Blog        Research        Costa Rica        Curriculum Vitae

From a poster presentation at the Entomological Society of America annual meeting Knoxville, 2012.

Density and Dispersion of Cavity Dwelling Ant Species in Nuts of Eastern US Forest Floors

Doug Booher 1, Joe A. MacGown 2, and Richard M. Duffield 3, Stephen P. Hubbell 1

1 University of California, Los Angeles, Los Angeles, CA, 2 Mississippi Entomological Museum, Mississippi State, MS, 3 Howard University, Washington, DC


Here we report on nut-nesting ant communities in Georgia, Maryland, and Mississippi. We (1) compare species diversity, ant abundance, and nut occupancy rates among the three sites, and (2) report the spatial dispersion of ant colonies in nuts in relation to housing opportunities and nest-site choice. We also (3) test the ability of artificial nest cavities (“Mobile Artificial Ant Pods”, MAAPs) to attract ant colonies, a method for sampling the ant fauna in litter.


Nut Collections: (Plate 1) We collected 6240 nuts of walnut (Juglans nigra), pecan (Carya illinoinensis), hickory (Carya glabra, C. ovata, C. myristiciformis, and C. tomentosa), and oak (Quercus alba, Q. nigra, and Q. rubra) on the forest floor at 64 tree sites. We collected only nuts that could be occupied, i.e., those with entrances suitable for ants. We collected 13 to 375 nuts in individual bags per location (Tree site).

Spatial Dispersion of Ants at Walnut Tree: Booher and Duffield mapped, collected, and labeled all possible nuts suitable for ant nests in a circular area of 10 m radius out from the base of a walnut tree. We recorded nut location and the entrance type. We replaced each collected nut with a flagged MAAP.

Testing the MAAP artificial traps (Plate 2): We put out a total of 16 sets of MAAPs in GA (4 sets at Sandy Creek Nature Center, Athens), MD (1 site each in Frederick County and Rockville), and MS (6 sites, two in Noxubee National Wildlife Refuge and 4 in Tombigbee National Forest). Each set had 50 MAAPs spaced 1 m apart in a 5 x 10 array, We collected the MAAPs between 6 and 10 weeks after placement and also collected 50 nuts with entrances next to the array to compare trap success to nuts (Fig. 6).

Statistics: We analyzed data using JMP, R, and Basic, as well as two custom programs, BiodiversityR12, and EstimateS11. We compared the rank abundance of species across states using a Wilcoxon test.

Photos: Joe MacGown took photos (Plates 1,2, and 4) using a Leica Z16 Macroscope, Leica DFC 495 digital camera, and Leica Application Suite V 4.1.0 with Montage Module.


Summary of 6240 nuts of 9 nut-bearing tree species from 64 sites in Maryland, Georgia, and Mississippi.

Figure 1. 36 ant species nest occupied 11.6±8.7% (site mean±SD) of nuts over all tree sites. Percent of occupied nuts did not vary significantly between states (GA=12.7±11.1%, MD=11.8±8.5%, MS=10.5±7.8%) (one way ANOVA p≤.8501). The range of occupancy was 0-42% of nuts by ants at a given tree site. Ant species richness was similar at state level (GA=20, MD=18, MS=16). Despite similar species richness paired states shared only 21% to 29% of ant species. All 3 states shared only 10.8% of the 36 nut-nesting ant species. The 10 most abundant ants differed in two-state comparisons (GA-MD and GA-MS) p=0.001. In our study we found no correlations between Fisher’s alpha (diversity) and sample size, tree species, or nut-nesting frequencies. In Rockville, MD, we found nests in walnuts of social parasites *V. nipponica with short wing V. emeryi queens, and long wing V. emeryi with short wing V. emeryi queens. Long wing V. emeryi and V. nipponica are native to Japan13 (Plate 4. G,H).

Figure 2. The rarefaction curve (red) generated by actual sampling method predicts a higher initial accumulation than if all nuts from all states were randomly sampled in sets of 100 nuts (black).

Figure 3. In the best-fit (cubic) polynomial, diversity increases with % nut occupancy up to 15% occupancy (R2=0.306, p≤0.002.)


