Limited ant co-occurrence and defensive mutualism in Acacia plants in a West African savanna

Our understanding of the role of fire and effect of ant species composition, beyond their diversity and abundance, on the effectiveness of mutualism defense is limited. Most of our knowledge of ant-plant defense in tropical Africa is biased toward East African savannas which have richer soil, higher primary productivity and a more diverse arthropods and mammal community than West African savannas. We assessed the diversity of ant species associated with Acacia species in the Pendjari Biosphere Reserve in the Dahomey Gap, and their impacts on elephant damage. Elephant damage, ant diversity and abundance were measured in stands of five Acacia species. Eleven ant species were identified in the Acacia stands. The composition of these ant communities varied across Acacia species. Pair of ant species co-occurred in only 2 % of sampled trees, suggesting a strong competitive exclusion. Within this annually burnt environment, ants were rare on small trees. The intensity of elephant-caused branch breaking did not vary between trees with ants and trees without ants, suggesting limited Acacia-ant mutualism. Such limited biotic defense may mask strong physical and chemical defense mechanisms of Acacia trees against elephant damage. Ant assemblages in West Africa, unlike those in the more productive East Africa, are particularly species-poor. However, there is a convergence between these two regions in low rate of ant co-occurrence which might indicate strong competitive exclusion. Our study suggests that such low ant species richness while limiting the efficacy of mutualism in controlling megaherbivore damage may mask a strong defense syndrome.

Methods
This data was collected in The Pendjari Biosphere Reserve which is the uppermost north-west Benin in the Sudanian region of the Dahomey gap (Adomou et al. 2006; Assédé et al. 2012) between the latitudes 10°30’ and 11°30’ N and the longitudes 00°50’ and 2°00’ E. We measured the diversity and abundance of ants once during the dry season after the annual vegetation fire in stands of Acacia sieberiana, Acacia seyal, Acacia gourmaensis, Acacia hockii and Acacia dudgeonii in the Pendjari Biosphere Reserve. Data on ants’ abundance were collected on Acacia trees sampled in 28 plots of 100 m x 100 m each. We sampled ants in eight plots for A. sieberiana sites and in five plots for each of the other Acacia species because A. sieberiana was more represented in the Pendjari Biosphere reserve. To evaluate the abundance and diversity of the ants, we used attractive traps consisting of 30 x 30 cm white ceramic tiles, each with a 4 cm spot of bait composed of honey mixed with canned tuna at its center. These attractive traps designed to evaluate the abundance and diversity of ants were each placed under the acacia trees and deployed for 30 minutes before the ants were counted and collected with a mouth aspirator. Ant counting was perfected in digital photographs of the samples (Dassou et al. 2015, 2016). Bait traps were placed under 10 randomly chosen acacia trees in each plot. To reinforce the collection of the diversity and abundance of ants, after collecting all the ant taxa in the bait trap we captured with an aspirator all ants that move on the acacia tree stem above the trap to a height of 2 m. These ant taxa collected on the tree stem were counted at the lab. In plots with less than 10 acacia trees, all individuals were sampled. All samples were obtained in the morning between 8:00 a.m and 12:00 p.m. Ants caught were preserved in 70% ethanol solution. Ant species collected in the acacia ecosystems were identified at the Entomological Museum of the International Institute for Tropical Agriculture, Benin Station.