The Changes in Ant Species Distribution During Ten Years Post-Fire Regeneration of a Heath

Abstract

An 8 ha area of S English heath sampled in 1961 for species of ants and 12 habitat factors (4 physical and 8 vegetational) was resurveyed after 10 yr by the same method. Between surveys the bare areas decreased and the ericaceous shrubs (Calluna vulgaris, Erica cinerea and E. tetralix) increased. Ulex minor remained the same, but Molinia caerulea and Agrostis setacea (both grasses) and Pteridium aquilinum (bracken) decreased. In 1971, 111 single and 46 double records, a total of 203, were obtained at 157 stations. In 1961, by contrast 18/157 stations were without ants and none had more than 1 sp. present. Ant species were the same and all but one, Lasius alienus, were more abundant. Mean values of the 13 habitat variaties (the height of Calluna was added to the 12 used in 1961) were calculated for each ant species and a matrix of correlation coefficients obtained. From this principal components were extracted. The 1st (accounting for 54% of the dispersion) represents a dry, sandy, bare, warm, soil with C. vulgaris and U. minor the principal plants (dry heath); it occurs at moderate heights. The 2nd (with 26% dispersion) represents dry, sandy, well-covered, cool soil with E. cinerea and A. setacea the principal plants (dry grass heath); it occurs quite high up. The 3rd, with 13% of dispersion, represents a moist, organic, moderately-covered, cold soil with gorse and bracken the main plants (scrub); it occurs low down. The 4th (with only 7%) represents a bracken-dominated area. The 1st component in 1971 is very similar to the 1st of 1961; others were less so. The relative positions of the ants in the component space are very similar in 1961 and 1971. L. alienus and Tetramorium caespitum lie close in the 1st dimension (I) and far from Lasius niger and Myrmica scabrinodis. Component II brings M. sabuleti away from the rest and component III brings Formica fusca out. The 4th dimension is of some interest as it alone separates T. caespitum and L. alienus. Each type of habitat contains 1 representative of both the subfamilies Myrmicinae and Formicinae. About 60% of stations had the same species in 1961 as in 1971. T. caespitum and F. fusca gained most, the former largely at the expense of L. alienus. However, the habitat it took from L. alienus was not different from the habitat that L. alienus held, and a possible ecodifferentiation between the 2 spp. in 1961 had vanished by 1971. The spread of vegetation probably tipped the competitive balance in favor of T. caespitum because it can build nest mounds in vegetation and live on seeds. Fire re-establishes a sparse plant cover more suitable for L. alienus, the 2 thus coexist through being adapted to opposite vegetational phases of the burn cycle.