Are all ant mosaics caused by competition?

In this paper we tested the hypothesis that ant mosaics are generated and maintained by stochastic processes, against the general idea that deterministic processes, such as competition, are the main forces that structure these communities. We analysed 14 published ant mosaics, described for crops and natural forests, transforming their data in presence/absence matrices. These matrices were submitted to species co-occurrence analysis, using software that generates null models. The observed species co-occurrence index was then statistically compared to the null matrices, to test whether there is evidence for biological processes structuring the ant communities. From the 14 mosaics analysed, seven had less and one had more co-occurrence than expected by chance. In six mosaics, species co-occurrence was within the 95% limits of frequency distribution of randomized matrices. Even though the observation of ant spatial distribution, together with association indices between species pairs, have usually been interpreted as a result of negative interactions, our results show that in only 50% of the studied mosaics there is evidence for this explanation. Furthermore, there are alternative hypotheses to explain the result of less co-occurrence than expected by chance, besides the competition hypothesis. Association of ant species with resources and conditions may also explain the observed pattern, although these hypotheses are rarely exploited. In conclusion, competition may be an important process when studying assembly rules in ant communities, although it is certainly not the only process involved. The importance of other biological processes and of stochastic events has to be taken into account to explain the species distribution patterns.