais S.A. (P&D 611—Descomissionamento da PCH Pandeiros: Uma experiência inédita na América do Sul). RMF and ADB were supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq grant 302462/2016–3 and CNPq grant 303903/2019-8). This study was part of Chaim J. Lasmar’s MSc. thesis at the Universidade Federal de Lavras that was supported by Coordenação de Aperfeiçoamento Pessoal (CAPES, Finance code: 001). Author contribution statement CJL CR, and CRR conceived and designed the experiments. CJL, CR, MMGI, GPA, and GBN conducted the field and lab work; CJL, CR, RMF, LNZ, LV, and ADB identified the collected species. CJL, ACMQ, and CAN analysed the data. CJL wrote the manuscript; all other authors provided editorial advice. Funding This study was funded by Fundação de Amparo à Pesquisa de Minas Gerais (FAPEMIG, Grant PPM 00243/14), Tropical Forest Conservation Act/Fundo Brasileiro Para a Diversidade (TFCA)/FUNBIO). Data availability All data produced from this study are provided in the Electronic Supplementary Material of this manuscript. 13 Declarations Conflict of interest The authors declare that they have no conflict of interest. Ethical approval All applicable institutional and/or national guidelines for the care and use of animals were followed. All animal experiments were approved by Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio), reference number 46564-1. References Acharya BK, Sanders NJ, Vijayan L, Chettri B (2011) Elevational gradients in bird diversity in the Eastern Himalaya: an evaluation of distribution patterns and their underlying mechanisms. PLoS ONE 6(12):e29097. https://doi.org/10.1371/journal.pone.0029097 Allen AP, Brown JH, Gillooly JF (2002) Global biodiversity, biochemical kinetics, and the energetic-equivalence rule. Science 297:1545–1548. https://doi.org/10.1126/science.1072380 Baccaro FB, Feitosa RM, Fernandez F, Fernandes IO, Izzo TJ, Souza JLP et al (2015) Guia para gêneros de formigas do Brasil. Editora INPA Manaus. https://doi.org/10.5281/zenodo.32912 Barry RG (2008) Mountain weather and climate, 3rd edn. Cambridge University Press, Cambridge Basset Y, Cizek L, Cuénoud P, Didham RK, Guilhaumon F, Missa O, Tishechkin AK et al (2012) Arthropod diversity in a tropical forest. Science 338:1481–1484. https://doi.org/10.1126/science. 1226727 Beck H (2006) A review of peccary–palm interactions and their ecological ramifications across the Neotropics. J Mammal 87:519–530 Binkenstein J, Klein AM, Assmann T, Buscot F, Erfmeier A, Ma K, Pietsch KA, Schmidt K, Scholten T, Wubet T, Bruelheide H, Schuldt A, Staab M (2018) Multi-trophic guilds respond differently to changing elevation in a subtropical forest. Ecography 41:1013–1023. https://doi.org/10.1111/ecog.03086 Birkhofer K, Wolters V (2012) The global relationship between climate, net primary production and the diet of spiders. Glob Ecol Biogeogr 21:100–108. https://doi.org/10.1111/j.1466-8238.2011. 00654.x Bollazzi M, Roces F (2002) Thermal preference for fungus culturing and brood location by workers of the thatching grass-cutting ant Acromyrmex heyeri. Insectes Soc 49:153–157. https://doi.org/10. 1007/s00040-002-8295-x Brescovit AD, Bonaldo AB, Bertani R, Rheims CA (2002) Araneae. In: Adis J (ed) Amazonian Arachnida and Myriapoda. Pensoft Publishers, Sofia-Moscow, pp 303–343 Brose U, Blanchard JL, Eklöf A, Galiana N, Hartvig M, Hirt MR, Kalinkat G et al (2017) Predicting the consequences of species loss using size-structured biodiversity approaches. Biol Rev 92:684–697. https://doi.org/10.1111/brv.12250 Brown JH (2014) Why are there so many species in the tropics? J Biogeogr 41:8–22. https://doi.org/10.1111/jbi.12228 Buckley LB, Hurlbert AH, Jetz