Aim: Foraging activity is critical for animal survival. Comprehending how ecological drivers influence foraging behavior would benefit our understanding of the link between animals and ecological processes. Here, we evaluated the influence of ecological drivers on ant foraging activity and relative resource use.

Location: Six Brazilian biomes: Amazon, Atlantic rainforest, Caatinga, Cerrado, Pampa and Pantanal.

Taxon: Formicidae.

Methods: We assessed ant foraging activity and resource use by sampling across 60 sites. We placed baited tubes that contained one of five liquid resources (sugar, lipids, amino acid, sodium and distilled water). We used model selection to assess the influence of ecological drivers (temperature, precipitation, temperature seasonality and net primary productivity) on ant foraging activity and relative resource use.

Results: Foraging activity was higher in wetter, more productive and less thermally seasonal environments. The relative use of amino acids increased at higher temperatures while the relative use of lipids decreased. The relative use of sugar increased in drier and less productive environments with high temperature seasonality while the relative use of amino acid and sodium decreased in those environments. The relative use of lipids was complex: increasing with increasing temperature seasonality and decreasing with increasing precipitation. Further, the relative use of sodium was greater where the foraging activity was high.

Main Conclusions: We demonstrate how ecological drivers are correlated to ant foraging activity and resource use in the field across large spatial scales. The search for resources encompasses different interactions involving ants with abiotic and biotic components in the ecosystem. Thus, we suggest that changes in climate and NPP, which influence the intensity and the way that ants search for resources, will result in changes in ant-mediated ecological processes.

Sampling of foraging ants

In each of the six biomes, we installed 10 transects as sampling units (60 transects in total) separated by at least 1 km, except for two in the Amazon biome and two in the Pampa biome where they were separated by 800 m due to spatial constraints. Each transect was 750 m long with 25 sampling points separated by 30 m. At each sampling point, we provided one of five liquid food resources in the epigaeic strata. The food resources were placed in 50-mL Fisher Scientific polypropylene centrifuge tubes with a 5 cm cotton ball containing 10 ml of the following solutions in distilled water: 1% sodium (NaCl), 20% sugar (CHO, made with sucrose), 20% amino acids (AA, made with unflavored whey protein isolate), lipids (100% extra virgin olive oil, without water), and distilled water as a control. Similar liquid resources have successfully been used in previous studies (e.g., Fowler, Lessard, and Sanders 2014; Peters et al. 2014; Tiede et al. 2017). We placed the baited tubes horizontally on the ground. Hence, along each transect, each liquid food resource was repeated five times in the same sequence along each transect (following the order: control, carbohydrates, lipids, amino acid and sodium). Consequently, there were five pseudoreplicates of each of the five resource types per transect (5 pseudoreplicates × 5 resource types = 25 sampling points per transect), giving a total of 250 sampling points per biome (25 sampling points x 10 transects) and 1500 sampling points across the study (250 x 6 biomes).

For all biomes, baiting was restricted to sunny periods, and never in rain or on totally cloudy days. We began placement of the baited tubes at 7:00 am at each site for all biomes except in the sites of Caatinga biome, where we delayed placement until 1:00 pm because light rainfall during the morning may have reduced ant activity in this period. By restricting the sampling to one period (mornings or afternoon) in each study area, we avoided large variation in temperature. We left all tubes open for three hours, after which we capped the tubes to collect the ants inside. A tube was classed as visited when there was at least one ant individual inside.

Ecological drivers

Data for climate and NPP were obtained for 60 1-km2 grid cells, which each contained one sampled transect. We obtained data for temperature, precipitation and climate seasonality from the WorldClim 2 database (Fick & Hijmans, 2017), which represents average climate between 1970 and 2000. For each transect, we extracted the WorldClim variables mean annual temperature (C°), annual precipitation (mm), temperature seasonality (standard deviation of annual mean temperature), and precipitation seasonality (coefficient of variation of annual precipitation). For the months in which we sampled ants in each biome, we also extracted the data for monthly mean temperature (°C), and monthly precipitation (mm; Table 1). We obtained net primary productivity (NPP) from MODIS, using annual NPP from 2000-2015 (MOD17) from the NASA Earth Observation System repository at the University of Montana (www.ntsg.umt.edu/), which has been improved by correcting for cloud-contaminated pixels and uses a model that considers the difference between gross primary productivity and autotrophic respiration (Zhao and Running, 2010).

We compared the WordClim estimates of climate to nearest climatic station (NCS) weather data from 1970 to the present (See Appendix 1). We did this to check whether (1) our sampling periods experienced extreme weather compared to the historical average and, (2) to assess whether WorldClim reliably predicted NCS estimates of local conditions. The NCS data showed that the time periods in which we sampled we not extreme relative to the historical record (Fig. S1.1 in Appendix 1), and that the WorldClim data closely matched the estimates of temperature and rainfall for all sites (Fig S1.2 in Appendix 1). Consequently, we opt to use WorldClim data in our analyses as a good representation of both local weather during sampling and long term climatic trends and because WorldClim provides better spatial cover of the sampling sites.

Ant species composition can be found in the supplementary material of the Journal of Biogeography.