From all (n=328) empty nuts with entrances within a 10m radius from the base of a walnut tree.

Figure 4. This site contained the highest number of species (n=10) and two invasive species, the Red Imported Fire ant (Solenopsis invicta) and the Asian Needle Ant (Pachycondyla chinensis). Ants occupied (10.1%) of sampled nuts.

Figure 5. Ants aggregate. The two lines represents separate regressions of average nearest neighbors from n1 to n30 for occupied to occupied nuts (blue, bottom line), and a nut chosen at random to its nearest neighbor (from n1 to n30, perm=100 times). The bars above and below are p=0.05 error bars showing the significance of non-overlap between paired nuts.


Preliminary results of the MAAPs versus nuts for occupancy by ants.

Figure 6. MAAPs attract more ants in areas where more ants use nuts. (Prob>t 0.0418*). As with nuts, number of species also increases with increased MAAP colonization (R=0.667, prob>F=0.0001*)


The lack of shared species between states, high numbers of single species, and the continued rate of accumulation when all nuts were sampled suggest many ants that uncommonly nest in nuts do so only opportunistically. However, nuts may be a vital resource to the most common or site abundant ants found in nuts. In our collections we found some cavity-dwelling specialists occur commonly (such as T. curvispinosus and Strumigenys rostrata, Plate 4: E,F), other rare species at specific sites in abundance (i.e. Nylanderia trageri, Vollenhovia emeryi, and Temnothorax tuscaloosae, Plate 4: A, G, C), and many species occur infrequently in low abundances.

Ants that nest in nuts have potential as models testing hypotheses of community organization in structured habitats. Colonies in nuts are easy to find, map, and manipulate and nuts are spatially distributed critical resources. The aggregation of ants (Fig. 5) predicts environmental factors and not competition organize the dispersion of ants, but this site had an average occupied densities close to the total mean and competition may influence dispersal and colony founding at resource rich aggregated areas within sites. Using MAAPs and nuts to manipulate densities through the addition and removal of both colonies and nesting opportunities promises to enhance understanding of community organization.


This research was supported by Mississippi Agricultural and Forestry Experiment Station State Project MIS-311080, the USDA-ARS Areawide Management of Imported Fire Ant Project (Richard L. Brown, P.I.), and Patricia Adair Gowaty (UCLA). Special cooperation has been provided by the Noxubee NWR, The Tombigbee NF, Randy Smith and the Sandy Creek Nature Center, Michael Oliveri for MAAP development support, Cecil Smith and Rick Hoebeke at The UGA Collection of Arthropods, donated supplies, and the Georgia Dept. of Natural Resources. Thanks to Audrey Sheridan (Mississippi State University) for assistance in collecting nuts.


1. MacGown, J.A. Marginalia Insecta. 2006. 1(1): p. 1-12. 2. Duffield, R.M. & G.D. Alpert. Psyche (Cambridge), 2011. 3. Foitzik, S. & J. Heinze. Behavioral Ecology, 1998. 9(4): p. 367-375. 4. Talbot, M. Ecology, 1957. 38(3): p. 449-456. 5. Backus, V.L. and J.M. Herbers. Northeastern Naturalist, 2009. 16(1): p. 113-124. 6. Greenslade, P.J.M. 1971. 8(2): p. 323-&. 7. Scharf, I. et al. BMC Ecology, 2011. 11: p. 9-Article No.: 9. 8. Pratt, S.C. and N.E. Pierce. Animal Behaviour, 2001. 62: p. 281-287. 9. Cao, T.T. and A. Dornhaus. Biology Letters, 2008. 4(6): p. 613-615. 10. Herbers, J.M. & V. Banschbach. Psyche (Cambridge), 1995. 102(1-2): p. 13-17. 11.Colwell, R.K.(2009) estimateS: 12.Kindt R and Coe R. 2005. Tree diversity analysis.Nairobi: World Agroforestry 13.Booher, D.B. & Duffield, R.M. (2013) U.S. Caste Variation in Vollenhovia emeryi and North American record of Vollenhovia nipponica In Maryland. Unpublished manuscript.

Blog        Research        Costa Rica        Curriculum Vitae

thanks to Alex Wild for letting me use his Dinoponera photo to make my blog logo