Kats were transferred from the 

 ... 'lo a polythene bag in which they 

 Si anacsthci.sed with ether. While 

 i„ ihc hag. they were weighed. They 

 A . a ih,-n ear- tagged with Moncl 

 imgerluig tags. Alter recovery, they 

 vvcie released. 



I weotv-Liiic rats. trapped away 

 from the grid, were killed ifl order 

 )0 obtain skins and skulls. Body 

 measurements of these were also 

 „ Ku le. Twenty skulls collected at 

 N.Hijec in (nppsland. by Mr R, W, 

 W.unekc of the Fisheries and Wild- 

 [jfe Department, were measured and 

 MfjWgfcd with measurements ol 

 the skulls of the island population. 

 Another ten skulls loaned by the 

 National Museum of Victoria, col 

 lected at Portland. Western Vie- 

 uiria. were also measured and com- 

 pared with the Gleiiuie population 



Two live rats, one of each son. 

 were removed from the island to be 

 used in breeding experiments, but 

 in fact they never bred in captivity, 

 lor reasons unknown 



Kist't is 

 I Drnsiry iMirmile\ 



the trapping programme was de- 

 signed for estimation ot home range 

 am! population density. Imp .suc- 

 cess was calculated by dividing the 

 total nnnibei of rats caught by the 

 number ol traps set and expressing 

 this as a percentage. Any trap which 

 wjs lound empty, with the bait 

 temoved, was no! counted as a set 



trap. Any trap found sprung, with 

 the bait still inside was counted as 

 half a set trap, because it would 

 prohably have been open, on the 

 average, for half a night. Traps o n 

 taining other species were not 

 counted as set traps. 



thus; 



I DON 



I rap success — _ 



r — ia r- fit (5753 I 



N = total individuals caught. 

 T = total traps set. 

 a traps found without halt, 

 h traps with other species, 

 C empty traps, sprung, with bail 

 inside. 



By this method, the overall trap 

 success for the period was 



101'/. In some cases, two rats were 

 lound in a trap. The trap success was 

 so high in each period that it could 

 not be used as a method of deter- 

 mining population density changes. 



ThC population density was cal- 

 culated by a method outlined in 

 libei hardl (1969). The formula used 

 is given below. 



.is I I. where N population si/c 

 n-i s totaJ captures. 



r total individuals 

 caught. 



The fust estimate (See lahle I) 

 wa.s made using data collected ovei 

 only three nights. This resulted in an 

 inaccurate estimation because Is-rl 

 is small. Consequently, a small error 

 in (s-ri results in a large error in N. 

 The estimates for later periods arc 

 probably more reliable. 



T vri i I 



Population density. 



Date 



Number captured 



fsiimnted 



ItuliMjuals 



M 



F 



Juv. 



population size 



per acre 



December l%7 

 l.miury |sV>X 

 Fsbnofji 



Hlnuuiv 

 (Intel' 

 December [?M 



30 

 .V) 

 IK 



33 



IK 

 W 



41 



25 



J 



67 



•n 



V7 



6<l 

 25 

 27 



33 



31 



2K 



6 



M2 



25 



fobruary. 1971 



