Print PageSagebrush rangelands in the western U.S. are being altered by non-native plants such as cheatgrass (Bromus tectorum). These changes (e.g., altered fire regimes) carry broad ecological and social implications, including wildlife habitat loss. Restoration of natives holds promise for abating this loss, yet effective initiatives remain elusive. Here, we examined potential impacts of seed-harvesting ants on the seed pool contributing to plant establishment at two sites in the Great Basin. We studied ant colonies in cheatgrass-degraded sites, reference sites, and prescribed burns to assess how management treatment and B. tectorum density (high vs. low) influence ant foraging.
Results/Conclusions
At Onaqui, central Utah, seeds were the most common item collected by ants, and were largely comprised of non-native species (e.g., 86% of collected seeds were non-native, n = 29 nests in 2007). Ceratocephala testiculata, Alyssum desertorum, and B. tectorum were the most common seeds returned, in addition to native bunchgrasses such as Elymus elymoides and Poa secunda. Forage items generally mirrored soil surface seed pool availability. Nests with high B. tectorum density had greater proportions of non-native seeds available and collected . Ants dramatically increased foraging following fire. For example, during Achnatherum hymenoides seed-removal trials in May 2007, ants detected (21 vs. 6) and emptied (10 vs. 0) more seed dishes in burn versus control conditions. Further, following the burn, ants collected more native seeds relative to the control. Differences were still evident, though less pronounced, after three years; in addition, native seed collection increased as the plant community reestablished. Results at Owyhee, northern Nevada, were different; litter was the most common item collected, and native seeds (P. secunda and E. elymoides) were more common among forage items (and in the seed pool) than non-natives. Nests with high B. tectorum density had greater proportions of B. tectorum seeds available and collected. Ants detected (2 vs. 10) and emptied (0 vs. 4) fewer seed dishes in B. tectorum-invaded than uninvaded areas. Our findings suggest: 1) ants collected and moved native and non-native seeds, with the relative proportions of each influenced by site, seed availability, and B. tectorum density; 2) nests with high B. tectorum density were less proficient at locating seed “patches,” likely because they focused on abundant seed resources at the time of sampling; and 3) ant foraging behavior appeared to increase strikingly following fire, presumably due to decreased seed resources. These results have implications for current post-fire reseeding practices.