W (2012) Broad-scale ecological implications of ectothermy and endothermy in changing environments. Glob Ecol Biogeogr 21:873–885. https://doi.org/10.1111/j.14668238.2011.00737.x Burnham KP, Anderson DR (2002) Model selection and inference: a practical information—theoretic approach, 2nd edn. Springer, New York Canty A, Ripley B (2012) Boot, Bootstrap R (S–Plus) Functions. R Package version1.3–4. https://CRAN.R-project.org/package=boot. Accessed 1 Feb 2020 Oecologia Cerdá X, Retana J, Cros S (1998) Critical thermal limits in Mediterranean ant species: trade-off between mortality risk and foraging performance. Funct Ecol 12:45–55. https://doi.org/10.1046/j. 1365-2435.1998.00160.x Colwell RK, Coddington JA (1994) Estimating terrestrial biodiversity through extrapolation. Philos Trans R Soc Lond B Biol Sci 345:101–118. https://doi.org/10.1098/rstb.1994.0091 Crawley MJ (2002) Statistical computing, an introduction to data analysis using S-Plus, 1st edn. Wiley, London Engelbrecht I (2013) Pitfall trapping for surveying trapdoor spiders: the importance of timing, conditions and effort. J Arachnol 41:133– 142. https://doi.org/10.1636/P12-57.1 Evans KL, Warren PH, Gaston KJ (2005) Species-energy relationships at the macroecological scale: a review of the mechanisms. Biol Rev 80:1–25. https://doi.org/10.1017/S1464793104006517 Fine PV (2015) Ecological and evolutionary drivers of geographic variation in species diversity. Annu Rev Ecol Evol Syst 46:369–392. https://doi.org/10.1146/annurev-ecolsys-112414-054102 Fox J, Weisberg S (2019) An R companion to applied regression, 3rd edn. Sage, Thousand Oaks. http://socialsciences.mcmaster.ca/jfox/ Books/Companion/. Accessed 01 Dec 2020 Galetti M, Camargo H, Siqueira T, Keuroghlian A, Donatti CI, Jorge MLS, Ribeiro MC (2015) Diet overlap and foraging activity between feral pigs and native peccaries in the Pantanal. PLoS ONE 10:1–10. https://doi.org/10.1371/journal.pone.0141459 Gebert F, Njovu HK, Treydte AC, Steffan-Dewenter I, Peters MK (2019) Primary productivity and habitat protection predict elevational species richness and community biomass of large mammals on Mt. Kilimanjaro. J Anim Ecol 88:1860–1872. https://doi.org/ 10.1111/1365-2656.13074 Gillooly JF, Brown JH, West GB, Savage VM, Charnov EL (2001) Effects of size and temperature on metabolic rate. Science 293:2248–2251. https://doi.org/10.1126/science.1061967 Gotelli NJ, Colwell RK (2011) Estimating species richness. In: Magurran AE, McGill BJ (eds) Frontiers in measuring biodiversity. Oxford University Press, New York, pp 39–54 Gotelli NJ, Ellison AM, Dunn RR, Sanders NJ (2011) Counting ants (Hymenoptera: Formicidae): biodiversity sampling and statistical analysis for myrmecologists. Myrmecol News 15:13–19 Haddad NM, Crutsinger GM, Gross K, Haarstad J, Knops JM, Tilman D (2009) Plant species loss decreases arthropod diversity and shifts trophic structure. Ecol Lett 12:1029–1039. https://doi. org/10.1111/j.1461-0248.2009.01356.x Hawkins BA, Field R, Cornell HV, Currie DJ, Guégan JF, Kaufman DM et al (2003a) Energy, water, and broad-scale geographic patterns of species richness. Ecology 84:3105–3117. https://doi.org/ 10.1890/03-8006 Hawkins BA, Porter EE, Felizola Diniz-Filho JA (2003b) Productivity and history as predictors of the latitudinal diversity gradient of terrestrial birds. Ecology 84:1608–1623. https://doi.org/10.1890/ 0012-9658(2003)084[1608:PAHAPO]2.0.CO;2 Hegel CGZ, Marini MÂ (2018) Large felids as predators of wild boars in the Atlantic forest: reconciling Verdade et al. and Rosa et al. Anim Conserv 21:363–364. https://doi.org/10.1111/acv.12402 Hegel CGZ, Santos LR, Marinho JR, Marini MÂ (2019) Is the wild pig the real “big bad wolf”? Negative effects of wild pig on Atlantic forest mammals. Biol Invas 21:3561–3574. https:// doi. org/ 10. 