Search

1. Harvester ants are major seed predators in arid environments. However, given that many harvester ants are partly omnivorous and therefore potentially attracted to the elaiosomes of myrmecochorous seeds, it is unclear if these ants act as predators or dispersers when removing myrmecochorous seeds. 2. We describe the outcomes of interactions between the harvester ant, Pogonomyrmex naegelii, and myrmecochorous plant, Microstachys serrulata, in a Brazilian savanna. We: (i) evaluated the role of elaiosome in seed removal by P. naegelii; (ii) investigated the fate and viability of removed seeds; (iii) tested if soils associated with P. naegelii nests are nutrient‐enriched; (iv) compared seedling survival; and (v) the density of seedling and adult M. serrulata nearby to P. naegelii nests compared with those away from these nests (i.e. controls). 3. Rates of removal of M. serrulata seeds were two‐fold higher with elaiosomes than without. The ant attractant oleic acid was the dominant fatty acid in elaiosomes, but it was absent from seeds. Removed seeds are taken into nests, and Tetrazolium tests indicated that 95% of seeds remain viable. Soils associated with P. naegelii nests were not nutrient‐enriched, and seedling survival was similar nearby to P. naegelii nests compared with control areas. However, densities of both seedling and adult M. serrulata were higher nearby to P. naegelii nests than in control areas. 4. Our findings show that P. naegelii switches its role from seed predator for most plant species to be the dominant seed disperser for M. serrulata, playing a key role in the distribution of adult plants. [ABSTRACT FROM AUTHOR]

1. Harvester ants perform important ecological functions, such as seed predation and redistribution of nutrients in the soil, through a complex of ecological interactions. Most studies are static descriptions of network structure, while their temporal organisations and the factors that modulate it have been neglected. 2. This study describes the temporal organisation and the influence of climatic variables (e.g. temperature and rainfall) on network structure between a harvester ant species and its food resources (seeds and invertebrates) in the Brazilian savanna over a 1‐year period. In addition, considering the ant–seed networks, the relationship between the size of seeds collected by ants and the distance travelled was examined, as well as how this relationship behaves over time. 3. For all network types (ant–seed, ant–invertebrate, and combined) there was great structural variation throughout the year. Among climatic variables, only rainfall influenced the properties of ant–invertebrate networks, presenting a positive influence on diversity of interactions and a negative influence on network specialisation. Finally, this study shows that harvester ants tend to collect larger seeds in the vicinity of the nest (< 5 m) and smaller seeds at greater distances, and this relationship appears to be constant over time. 4. This study highlights that ecological networks are useful tools with which to understand the temporal organisation of ecological functions. In addition, climatic changes, mainly in the rainfall regime, may interfere directly in the availability of resources and ultimately in the interactions between harvester ants and their food resources. [ABSTRACT FROM AUTHOR]

• Management variables known to affect dung beetles did not affect their functions. • Pasture age had a weak negative relationship with soil excavation by dung beetles. • Beetle community attributes did not explain dung removal or soil excavation. • Functions respond weakly, or not all, to subtle habitat changes within a land cover. • Studies need to determine how to use beetles as indicators of subtle habitat changes. Studies comparing different land covers clearly show that land-use change commonly affects animal communities and the ecological functions they play in ecosystems. However, we lack a good understanding of the effects that more subtle changes, those occurring within a land cover type, can have on ecological functions of animals, and if these are also mediated by shifts in community attributes. We studied two important ecological functions of dung beetles in tropical planted pastures with exotic African grasses. Due to differences in management, pastures varied in grass cover, tree density, pasture age, cattle density, and in the application frequency of veterinary medical products. We sampled beetles and their functions in 24 sites to address two questions: (1) Do management variables known to strongly affect beetle communities also affect their functions?; (2) Are the dung beetle-mediated ecological functions of dung removal and soil excavation correlated, and are they influenced by community attributes? We found no influence of management variables on dung removal, while soil excavation was negatively affected by pasture age, but only weakly. Contrary to our expectations, no community attribute was correlated with either dung removal or soil excavation, but both functions were strongly and positively correlated with each other. Our results indicate that responses of dung beetle communities and their ecological functions to subtle changes within a land cover type, are decoupled, idiosyncratic and context-dependent, making predictions and generalizations difficult. We believe that there is a need of more studies investigating the effects of fine-scale changes within habitats to better understand the consequences of management decisions on ecological processes mediated by animals. [ABSTRACT FROM AUTHOR]

• Dungbeetles and ants showed different responsesto pastoral land management. • We link these responses to different ecological traits between and among these taxa. • Dungbeetles and leafcutter ants were mainly affected by parasiticide use. • Ants and small-sized roller beetles responded to vegetation structure. • Beetles best indicate the use of parasiticides and ants the variation in plant cover. Biological responses to a disturbance can vary among taxa, which challenges the use of bioindicators for representing biodiversity responses more broadly. Linking ecological traits to disturbance response helps clarify what different bioindicator groups are actually indicating, providing a mechanistic basis for predicting the responses of other taxa. Dung beetles and ants are ecologically contrasting but widely used bioindicator taxa. Here we link variation in dung beetle assemblages to body size and dung-feeding behavior, and variation in ant communities to habitat preferences, foraging behaviour and competitive dynamics, to clarify what dung-beetles and ants are indicating in terms of biodiversity responses to pastoral land management in a Brazilian savanna landscape. Dung beetles and ants were sampled at 24 pastoral sites that varied in grass cover, tree density, pasture age, cattle density, and frequency of parasiticide use. Community metrics (abundance, richness, composition, functional group composition) of both dung beetles and ants varied markedly with pastoral land management, but this variation was not correlated, or only weakly, between taxa. For dung beetles, abundance, biomass and richness were all best predicted by frequency of parasiticide use, whereas species composition was best predicted by cattle density. Tree density was the most important variable for explaining ant abundance (both total and of several functional groups) and was also significantly related to ant species composition. Dung beetles and ants provided different indications of the impacts of pastoral management on biodiversity. However, by linking responses to ecological traits we were able to clarify what they were actually indicating. Leaf-cutter ants showed a similar response to dung beetles; collectively they are indicating an impact of parasiticide use on dung-feeders. The abundance of small-sized roller beetles was negatively related to grass cover; along with ants they are indicating the importance of vegetation structure. Such a linking of disturbance responses to ecological traits allows for more-robust predictions of likely responses of other faunal groups. We believe that such an approach provides a substantially improved foundation for the use of bioindicators in land management. [ABSTRACT FROM AUTHOR]