Recruitment in the harvester ant, Pogonomyrmex occidentalis: effects of experimental removal

We examined the importance of experimental removal of mature colonies on colony recruitment in the western harvester ant Pogonomyrmex occidentalis. To test the common assumption that established colonies suppress the establishment of new colonies we removed all colonies in ten 0.25 ha plots in 1996 and an additional five plots in 1997 and measured new colony recruitment in 1997, 1998, and 1999. Using a blocked, paired plot design we found that removal of colonies increased new colony recruitment in some areas of the site, but not others. Spatial variation in the importance of established colonies to recruitment was consistent across years; blocks in which density dependence was important in one year exhibited density dependent recruitment in following years. We estimated that in the blocks where recruitment was affected by established colonies, they accounted for less than 10% of the mortality of foundress queens. The increase in the number of new recruits (on average two additional new colonies) was considerably less than the number of colonies removed; average colony density in the removal plots was 14 colonies per 0.25 ha plot. The consistent lack of importance of established colonies to recruitment in one block and the relatively small response to colony removal in the other blocks suggests that the number of new colonies in a year may not be equivalent to the number of deaths of established colonies in that year. Space limitation is an important influence on recruitment in P. occidentalis, but the magnitude of the limitation varies spatially.