1007/s10530-019-02068-9 Hillebrand H (2004) On the generality of the latitudinal diversity gradient. Am Nat 163:192–211. https://doi.org/10.1086/381004 Janzen DH (1967) Why mountain passes are higher in the tropics. Am Nat 101:233–249. https://doi.org/10.1086/282487 Karger DN, Conrad O, Böhner J, Kawohl T, Kreft H, Soria-Auza RW, Zimmermann NE, Linder P, Kessler M (2017) Climatologies at high resolution for the Earth land surface areas. Sci Data 4:170122. https://doi.org/10.1038/sdata.2017.122 Kaspari M (2001) Taxonomic level, trophic biology, and the regulation of local abundance. Glob Ecol Biogeogr 10:229–244. https://doi. org/10.1046/j.1466-822X.2001.00214.x Kitching RL, Dahlsjö CAL, Eggleton P (2020) Invertebrates and the complexity of tropical ecosystems. Biotropica 52:207–214. https://doi.org/10.1111/btp.12768 Körner C (2007) The use of ‘altitude’ in ecological research. Trends Ecol Evol 22:569–574. https://doi.org/10.1016/j.tree.2007.09.006 Kury AB, Giupponi APL, Mendes AC (2018) Immolation of Museu Nacional, Rio de Janeiro—unforgettable fire and irreplaceable loss. J Arachnol 46:556–558 Lasmar CJ, Queiroz ACM, Rabello AM, Feitosa RM, Canedo-Júnior EO, Schmidt FA, Cuissi RG, Ribas CR (2017) Testing the effect of pitfall-trap installation on ant sampling. Insectes Soc 64:445–451. https://doi.org/10.1007/s00040-017-0558-7 Lasmar CJ, Ribas CR, Louzada J, Queiroz ACM, Feitosa RM, Imata MM, Alves GP, Nascimento GB, Neves FS, Domingos DQ (2020) Disentangling elevational and vegetational effects on ant diversity patterns. Acta Oecol 102:103489. https://doi.org/10.1016/j.actao. 2019.103489 Lasmar CJ, Bishop TR, Parr CL, Queiroz ACM, Schmidt FA, Ribas CR (2021) Geographical variation in ant foraging activity and resource use is driven by climate and net primary productivity. J Biogeogr 48:1448–1459. https://doi.org/10.1111/JBI.14089 Lessard JP, Sackett TE, Reynolds WN, Fowler DA, Sanders NJ (2011) Determinants of the detrital arthropod community structure: the effects of temperature and resources along an environmental gradient. Oikos 120:333–343. https://doi.org/10.1111/j.1600-0706. 2010.18772.x MacArthur RH (1972) Geographical ecology: patterns in the distribution of species. Princeton University Press, Princeton Mayr AV, Peters MK, Eardley CD, Renner ME, Röder J, SteffanDewenter I (2020) Climate and food resources shape species richness and trophic interactions of cavity-nesting Hymenoptera. J Biogeogr 47:854–865. https://doi.org/10.1111/jbi.13753 Mazerolle MJ (2020) AICcmodavg: model selection and multimodel inference based on (Q)AIC(c). R package version 1.30. http:// CRAN.R-project.org/package=AICcmodavg. Accessed 01 Dec 2020 McCain CM, King SRB, Szewczyk T, Beck J (2018) Small mammal species richness is directly linked to regional productivity, but decoupled from food resources, abundance, or habitat complexity. J Biogeogr 45:2533–2545. https://doi.org/10.1111/jbi.13432 Moorthi SD, Schmitt JA, Ryabov A, Tsakalakis I, Blasius B, Prelle L, Tiedemann M, Hodapp D (2016) Unifying ecological stoichiometry and metabolic theory to predict production and trophic transfer in a marine planktonic food web. Philos Trans R Soc B 371:20150270. https://doi.org/10.1098/rstb.2015.0270 Oksanen L, Fretwell SD, Arruda J, Niemela P (1981) Exploitation ecosystems in gradients of primary productivity. Am Nat 118:240– 261. https://doi.org/10.1086/283817 Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, Minchin P, O’Hara R, Simpson G, Solymos P, Stevens M, Szoecs E, Wagner H (2020) vegan: Community Ecology