Foraging strategies and resource partitioning of desert harvester ants Ali Brunschwiler1, Amelia Maurer2, John Parsons3, Sage Radecki2, Joe Scullin4, Selena Vengco2 1University of California, Davis, 2University of California, Santa Cruz, 3University of California, Santa Barbara, 4University of California, San Diego ABSTRACT Resource limitations force desert species to forage efficiently and partition resources. Ant species have different foraging strategies, which may act as a method of niche partitioning. We aimed to measure differences in foraging behavior between colonies of Messor and Pogonomyrmex. Using an experimental treatment of seeds, we measured collection rate and resource preference between high and low density resources between the two species. We found significant differences in multiple aspects of foraging behavior, including overall abundance of seeds collected, proportion of different resources collected, and the effect of time on resource preference. These differences suggest a mechanism for resource partitioning between Messor and Pogonomyrmex. Additionally, the role of these two species in seed dispersal may be influenced by their foraging strategies. INTRODUCTION competition by partitioning resources (Hutchinson 1978, Begon et. al 1986, Deserts are among the most Ricklefs 1990). abiotically stressful ecosystems due to Sedentary ants form colonies in fixed low rates of precipitation and extreme locations and rarely move, limiting their temperatures. As a result, natural foraging abilities to a small area just selection has favored species that adapt around their nest. Since limiting to these extreme conditions by resources may impose strong selective optimizing energy efficiency (Davidson pressures in deserts, foraging behavior 1977). Optimal foraging theory predicts and behavioral cohesion of a colony that animals will maximize the amount may reflect their effort to maximize the of energy gained when foraging while difference between resources gained minimizing the amount of time and and energy lost while foraging energy lost. Additionally, species with (Bernstein 1975). In areas of higher similar niche requirements can reduce abundance, sedentary ants have been energy losses associated with observed utilizing a solitary foraging CEC Research | https://doi.org/10.21973/N3R36F Fall 2017 1/6 strategy where scouts leave the nest in can coexist they must be able to a certain direction until food is found. resource partition through foraging Once the food is returned to the nest, strategy. Our research allowed us to the ant will return directly to the place assess the resource dispersion where it last found food and begin preference of both Messor and searching for more food. This limits Pogonomyrmex. competition between foragers of the same colony, and maximizes the total METHODS surface area covered by a single colony 2.1 Study Area while foraging. In areas where resources are scarce, sedentary ants are typically We conducted our study at the observed utilizing a group foraging Sweeney Granite Mountains Desert strategy where scouts forage in Research Center in the Mojave Desert of different directions and recruit more California (34.7813° N, 115.6538° W). foragers once a food source is found. The reserve hosts a wide number of This maximizes their gathering potential granivorous ant species, and is in areas of scarce resources by putting composed of mixed cactus and yucca the colony’s efforts towards high scrub habitats. We observed ant density food resources. colonies located at Norris Camp and Two common ants found in arid Cottonwood Basin, locations that were regions of the southwestern United similar in habitat and abiotic conditions States are Messor sp. and including temperature and vegetative Pogonomyrmex sp. (both hereafter features. Our sample sites were at least referred to by Messor and 5 m apart to avoid inter-colony Pogonomyrmex, respectively). Both interactions, and we only surveyed genera are harvester ants although actively foraging colonies. We Messor tends to utilize the group considered a colony to be active if there foraging strategy while Pogonomyrmex were more than 10 ants outside a radius generally uses the solitary foraging of 0.5 m from the colony. We surveyed strategy (Rissing and Pollock 1984). In 12 Pogonomyrmex and 11 Messor order to quantify the differences colonies, 23 in total. between these two foraging techniques 2.2 Sampling Design we examined ant responses to high We conducted our field sampling over density (clumped) food sources vs. low the course of five days, November 4–8, density (dispersed) food sources. 