Working Title: Invasions dynamics of W. auropunctata in Mexico and Puerto Rico
Biological invasions are major drivers of global change and pose a major challenge for the maintenance of biodiversity and human welfare (Vitousek 1990, Simberloff 2013). Invasive ants have successfully spread across the globe as is evident from their continual expansion across distinct biogeographical regions (McGlynn 1999) Invasive ants can reduce the abundance and richness of native ant species in regions with high densities of invasive ants (Holway 2002, Lach 2009). In invaded habitats, fewer native ants have been found to co-occur with some estimates showing 70 % reduction in native ants species richness (Porter 1990). Interspecific competition is a major factor structuring of ecological communities (Hölldobler 1990). Studies on invasive ant and competitive dynamics have primarily focused on interference and exploitative competition (Holway 2002). Interference competition (i.e. contest competition) involves direct aggressive encounters among individuals by hindering foraging activities or reproductive efforts. Exploitative competition (i.e. scramble competition) typically involves indirect encounters through competition for limited resources such as food and nesting sites. Rapid recruitment of workers and movement of nests to resources are examples of exploitative competition whose control allows colonies to outperform competitors (Parr 2009). The relative degree to which interference and exploitative strategies enable invasive ants to be successful is not well understood. Within native communities, interference and exploitative traits can promote species coexistence in ant communities (Adler 2007). The mechanistic explaination for this phenomena relies on the discovery-dominance trade-off describing a negative relationship between the ability of a species to discover a resource rapidly versus the ability of a species to recruit workers and dominate resources (Fellers 1987). Differences in discovery-dominance trade-offs can promote coexistence among many competing ant species (LEBRUN 2007). Invasive species are thought to break the discovery-dominance trade-off by excelling at both traits. For instance, in California the invasive Argentine ant Linepithema humile was found to excel at discovering and dominating resources, while native ants differed in their discovery and dominance abilities (Holway 1999). This abillity to excel at discovery and domination has been attributed to invasive success (Human 1996). However, we continue to lack information on whether other invasive species share similar competitive traits resources in both their native and introduced ranges. In one study, interactions between the invasive ant S. invicta and local ant community showed a discovery-dominance relationship in only 1 out of 3 sites within the native range of Brazil (Feener 2008). Recent work shows a negative relationship between discovery and dominance among four of the worst invasive ant species including L. humile, L. neglectus, P. megacephla, W. auropunctata, in a laboratory setting allowing all four species to coexist (Bertelsmeier 2015). Yet, these patterns will likely change for each invasive species when considering competitive interactions with native species under natural field conditions. To this end, we conducted a comparative study between W. auropunctata and competitors within the native range of Mexico and introduced range of Puerto Rico. W. auropuncata is considered to be one of the globe's worst invasive ant species due to negative associations with native ant communities (citation needed) . W. auropunctata is native to the Neotropics and has in recent decades spread to other tropical and subtropical regions of the world including Caribbean and Pacific Oceans, Australia, Florida, and more recently Israel (Wetterer 2013, Vonshak 2011). Within it's native range, W. auropunctata coexists with other dominant species in both forest and human dominated landscapes (Majer 1999, Armbrecht 2003)(ORIVEL 2009). However, in the introduced range W. auropunctata has been found to displace native ant species C(Clark 1982), with high reductions in species density reported along forest edges in New Caledonia (Breton 2003) and infested forested habitats in Gabon (Walker 2006). Laboratory experiments reveal that W. auropunctata does not initiate aggression and instead behaves as an insinuator species by foraging undercover alongside dominant species V(Vonshak 2011). Furthermore, W. auropunctata is slow at discovering and recruiting workers to resources (Bertelsmeier 2015a). This suggests that invasive species are not necessarily homogeneous as a group but may display different behavioral strategies that need further investigation. Our study compared the invasion dynamics of W. aurpunctata populations between their native and invasive ranges. We were interested in determining biotic resistance mechanisms by dominant species in the native range that were absent in the introduced range enabling W. auropunctata to become a successful invader. We compared discovery-dominance trade-offs between W.auropunctata and dominant ant species under natural and laboratory conditions in the native range of Mexico and introduced range of Puerto Rico. Our study considered both short-term and long-term competitive dynamics. We hypothesize that in the native range W. auropunctata coexists with native species because of competitive trade-offs operating at smaller spatial scales whereas in the introduced range W. auropuncta dominates the local ant community by excelling in both discovery and dominance traits.
