• Formicidae;
  • longitudinal gradient;
  • path analysis;
  • productivity hypothesis;
  • species richness

Abstract.  1. We analysed ant species richness-environment relationships across the Subantartic-Patagonian transition, in southern South America. We tested the predictions of the (i) thermal limitation hypothesis: temperature limits ant species richness, (ii) the productivity hypothesis: ant richness is driven by the indirect effect of climate (temperature and precipitation) mediated by changes in plant environment (plant cover and litter accumulation). We also evaluated the effects of (iii) plant species richness, and (iv) habitat use by cattle on richness.

2. We collected ants using 450 pitfall traps within a ∼150 × 150 km area. We used path analysis to evaluate the associations of ant richness with environmental predictors. We evaluated the influence of local abundance variation on richness-environment relationships; we applied principal co-ordinates of neighbour matrices to model the spatial autocorrelation in the data and combined spatial and ecological descriptors of ant richness into partial regression.

3. Ant richness responded positively to indirect effects of precipitation mediated by shrub cover and tree cover, towards the west, and by litter accumulation, towards the east of the transition. Direct climatic effects on richness were negative or weak throughout. The local increase in plant species richness and habitat disturbance by cattle promoted an increase in ant richness, mostly towards the east of the transition. These results were not seriously affected by variation in local abundance and the spatial autocorrelation in the data.

4. We suggest that climate, either directly or indirectly, through its biological effect on plant environment interacts with micro-spatial variation in habitat heterogeneity and disturbance to account for ant species diversity across this transition.