Package. R package version 2.5-7. https://CRAN.R-project.org/package= vegan. Accessed 01 Dec 2020 Owen JG (1990) Patterns of mammalian species richness in relation to temperature, productivity, and variance in elevation. J Mammal 71(1):1–13 Peters MK, Hemp A, Appelhans T, Behler C, Classen A, Detsch F et al (2016) Predictors of elevational biodiversity gradients change from single taxa to the multi-taxa community level. Nat Commun 7:13736. https://doi.org/10.1038/ncomms13736 Peters MK, Classen A, Müller J, Steffan-Dewenter I (2020) Increasing the phylogenetic coverage for understanding broad-scale 13 Oecologia diversity gradients. Oecologia 192:629–639. https://doi.org/10. 1007/s00442-020-04615-x Pineda-Munoz S, Alroy J (2014) Dietary characterization of terrestrial mammals. Proc R Soc B Biol Sci 281:20141173. https://doi.org/ 10.1098/rspb.2014.1173 Pinto-da-Rocha R, Giribet G (2007) Harvestmen: the biology of the Opiliones. Harvard University Press, Cambridge R Development Core Team (2020) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. http://www.rproject.org. Accessed 01 May 2021 Rahbek C, Borregaard MK, Colwell RK, Dalsgaard B, Holt BG, Morueta-Holme N, Nogues-Bravo D, Whittaker RJ, Fjeldså J (2019) Humboldt’s enigma: what causes global patterns of mountain biodiversity? Science 365:1108–1113. https://doi.org/10.1126/ science.aax0149 Rogala JK, Hebblewhite M, Whittington J, White CA, Coleshill J, Musiani M (2011) Human activity differentially redistributes large mammals in the Canadian Rockies National Parks. Ecol Soc 16(3):16 Sabu TK, Shiju RT, Vinod KV, Nithya S (2011) A comparison of the pitfall trap, Winkler extractor and Berlese funnel for sampling ground-dwelling arthropods in tropical montane cloud forests. J Insect Sci 11:2011–2028. https://doi.org/10.1673/031.011.0128 Safford HD (1999) Brazilian Páramos I. An introduction to the physical environment and vegetation of the campos de altitude. J Biogeogr 26:693–712. https://doi.org/10.1046/j.1365-2699.1999.00313.x Sanders NJ, Lessard JP, Fitzpatrick MC, Dunn RR (2007) Temperature, but not productivity or geometry, predicts elevational diversity gradients in ants across spatial grains. Glob Ecol Biogeogr 16:640–649. https://doi.org/10.1111/j.1466-8238.2007.00316.x Smith JE, Lehmann KD, Montgomery TM, Strauss ED, Holekamp KE (2017) Insights from long-term field studies of mammalian carnivores. J Mammal 98:631–641. https://doi.org/10.1093/jmamm al/gyw194 Srbek-Araujo AC, Chiarello AG (2013) Influence of camera-trap sampling design on mammal species capture rates and community structures in southeastern Brazil. Biota Neotrop 13:51–62. https:// doi.org/10.1590/S1676-06032013000200005 Srbek-Araujo AC, Kierulff MCM (2016) Mamíferos de médio e grande porte das florestas de Tabuleiro do norte do Espírito Santo: grupos funcionais e principais ameaças. In: Rolim SG, Menezes LFT, Srbek-Araujo AC (eds) Floresta Atlântica de Tabuleiro: diversidade e endemismos na Reserva Natural da Vale. Rupestre, Belo Horizonte, pp 469–479 Straneo SL, Ball GB (1989) Synopsis of the genera and subgenera of the Tribe Peleciini, and revision of the neotropical and oriental species (Coleoptera: Carabidae). Insecta Mundi 3(2):73–178 Sundqvist MK, Sanders NJ, Wardle DA (2013) Community and ecosystem responses to elevational gradients: processes, mechanisms, and insights for global change. Annu Rev Ecol Evol Syst 44:261– 280. https://doi.org/10.1146/annurev-ecolsys-110512-135750 Terborgh JW (2015) Toward a trophic theory of species diversity. Proc Natl Acad Sci USA 112:11415–11422. https://doi.org/10.1073/ pnas.1501070112 Thackeray SJ, Henrys PA, Hemming D, Bell JR, Botham MS, Burthe S et