In the summer of 2012, a 45-ha plot was surveyed to map the spatial distribution of W. auropunctata colonies in a 300 ha shaded coffee farm in the state of Chiapas in southern Mexico (15.1735835,-92.3382748). In addition, a 30-ha plot was surveyed in a low-shade conventional farm ((15.172465,-92.3301377). We observed a total of 4 colonies in the 45-ha plot and 1 colony in the 30-ha plot ranging from 0.25-1 ha in size (Fig 1a). Within the largest W. auropunctata patch (1 ha ) we established a 50 by 50 m plot during the summer of 2013. Descriptions of the farm in Mexico and the 45-ha plot have been reported elsewhere (Vandermeer 2010). In the summer of 2014, we surveyed 11 small coffee farms in the mountainous regions of Puerto Rico within the municipalities of Orocovis, Lares, Adjuntas, and Utuado (18.175850, -66.4155700). Within the largest shaded coffee farm in Orocovis (5 ha), we established a 20 x 20 m plot to map the spatial distribution of W. auropunctata _ populations. Site locations in both Mexico and Puerto Rico were selected so that native ant species, both arboreal and ground-foraging, overlapped in their ranges with _W. auropunctata populations. Differences in farm size reflected the fact that coffee farms in the Soconoscu regions of Mexic are relatively large in size (300 ha) as compared to the central region of Puerto Rico, where the average farm size is considerable small (5 ha).
In Mexico, a 4- by 4 m grid was established within the 50 x 50 m plot during the summer of 2013. Ants were sampled once with tuna baits in the summer of 2013, 2014, and 2015. Tuna baits were placed on the ground and the nearest coffee plants, yielding a total of 154 grid points. Species were checked at each grid point after 30 minutes and their presence or absence was recorded. Identification of ants species was primarily done in the field except for unknown species which we identified in the lab to genus level and assigned a morphological species list. While cross-yearly morphological species list was approximate, the five most commonly occurring species forming spatial patterns were readily identifiable in the field. In Puerto Rico, the same sampling protocol was implemented using a 4-by-4m grid within a 20 x 20 m during the summer of 2014. The main difference is that in Puerto Rico the species richness was considerably lower (~ 10 species) as compared with our Mexican field site (> 100 species).
In Mexico, we conducted discovery-dominance competition experiments in the field between W. auropunctata and dominant ant species. In particular, we conducted short-term competition experiments between W. auropunctata and ground-foraging ants such as Solenopsis geminata and Pheidole protensa as well as with arboreal ants such as Solenopsis picea and Pheidole synanthropica. We estimated the timing of discovery and dominance by placing small tuna baits in a 1 meter line right at the border between ant colonies and subsequently checked the baits at 30-s intervals for up to 90 minutes (Perfecto 1994). For the ground-foraging experiment ten replicates were carried out per species. For our arboreal competition experiments, we placed tuna baits on bamboo branches between coffee trees. We conducted 5 replicates for each of the arboreal species. We recorded the time until discovery of baits ( arrival of first ant), time until recruitment of ants to baits (5 or more ants), and the total number of ant workers at baits. For the long-term laboratory competition experiments between W.auropunctata and S.picea, we assessed the total number of live ant workers after 14 days of running the experiment. In Puerto Rico, we followed the same protocol for assessing the competitive dynamics between W. auropunctata and dominant ants. We performed W. auropunctata competition experiments with the ground-foraging ant Solenopsis invicta and arboreal ant Linipethema iniquum for 90-minutes. Lastly, we conducted a long-term competition experiment between W. auropunctata and L. iniquum for 14 days.
