BIOLOGY OF RODENT S 



207 



TABLE 1 



Summary of Female Rat Reproductive Data from 

 Northern Marshall Islands 



•From Temme (1981). 



mainland; dispersal is nearly absent. In this context, Krebs, 

 Keller, and Tamarin (1969) have shown that blocking 

 dispersal leads to unusually high densities in some rodent 

 species. 



Adrenal gland weight has been widely used to assess 

 the role of agonistic behavior in inducing a physiological 

 stress response. Laboratory and field studies have demon- 

 strated a positive correlation between adrenal gland weight 

 and both population density and loss of fights (reviewed by 

 Christian, 1978). Other work has demonstrated inhibitory 

 effects of crowding and aggression on reproductive func- 

 tion and disease-defense mechanisms. In this Enewetak 

 study (1977 to 1978), the relationships between popula- 

 tion density, adrenal size, wounding, and parasite load 

 were examined. 



Trapping 



Conventional Victor kill traps, baited with fresh 

 coconut, were set at approximately 7-m intervals before 

 sunset, emptied and rebaited several times during the 

 night, and pulled at midnight or the next morning. Because 

 of the variability in trapping effort, results are expressed in 

 rats caught per trap hour rather than per trap night. 



Necropsy 



Routine body measurements and reproductive data of 

 scrotal males and sexually mature, nonpregnant females 

 were noted; adrenals were preserved in 10% buffered for- 

 malin and later cleaned and weighed wet. The ratio of the 

 combined adrenal weights to the head and body length 

 was used in all subsequent calculations. Small intestines 

 were removed and also fixed in buffered formalin; later 

 they were opened and tapeworms counted. After clipping 

 hair from the posterior one-third of the back, remaining 

 hair was removed with a depilatory; fresh wounds were 

 counted. 



Population Density and Adrenal Weight 



On five Polynesian rat islets, with densities ranging 

 from 0.06 to 0.58 rats caught per trap hour, no consistent 

 relationship between population density and adrenal weight 

 was apparent for either sex. 



For five roof rat islets, with densities ranging from 

 0.006 to 0.062 rats per trap hour, correlation analysis 

 revealed a significant, positive relationship between density 

 and adrenal weight in both sexes for this species (Fig. 3). 

 Enewetak Islet had a low population density, because of a 

 control program, and the lowest adrenal weights. Medren, 

 with large numbers of abandoned buildings, supported the 

 densest population and the heaviest adrenal weights. Other 

 islets were intermediate. Population densities on Enjebi 

 increased from 0.029 in 1977 to 0.045 in 1978, probably 

 because the island was chained between samples, and the 

 cut vegetation was placed in piles as part of a cleanup pro- 

 gram, thus concentrating the survivors. Adrenal weight/ 

 body length ratios correspondingly increased from 0.28 

 (males) and 0.27 (females) to 0.40 (males) and 0.44 

 (females) (Table 2). 



Population Density and Wounding 



The mean number of back wounds per rat (species 

 combined) increased as population density increased (Fig. 

 4). Males generally had more wounds than females, partic- 

 ularly on the densely populated Polynesian rat islets of 

 Bijire and Aomon. On low density, mainly roof rat islets, 

 sex differences were not great. 



Wounding and Adrenal Weight 



Polynesian rats with fewer than two wounds had signifi- 

 cantly lower adrenal weights than did those with two or 

 more wounds (t-test, P < .001, both sexes; Table 3). For 

 the less frequently wounded roof rats, both males and 

 females with no wounds had significantly lighter adrenals 

 than did those with wounds (t-test, P < .025). 