al (2016) Phenological sensitivity to climate across taxa and trophic levels. Nature 535:241–245. https://doi.org/10.1038/natur e18608 Turney S, Buddle CM (2016) Pyramids of species richness: the determinants and distribution of species diversity across trophic levels. Oikos 125:1224–1232. https://doi.org/10.1111/oik.03404 Twomey M, Brodte E, Jacob U, Brose U, Crowe TP, Emmerson MC (2012) Idiosyncratic species effects confound size-based predictions of responses to climate change. Philos Trans R Soc B 367:2971–2978. https://doi.org/10.1098/rstb.2012.0244 Voigt W, Perner J, Davis AJ, Eggers T, Schumacher J, Bährmann R, Fabian B, Heinrich W, Köhler G, Lichter D, Marstaller R, Sander FW (2003) Trophic levels are differentially sensitive to climate. Ecology 84:2444–2453. https://doi.org/10.1890/02-0266 Voigt W, Perner J, Hefin Jones T (2007) Using functional groups to investigate community response to environmental changes: two grassland case studies. Glob Change Biol 13:1710–1721. https:// doi.org/10.1111/j.1365-2486.2007.01398.x Wang S, Brose U (2018) Biodiversity and ecosystem functioning in food webs: the vertical diversity hypothesis. Ecol Lett 21:9–20. https://doi.org/10.1111/ele.12865 Welti EAR, Kuczynski L, Marske KA, Sanders NJ, de Beurs KM, Kaspari M (2020) Salty, mild, and low plant biomass grasslands increase top-heaviness of invertebrate trophic pyramids. Glob Ecol Biogeogr 29:1474–1485. https://doi.org/10.1111/geb.13119 World Spider Catalog (2021) World Spider Catalog. Version 22.0. Natural History Museum Bern. http://wsc.nmbe.ch. Accessed 20 May 2021. https://doi.org/10.24436/2 Wright DH (1983) Species-energy theory: an extension of species-area theory. Oikos 558(41):496–506. https://doi.org/10.2307/3544109 Xu G, Zhang S, Zhang Y, Ma K (2018) Environmental correlates underlying elevational richness, abundance, and biomass patterns of multi-feeding guilds in litter invertebrates across the treeline. Sci Total Environ 633:529–538. https://doi.org/10.1016/j.scito tenv.2018.03.212 Authors and Affiliations Chaim J. Lasmar1 · Clarissa Rosa2 · Antônio C. M. Queiroz1 · Cássio A. Nunes3 · Mayara M. G. Imata1 · Guilherme P. Alves1 · Gabriela B. Nascimento1 · Ludson N. Ázara4 · Letícia Vieira5 · Júlio Louzada6 · Rodrigo M. Feitosa7 · Antonio D. Brescovit8 · Marcelo Passamani9 · Carla R. Ribas1 1 2 Programa de Pós-Graduação em Ecologia Aplicada, Departamento de Ecologia e Conservação, Instituto de Ciências Naturais, Laboratório de Ecologia de Formigas, Universidade Federal de Lavras, PO Box 3037, Lavras, MG 37200-900, Brazil Coordenação de Biodiversidade, Instituto Nacional de Pesquisas da Amazônia, Manaus, Amazonas 69067-375, Brazil 13 3 Programa de Pós-Graduação em Ecologia Aplicada, Departamento de Ecologia e Conservação, Universidade Federal de Lavras, Lavras, Brazil 4 Laboratório de Aracnologia, Departamento de Invertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista, São Cristóvão, 20, Rio de Janeiro, RJ 940-040, Brazil Oecologia 5 Laboratório de Ecologia Florestal, Departamento de Ciências Florestais, Universidade Federal de Lavras, Lavras, MG 37200-000, Brazil 6 Laboratório de Ecologia de Invertebrados, Departamento de Ecologia e Conservação, Instituto de Ciências Naturais, Universidade Federal de Lavras, Lavras, MG 37200-000, Brazil 7 Departamento de Zoologia, Universidade Federal do Paraná, CP 19020, Curitiba, PR 81531-980, Brazil 8 Laboratório de Coleções Zoológicas, Instituto Butantan, São Paulo, SP 05503-900, Brazil 9 Laboratório de Ecologia e Conservação de Mamíferos, Departamento de Ecologia e Conservação, Instituto de Ciências Naturais, Universidade Federal de Lavras, Lavras, MG 37200-000, Brazil 13