CHEEK POUCH CAPACITIES AND LOADING RATES OF HETEROMYID RODENTS

Rodents of the family Heteromyidae are proficient gatherers and hoarde rs of seeds. A major component of their adaptive specialization for ha rvesting and transporting seeds is their spacious, fur-lined cheek pou ches. Precise measurements of cheek pouch capacities are essential if ecologists are to understand the foraging ecology, possible constraint s on locomotion patterns, and competitive relationships of heteromyid rodents. To measure the size of these cheek pouches and the rate at wh ich animals load seeds into their pouches during seed harvest, we attr acted 56 individuals representing ten species of heteromyid rodents to bait stations in the field and allowed them to fill their cheek pouch es with seeds several times while we observed and timed the events wit h the aid of night-vision equipment. The largest load taken by each in dividual was used as an estimate of its cheek pouch capacity. At the e nd of observations, each subject was captured and its mass and other d ata gathered. The allometric relationship between cheek pouch capacity and body mass for ten species of heteromyids was significant [pouch c apacity (ml) = 0.148 body mass (g)(0.992), r(2) = 0.91, P < 0.0001]. T he regression coefficient is approximate to 1.0, which indicates that the volume of the cheek pouches scales in direct proportion to body si ze. When the data were subdivided into quadrupeds (Perognathus and Cha etodipus) and bipeds (Dipodomys) (n = 5 for each), the relationships b etween pouch capacity and body mass were significant, but the two regr essions were not significantly different from each other. When all loa ds (full and partial) were considered, subjects filled their cheek pou ches an average of 93 +/- 10% of pouch capacity (n = 185). Cheek pouch capacities from published studies of artificially filled pouches of h eteromyids in the laboratory averaged about 40% below the field measur ements obtained here. The allometric relationship between mean loading rate and body mass was also significant [seeds/s = 1.067 bodymass (g) (0.830), r(2) = 0.85, P = 0.0011), but when quadrupeds and bipeds were considered separately, the relationships were not significant. Seed d ensities and bulk densities were used to calculate packing coefficient s for seed species, which, when used in conjunction with the allometri c relationship between cheek pouch capacity and body size, can be used to estimate the maximum load carried by a heteromyid. Except for the very largest kangaroo rat species, a full pouch load of Indian rice-gr ass seeds represents less than the daily energy requirements of an act ive heteromyid.