The size grain-hypothesis in ants: ¿conflicting evidence or confounded perspective?

ALEJANDRO GUSTAVO FARJI BRENER

ECOLOGICAL ENTOMOLOGY

WILEY-BLACKWELL PUBLISHING, INC

Año: 2004 vol. 29 p. 380 - 380

0307-6946

Longer legs decrease the cost of running and permit animals to walk over environmental interstices instead of through them. However, long legs can complicate the ability of an organism to penetrate interstices that contain food sources and refugia. Since the perception of environmental interstices (i.e. rugosity) increases as body size decrease (Size-Grain Hypothesis, hereafter SGH), Kaspari and Weiser (1999) propose that the benefits of long legs are outweighed by the cost of leg construction in small ants, and thus natural selection should favor proportionally smaller legs with decreasing body size. Two studies have demonstrated that leg length decreases disproportionally to body size in 142 ant species belonging to five subfamilies mainly from the New World (Kaspari & Weiser, 1999; Espadaler & Gómez, 2001). However, Parr et al. (2003) recently obtained apparently contradictory results.  Using data from Southern African ants, they analyzed the scaling of leg length and body length of 86 ant species using phylogenetic independent contrasts, and also compared the body sizes of ant species from areas of contrasting habitat complexity (unburned and experimentally burned plots). The allometric approach using the phylogenetic independent contrast method provided strong support for the SGH; but the smaller body size classes were no more speciose on the control (rugose) plots than on the annually burnt (planar) plots. Parr et al. (2003) interpreted this field evidence contrary to the predictions of the SGH. I believe that this field evidence does not contradict the SGH because the manipulative nature of the experiment of Parr et al. (2003) represents an ecological rather than an evolutionary test. The SGH proposes that natural selection should favor proportionally smaller legs with decreasing body size, and thus that the degree of rugosity of an environment will favor or disfavor different body sizes in different lineages of ants. Because natural selection depends on sustained interactions between the biological actors implicated (in this case, ant species and the rugosity level of the environment) over evolutionary time, different body size distributions of ant species between burned and unburned plots are not necessarily predicted by the SGH. Although for the ants, burned plots are more planar than unburned plots, the experimental situation of the Parr et al. (2003) experiments is probably not the natural environment where the morphology of the ants evolved. Moreover, because foundress queens of many ant species can fly great distances and disperse widely from their nest to establish new colonies (Hölldobler & Wilson, 1990), the spatial association between ants and the rugosity condition of the plots may be temporally inconsistent even for a given ant species, making the allometric relationships less consistent than predicted by the SGH. I agree with Parr et al. (2003) that this hypothesis needs more field tests to reinforce its validity, including measurements of the functional relationship between the allometric pattern proposed by the SGH and habitat perception (e.g. Farji-Brener et al., 2004). Because the SGH concerns evolutionary patterns, natural habitats with different levels of rugosity (e.g. Yanoviak & Kaspari, 2000), rather experimental plots, should offer more useful tests of predictions of SGH concerning body sizes classes of ant assemblages.