10 likes | 143 Views
Interspecific spatial patterns support indirect facilitation of harvester ants by kangaroo rats. Andrew J. Edelman Dept. of Biology, University of New Mexico, Albuquerque, New Mexico 87131 (andrewe@unm.edu). Background Information
E N D
Interspecific spatial patterns support indirect facilitation of harvester ants by kangaroo rats Andrew J. Edelman Dept. of Biology, University of New Mexico, Albuquerque, New Mexico 87131 (andrewe@unm.edu) • Background Information • Heteromyid rodents (e.g. kangaroo rats) are hypothesized to indirectly facilitate granivorous ants in the Chihuahuan Desert by influencing ant food resources. • Heteromyid rodents forage selectively on large-seeded winter annuals, whereas granivorous ants primarily forage on small-seeded winter annuals. • Long-term removal experiments have shown that Heteromyid rodents reduce the abundance of large-seeded winter annuals, which competitively releases small-seeded winter annuals. • However, increases in small-seeded winter annual abundance does not lead to large increases in granivorous ant abundance as predicted. • I am examining interspecific spatial patterns of kangaroo rats and harvester ants for evidence of indirect facilitation effects. • Study Organisms • These species are the largest and most dominant granivores in rodent and ant communities of the Chihuahuan Desert. • Banner-tailed kangaroo rat (Dipodomys spectabilis) • Builds large mounds (4-m diameter) that contain burrows and seed caches. One adult lives at each active mound. • Rough harvester ant (Pogonomyrmex rugosus) • Builds large underground colonies with cleared surface disc (1-m diameter). Each colony has 1 queen and several hundred workers. • Are spatial associations influenced by recruitment of ant colonies? • No • Kangaroo Rat Mounds vs. Young Ant Colonies • Research Questions • Do kangaroo rat mounds and harvester ant colonies exhibit positive spatial association as predicted by the indirect facilitation hypothesis? • Are spatial associations influenced by recruitment or mortality patterns of ant colonies? • Do fitted spatial models support the indirect facilitation hypothesis? • Do fitted spatial models support the indirect facilitation hypothesis? • Yes, the best fit model includes a positive spatial interaction between species at scales ≤ 10 m. ↓ large-seeded winter annuals Table 1. Summary of multi-type Strauss hard-core models fitted to locations of kangaroo rat mounds and older ant colonies. The full model was a significantly better fit than the null model (P < 0.01). ↑ ant abundance ↑ small-seeded winter annuals • Methods • Study area: 8.5-ha grassland site at the Sevilleta National Wildlife Refuge, New Mexico. • All occupied mounds (n = 48) and discs (n = 209) were mapped to sub-meter accuracy on the study area during Summer 2007 using a GPS unit. The study area was resurveyed during Summer 2008 and new colonies (n =166) were mapped. • Ant colonies were checked for activity and diameter of disc was measured (index of colony age). • Occupancy of kangaroo rat mounds was monitored monthly since March 2005 by live-trapping and marking individuals. • All spatial analyses and model fitting were performed using the R package spatstat. • Mean intensity of opposing species sites (transformed bivariate Ripley’s K with edge correction) was calculated for a range of distances. Test significance was constructed using a Monte Carlo procedure (1000 simulations). Rejection limits (P < 0.05) based on spatial independence were used to determined at what distances the observed intensity differed from expected. • Multi-type Strauss hard-core models were fitted to spatial locations of kangaroo rat mounds and older ant colonies. A Monte Carlo test (100 simulations) was used to determine the best fit model. Fig. 3. Spatial association of kangaroo rat mounds and young ant colonies (n = 53, disc < 64 cm) during 2007. Kangaroo Rat Mounds vs. New Ant Colonies • Discussion • Kangaroo rat mounds and harvester ant discs were positively associated at fine scales (< 12 m) as predicted by the indirect facilitation hypothesis. • Recruitment of harvester ant colonies was spatially random with respect to mounds. • Positive spatial association between mounds and older colonies may indicate that survivorship of colonies near mounds is higher than away from mounds. • The best fit spatial model includes a positive spatial interaction between kangaroo rat mounds and ant colonies supporting the indirect facilitation hypothesis. • Kangaroo rats may indirectly facilitate harvester ants by increasing the seed abundance around their mounds. • Colonies near mounds have shorter foraging times and greater seed availability. Fig. 4. Spatial association of kangaroo rat mounds and new ant colonies (n = 166) during 2008. Results Do kangaroo rat mounds and harvester ant colonies exhibit positive spatial association? Yes, at scales < 12 m Kangaroo Rat Mounds vs. Ant Colonies • Are spatial associations influenced by ant colony mortality. • Yes, at scales < 12 m • Kangaroo Rat Mounds vs. Older Ant Colonies • Future Research • Monitor survival and recruitment of harvester ant colonies in 2009. Acknowledgments Funding for this project was provided by grants from Sevilleta LTER, NSF, ASM,UNM RPT, UNM SRAC, and UNM BGSA. I thank J. Brown, M. Friggens, S. Johnson, A. Kodric-Brown, F. Smith, E. Tuttle, and the personnel of the Sevilleta LTER and NWR for their assistance. Fig. 1. Study area with mapped locations of kangaroo rat mounds (○) and harvester ant colonies (∆) Fig. 2. Spatial association of kangaroo rat mounds (n = 48) and harvester ant colonies (n = 209) during 2007. Fig. 5. Spatial association of kangaroo rat mounds and older ant colonies (n = 156, disc > 64 cm) during 2007.