2017. At each colony, we baited ants In our study, we examined how these with 8 g of bulgur seeds. Our two two ant genera utilized high vs. low treatments included seeds set out at density resources. By laying out equal high (clumped) and low (dispersed) quantities of clumped and dispersed densities. For the high density seeds, we were able to compare seed treatments, we set out four 1.0 gram intake and proportion of each resource clumps of seeds 4 cm in diameter and type collected between the two species 0.5 meters from the colony entrance at over time. If Messor and Pogonomyrmex CEC Research | https://doi.org/10.21973/N3R36F Fall 2017 2/6 each of the four cardinal directions. For We excluded the last fifteen minutes the low density (dispersed) treatments, of our observation period, because we we sprinkled 4.0 g of seeds evenly observed that the number of seeds collected per interval began to decline after 45 minutes, indicating that seed resources may have been exhausted. Thus, seeds may have been collected based on availability rather than preference. To test for differences in handling time between species we used a two- tailed t-test. We defined handling time as the time between an ant first picking Figure 1. Experimental layout. Large red circles up and first returning a seed to the represent 1.0 g clumped seed treatments and colony. We did not distinguish between blue dotted line represents 4.0 g of seeds evenly clumped and dispersed resources for dispersed 0.5 m from colony entrance. handling time. In order to determine if there were differences in the number of around the colony entrance at a 0.5 m total seeds collected by species per five radius (Figure 1). In order to minute time interval, we performed a differentiate between clumped and two-tailed t-test. A two-tailed t-test was dispersed seeds in the field, we dyed also used to test for differences in the seeds in different treatments different proportion of clumped resources that TM colors using food coloring (Watkins each species collected over the whole Assorted Food Coloring). One color of time period. We used a ANCOVA testing seeds was used for the four high density the effects of ant genus on proportion clumps, and a different color for the of clumped seeds taken with time dispersed arcs. To avoid potentially interval as a covariate. confounding effects of different colors of dye on experimental treatments, we RESULTS randomized colors of clumped and We found no significant difference in dispersed seeds each sampling day. handling time between species (n = 207, At each colony we observed ant p = 0.467). On average, Messor activity for 60 minutes, beginning when collected significantly more seeds per all seeds were set out. We divided the five minute time interval than observation period into five minute Pogonomyrmex (n = 207, p < 0.0001). intervals to observe changes in seed Messor collected over twice as many collection over time. We recorded the seeds per interval than Pogonomyrmex color of every bulgur seed carried into (Figure 2). Additionally, 31.3% of the the colony in order to distinguish which seeds that Messor collected were from treatment the ant collected it from. clumped resources, whereas 2.3 Statistical Analysis Pogonomyrmex collected 21.6% from clumped resources (n = 207, p < 0.0134 ; CEC Research | https://doi.org/10.21973/N3R36F Fall 2017 3/6 Figure 4). The proportion of clumped seeds taken by ants increased significantly over time (n = 207, p < 0.003; Figure 4). Additionally, there was Figure 4. Proportion of clumped resources per time interval per species. Black line represents Messor, red line represents Pogonomyrmex. a marginally significant interaction between the effects of species and time on proportion clumped seeds collected Figure 2. Total seeds per time interval by (n = 207, p < 0.056 ; Figure 4). Over species. Mean number of seeds collected per time, the proportion of clumped seeds interval ± se between ant species. taken by Messor increased by a 4% per interval, while the proportion taken by Pogonomyrmex did not increase over time. DISCUSSION We determined that Messor and Pogonomyrmex exhibit differences in many aspects of their foraging strategies, including harvesting rate, resource preference, and shifts in foraging behavior over time. Despite the similarity of individual efficiency between these two species, Messor Figure 3. Proportion of clumped resources by colonies as a whole were able to collect ant species. Mean proportion of clumped resources. seeds at twice the rate of Pogonomyrmex colonies. This suggests that differentiation between the two species is occurring at the level of the colony, not at the level of the individual forager. Because foraging efficiency does not differ between individuals of Messor and Pogonomyrmex, colony size and activity levels may be driving the CEC Research | https://doi.org/10.21973/N3R36F Fall 2017 4/6 differences in collection rates we in open areas (Nelson and Chew 1977). observed. Messor may be specialized to foraging in Furthermore, Messor colonies focused areas of higher seed density, while a higher proportion of their foraging Pogonomyrmex may be better adapted effort on clumped seeds than to open areas with low seed density due Pogonomyrmex colonies. This suggests to their lower preference for clumped that Messor is better at exploiting high- seeds. This is supported by the fact that density resources than Pogonomyrmex, Messor has a higher elevation range perhaps by use of a group-foraging than Pogonomyrmex and is therefore strategy.