We analyzed the differences among species with respect to total number of workers at baits, time to discovery, and the time to recruitment to baits. Prior to conducting statistical analyses, we plotted the data to test for normality using the Shapiro-Wilk test. Because of the non-normality, we used the non-parametric Kruskal-Wallis rank sum test with the library stats( R Core Development Team 2013). Nemenyi's post-hoc tests were applied for pairwise multiple comparisons using the PMCMR package in R, v. 2.15. In situations, where discovery or recruitment time to resources was not observed, we assigned the maximum time to each species (90 minutes). Since we applied a non-parametric rank based test, species that did not discover or recruit the minimum workers were assigned the last rank (Bertelsmeier 2015).
Competition experiments in Mexico tested species differences in the average number of ants, discovery, and recruitment time to baits. Short-term competition experiments showed that species differed significantly in the number of ants simultaneously occurring at baits in the field. Ground-foraging ants S. geminata and P. protensa were superior competitors against W.auropunctata populations (X^2=55.976, P < 0.0001; X^2=15.641, P < 0.0001). For the arboreal ants, S.picea and P. synanthropica populations displayed superiority against W. auropunctata (X^2= 9.8396, P = 0.0017; X^2= P < 0.0001). In our long-term experiment, we observed significant differences in the total number of live ant workers found (F=5.4371, P= 0.01356) after 21 days in treatments. In most cases, species did not differ significantly in discovery time. Except in the case of P. protensa (X^2= 4.1823, P=0.041), we did not find a difference in discovery time for the ground-foraging ant S. geminata (X^2= 0.33037, P=0.57). The arboreal species S.picea and P.synanthropica did not differ significantly with respect to W. auropunctata (X^2= 0.25204, P=0.62; X^2=2.0979, P=0.15). We also did not observe any differences in recruitment among ground-foraging ants S. geminata and P. protensa (X^2= 2.082, P= 0.15; X^2=2.3475, P=0.13) and arboreal ants S. picea and P.synanthropia (X^2=0.327, P=0.57; X^2=2.79, p=0.095). In Puerto Rico, we found 2 dominant ants that competed for resources with W.auropunctata. Competitive interactions with the ground-foraging ant S. invicta showed no significant differences at baits (X^2= 3.485, P= 0.062), while W.auropunctata showed superiority against the arboreal ant L.iniquum in the short-term competition experiment (X^2= 47.982, P <0.0001). However, we did not find a difference in the number of live ant workers found in the long-term competition experiment (F=2.52, P=0.07). Similarly, there was no difference in the discovery rate of resources with the ground-foraging ant S. invicta (X^2 = 0.09, P=0.76) and arboreal ant L. iniquum (X^2=2.6853, P=0.11). Lastly, no differences were found in recruitement time for S. invicta (X^2 = 1.05 , P=0.3) and L. iniquum (X^2= 3.35 , P= 0.6 ).
