Lincoln University Wildlife Management Report 18 THE ROLE OF NON-TOXIC CEREAL PREFEED AND POSTFEED (DYED AND UNDYED) IN THE DEVELOPMENT AND MAINTENANCE OF 1080 CEREAL BAIT SHYNESS IN CAPTIVE POSSUMS James Ross Ecology and Entomology Group, PO Box 84, Lincoln University Funding FRST programme November 1997 The role of non toxic cereal prefeed and postfeed (dyed & undyed) in the development and rna intenance of 1080 cereal bait shyness in captive possums James Ross Ecology and Entomology Group Soil, Plant and Ecological Sciences Division, PO Box 84, Lincoln University Canterbury, New Zealand INTRODUCTION Previous research trials with captive possums in pens, and in the field, have shown that a long-lasting aversion to 1080 cereal bait can be induced in this species (Hickling 1994; Morgan et al. 1996). Research investigating methods of combating this aversion has branched into two main directions. The first involved an investigation into the effectiveness of alternative toxicants as a means of eradicating 1080 shy possums. The second has been an investigation to identify which components of 1080 cereal bait are involved in maintaining the aversion. It is this second research direction that is the focus of this report. f Research investigating toxic bait aversions has already been carried out for other mammalian pest species. In particular, rodent research has shown that toxic bait aversion is influenced by the taste (Naheed & Khan 1989) and texture (Prakash 1988) of the bait material. Possum research has generated similar results, with possums showing aversion to specific components of the 1080 cereal bait, such as flavour of the lure and the cereal pellet bait base (O'Connor and Matthews 1995). This result has been reasonably well investigated and researchers have shown that acute toxicant shyness can be largely overcome by switching to a non-cereal bait matrix with an different lure (Morgan et al. 1996; O'Connor and Matthews 1995). Other rodent research has shown that if rodents are exposed to a flavour (or bait) prior to poisoning this familiarisation inhibits the formation of bait aversion (Bhardwaj and Prakash 1982). Recent possum pen and field trials have likewise demonstrated that possum 1080 bait 1 aversion is reduced by 'prefeeding' non-toxic bait prior to poisoning (Moss et al. 1997). The pen trial had three treatment groups with possum consuming no prefeed, undyed prefeed or dyed prefeed. The possums then consumed a sublethal dose of 1080 in a dyed cereal pellet. On subsequent re-exposure to the toxic 1080, bait aversion was highest in the group that was not prefed and lowest in the group that ate dyed prefeed. While Moss's et al. (1997) trial was able to demonstrate that prefeeding prior to the sublethal dose influenced the development of 1080 bait aversion the researchers were unable to determine whether the level of 1080 bait aversion is also influenced by 'postfeeding' nontoxic bait; that is, providing additional non-toxic bait after ingestion of the sublethal dose and prior to any further attempt at poisoning. This workplan proposes to confirm the results of this previous pen trial and investigate the effect of postfeeding after sublethal poisoning. This research is important because prefeeding with non-toxic bait is a common strategy in possum pest control. I suggest postfeeding may have an additional benefit not previously demonstrated in possum control. OBJECTIVES • To determine the influence of non-toxic cereal prefeed (dyed & undyed) on the number of possum developing 1080 cereal bait shyness . • To determine the effectiveness of non-toxic cereal postfeeding (dyed & undyed) as a means of combating 1080 cereal bait shyness once it has been induced. METHODS This trial was undertaken from July to September 1997 using 48 wild caught possums housed in indoor pens at the Landcare Research (NZ) Ltd. captive animal facility at Lincoln. These animals were quarantined from all other possums because of a virus which had been killing at least 30% of possums in other indoor pen trials. Due to the virus these animals were also not acclimatised for the usual six week period (the virus has a four week incubation period and the three week trial could be run before any sickness took hold). 