A Long-Term Study of Rattus tiomanicus Populations in an Oil Palm Plantation in Johore, Malaysia: I. Study Methods and Population Size Without Control

Abstract

R. tiomanicus (Miller), usually virtually the sole rat species causing damage to oil palms in the Malay Peninsula, has been studied since 1969, and the first 10 yr are reported. Populations in a saturated population area (SPA-80 ha) undisturbed by control measures, and a depopulated area (DPA-410 ha) subject to control episodes, were estimated regularly by mark release recapture (MRR). In recurrent trapping plots (RTP) traps are set for a 4-night period at intervals (i) of 4 wk (i1-32), 5 wk (i33-100) and 7 wk (i101 onwards). In destructive capture plots (DCP), MRR is followed by hunting, and from i17, also by breakback trapping. The palms were planted in 1959 on a 9.14 m equilateral triangle pattern, which is used as the basis of the trapping grid, at 3 traps/palm. Overall trapping success in RTP of SPA was 25% occupation in 132,000 opportunities. Population estimates are made from the RTP, by the method of Jolly (1965).sbd..cxa.N (Jolly).sbd.and from the DCP by Lincoln index (adjusted by Bailey''s method) LI (.cxa.N)B, where the number of marked animals (M) is composed of the total previously marked during the trapping period, and the index sample (n) is obtained by hunting. Both capture techniques take a reasonable proportion of rats; marks are permanent and do not affect the animals too adversely; evidence is that short-term changes and movement (edge effects) do not bias estimates unduly (if the plot size is 5 rows .times. 10 palms or more). Capturing n by an indepent method reduces the bias caused by individual variation in trappability in MRR alone. Apparently the latter influence is smaller in RTP, probably due to the frequent presence of traps and the inclusion in analysis of rats which are captured only rarely. The techniques appear reliable, that of the DCP for a single-occasion estimate (with allowance for a relatively large sampling error), and of the RTP for trend comparisions. In very low populations, error can be proportionately greater. .cxa.N (Jolly) relates quite closely to numbers seen (i.e., those captured in a given i plus others seen both earlier and later.sbd.ni, + zi) while LI .cxa.N(B) is mostly about 1.5 to 3 times greater than number seen (M + n - m). When number seen is < 6, or .cxa.N (Jolly) does not compute due to zeros, number seen times 1.5-3 is used as a basis to estimate a range. Only 27 records of other rat species were made (cf. 46,716 of R. tiomanicus) comprising R. exulans, R. whiteheadi and R. rajah, all in DPA when R. tiomanicus was at low numbers following control. In the SPA, the population in the RTP fluctuated slowly, reaching its lowest .cxa.N (Jolly) of 183 ha-1 in mid-1972 and a high plateau of .apprx. 537 in 1976. Successive troughs were separated by about 6-7 yr.