Our study compared ecological dynamics, both short and long-term, of W. auropunctata invasion within the native range of Mexico and introduced range of Puerto Rico. Combining field and laboratory experiments, we assessed differences in the number of workers at baits. In Mexico, the ground-foraging ants S.geminata and P.protensa displayed greater superiority against W. auropunctata, while the arboreal ants S.picea and P.synanthropica showed dominance over W.auropunctata populations. However, long-term competitive dynamics in the laboratory showed that W.auropunctata is superior against S. picea. In Puerto Rico, W. auropunctata was not found to be superior in areas dominated by _S. invicta, but laboratory experiments with L. iniquum showed that W. auropunctata was superior at baits in the short-run. This pattern of W. auropunctata dominance did not hold in long-term experiments with L. iniquum. We did not however observe asymmetries in discovery and recruitment time to baits among ants, except in the case of P. protensa which was found to be a faster resource discoverer in Mexico. Our results are consistent with the biotic resistance hypothesis (Elton 1958) as we found that W. auropunctata populations were repelled by resistance ants as a result of inter-specific competition in their native range of Mexico, while in the introduced range of Puerto Rico W. auropunctata populations faced less resistance. We have found that biotic resistance from dominant competitors occurs both terrestrially and arboreally. S. geminata is known to be an aggressive competitor that can withstand higher temperatures and outcompeted other ants in open patches with light gaps (Perfecto 2011). However, the presence of specialized phorid parasitoids limit the spatial distribution of S. geminata and enable W.auropunctata to thrive (Morrison 1999). Another ground-nesting ant P. protensa is a slow moving forager whose colony consists of many small nests allowing foragers to quickly overtake resources (citation needed). P. synanthropica nests in the ground, but also forages in the coffee trees to tend hemipterans. Workers recruit rapidly to resources and they are frequently found to swarm baits thereby excluding W. auropunctata. The arboreal ant S. picea commonly nests in the tree trunk of coffee trees and it's polydymous nature enables it to have spatially distributed colonies making it difficult for W. auropuctata to occupy and dominate in trees. Although we found that S. picea outperformed W. auropunctata in the field, S. picea populations were not able to dominate W. auropunctata in the long-term experiment in the laboratory. Previously, Vonshak et al. (2012) conducted both short (2 h) and long-term (21 days) competition experiments between W. auropunctata and two other ant species and found that in the short-term W.auropunctata was the poorest competitor, while in the long-term it was considered the most aggressive species as it dominated baits and invaded opponent's nests. This result is consist with our findings of W. auropunctata competitive dynamics. The density of S. picea nests in coffee bushes appears to confer an advantage in the short-term since workers rapidly detect resources all the while W. auropunctata takes a long time to exploit and recruit to baits. However, during the long-term experiment W. auropunctata was by far the most aggressive competitor overtaking most of the baits and moving it's entire nest (including queens, brood, and workers) into nests occupied by S. picea colonies thus suggesting a high degree of behavioral plasticity. Previous research into invasive ant species suggests that their success is in part due to breaking the discovery-dominance trade-off, being good at both discovering and dominating resources as compared to their native counterparts, and that such a competitive advantage provides a direct mechanism to explain invasiveness (Holway 1999). However, our results suggest that W. auropunctata populations are not equally good exploiters and recruiters to baits. Comparisons between Mexico and Puerto Rico showed that W. auropunctata was neither good at exploiting or dominating resources in the short-term, while long-term competitive dynamics in the lab showed that W. auropunctata were good at dominating baits against S. picea and L. iniquum. Our results suggest that W. auropunctata's invasive success is not correlated with superiority in discovery and dominance strategies. Very few studies have actually examined competitive interactions between invasive and native communities at contact zones (Krushelnycky 2009). Studies on Argentine ant invasions in California (i.e. introduced range) have shown that Argentine ants outperformed native ants in discovering and recruiting workers to baits (Holway 1999). Among the four worst invasive ant species, recent laboratory studies showed that W. auropunctata was the slowest at discovering and recruiting to baits (Bertelsmeier 2015), while dyadic confrontations showed that W. auropunctata was the most dominant species (Bertelsmeier 2015a). While this patterns is in general agreement with our short-term experiments, it's important to consider long-term competitive dynamics (i.e. weeks) that might provide insights into changing behavioral strategies. As (Vonshak et al. 2012) noted that W. auropunctata behaves as an insinuator species in the short-term while showing aggression over longer time periods. Our study focused primarily on biotic interactions between W. auropunctata and resident ants and the way in which discovery-dominance trade-offs can explain the astounding invasive success of W. auropunctata populations. While competitive interactions