2 Health of individual possums was monitored by checking their body weight. All possums were weighed (± SOg) two times prior to the first treatment and at post-mortom. Any significant weight loss (>20%) was considered indicative of stress and these possums were withdrawn from the trial. During the trial the possums were given fruit, vegetables and cereal food. They did not receive any food in pellet form as this could have influenced their responses when they encountered cereal baits later in the trial. Remaining healthy possums were divided into four groups balanced for weight and sex (Figure 1) and given non-toxic prefeed (RSS pellet containing 0.1% cinnamon and, for one of the four groups, 0.1% green Lavanyl dye) for one week (Treatment 1). After this week each possum received Ig of 0.08% 1080 cereal bait per kg liveweight (containing 0.1% cinnamon lure and, and for three of the four group 0.1 % green Lavanyl dye; Treatment 2). At this concentration the bait represented an LDzo for possums (R. Henderson, unpublished data). After a further week the level of shyness was assessed by re-exposing each possum to 1g of the 1080 cereal bait (Treatment 3). If the possum ate less than O.S g it was categorised as shy and moved to the next phase of the trial. Possums eating more than O.Sg were presented a further two baits over the next two nights. Once the level of shyness was assessed groups 1, 2 & 3 were to prefed for a further week (Treatment 4). However, this was dependent on the number of possums still alive and treatment 4 had to be modified due to the low number of shy possums (Figure 3). The remainder of the trial continued as per the original design and the level of shyness was then reassessed by exposing the possums to a potential lethal dose of 1080 cereal bait (20g of 0.08% cone.) over two nights (Treatment S). All surviving shy possums were then exposed to a potential lethal dose of 1080 gel (containing 0.1 % cinnamon lure and 0.1% green Lavanyl dye) bait (20g of 0.13% cone.) to assess the effectiveness of this treatment as a means of overcoming 1080 cereal bait shyness (Treatment 6). All possums had access to their normal diet throughout the trial. All bait was made up by Landcare Research, Lincoln. RSS cereal bait is a proprietary mix of cereal and sugar sweeteners supplied by Animal Control Products, Waimate. Bait toxicity was confirmed by Landcare laboratory assay. 3 Treatment I r Treatment 2 Group 1 I prefced with ~dyed RS5 I~ ~ i Group 2 prefeed with undyedRS5 (--Group 3 1 i pre feed with 1----- l__undyed RS5 j Sub-lethal dose with undyed ~80 Treatment 3 Shyness confirmed Treatment 4 TreatmentS Treatment 6 l ~r~_u_n_dy_e_d~}-J,:80~"l ~ Prefeedwith _ - undyed RS5 1080 Sub-lethal dose with dyed 1080 Shyness confirmed Prefeed with undyed RS5 Sub-lethal dose with dyed 1080 Shyness confirmed No prefeed Sub-lethal dose with dyed 1080 Shyness confirmed Prefced wi th dyed RS5 Dyed 1080 Shyness confirmed No Prefeed Dyed 1080 12 possums _ Dyed 1080 1080 Gel 1~_D_l~_~_~_~I-1IOW ~I I 1080 Gel 12 possums 12 possums 12 possums 33 possums Figure 1: Flow diagram of experimental design. *Group 5 data was derived from an earlier trial; these possums were housed in the same room under identical conditions. Accordingly, they were unavailable for the final 1080 gel treatment. RESULTS Consumption of prefeed by naive possums All 48 prefed possums consumed prefeed during the first week of the trial. On average each individual consumed 36.1g (+ 2.5g S.E.) per night (Figure 2) with no significant difference in the consumption of dyed and undyed prefeed (rank sum two sample test; U=264, p=0.26). 4 o Undyed (n=36) 100 o Dyed (n=12) rrrf- § Frf- rr-rr rr- T "C <1.l E -±--f• rrr.I- rr-f- ::I ~ 10 o o "C <1.l ~ ~ ... a.. 1 1 2 4 3 5 6 7 Night Figure 2: Consumption of dyed and undyed RSS non-toxic prefeed by naive possums. Means presented + S.E. Consumption of sub-lethal dose All 81 possums consumed their sub-lethal dose on the night is was provided. There was no initial shyness of the toxic 1080 bait for any treatment group. The sublethal dose killed 15 of the 81 possums (18.5%). Bait shyness following a sub-lethal dose Group 5, which had not been prefed, had significantly more shy animals on every night (l test for independence; p<0.0003 for all nights). Possums that ate the same coloured prefeed as their sub-lethal dose (Groups 1 & 4) had the fewest shy animals following each re-exposure, however, this result was not statistically significant on any night from than possums that ate different coloured prefeed (Groups 2 & 3; X2 test for independence; p>0.22 for all nights). 5 100 OGroup 1 (n=12) o Group 2 (n=24) ~80 IillGroup 3 (n=12) '-" rn rn l1li Group 4 (n=33) Q) ~60 .s:::. rn Q) 1; 40 .!!! :J E § 20 u 0 2 1 3 Number of re-exposures to 1080 Figure 3: Percentage of possums developing shyness following reexposure to 19 sub-lethal doses of 1080; sample size in parentheses. After the three re-exposures there were insufficient numbers of shy possums from Groups 1, 2, 3 and 4 to proceed as per the original plan. Possums that were still not shy received four more sub-lethal doses over the next two weeks (Table 1). Table 1: Amount of bait provided, and the cumulative proportion and number of shy possums after each of7 re-exposures to 0.08% 1080 cereal bait. Re-exposure Bait Cumulative proportion Cumulative No. shy provided (g) Same prefeed1 Different prefeed2 Same Different 2 3 4 5 6 7 I 2 1 1 2 3 4 1 5% 10% 14% 17% 22% 53% 61% 17% 17% 22% 22% 30% 57% 63% 4 4 5 5 6 11 12 1 2 3 3 4 7 8 Groups 1 and 4 Groups 2 and 3 The 20 shy animals obtained from these four additional exposures were still insufficient for four treatment groups. Treatment 4 was therefore modified for Group 2 possums (Figure 4) so that 10 of these animals received prefeed with the other 10 receiving no prefeed. The possums eating prefeed were made up from Groups 1 and 3; these continued to consume prefeed which was the same colour as their sub-lethal dose. Therefore, the presence/absence of dye was not a confounding factor. The remainder of the trial proceeded as per the original plan. 6 Groups Group I Treatment 4 ~ Prefeed with undyed RS5 6 possums 4 possums Group 3 ~ No prefeed 6 possums G roup 4 ~ I ~ Prefeed with dyed RS5 4 possums Group 5 ~ No Prefeed 10 possums Figure 4: Modifiedexpe.-imental-desigu. Consumption of postfeed by 1080 shy possums All 10 shy possums consumed some postfeed during the week of treatment 4, however nightly consumption dropped to 12.9 (±5.2), which was significantly lower than the earlier prefeed treatment prior to the sub-lethal dose (Wilcoxon signed-rank test; T=2.4, p=O.Ol). Mean nightly postfeed consumption tended to increase over time rising, from a mean of 7.5g (± 4.7g) on the first night to 16.9g (± 6.5) on the last night (Figure 5), however this increase was not statistically significant on any night (linear regression on log transformed data; r=0.20, d.f.=5, p=O.31). 7 100 § "0 Q) E ::l CIJ c 0 U 10 "0 Q) - ~ CIJ 0 Cl. 1 1 2 3 4 Night 5 6 7 Figure 5: Consumption of non-toxic postfeed by 1080 shy possums. Means presented ±SE. Consumption of 1080 cereal bait by 1080 shy possums following postfeeding Possums that had not eaten any prefeed or postfeed (Group 5) ate the least amount of 1080 on the subsequent re-exposure (Table 2) however, individual consumption was variable and the difference in consumption between groups was not statistically significant (Kruskal-Wallis test, H=2.99, d.f=2, p=0.22). Possums that had been both prefed and postfed consumed the most 1080 and consequently had the highest mortality rate, however again the differences between the groups were not significant (X 2 test for independence; p= 0.09). Table 2: Daily consumption of 1080 RS5 cereal bait by shy possums after 1 week of exposure to non-toxic postfeed. Means are presented + SE. Groups Prefed before Postfed after Mean 1080 Mortality No. sub-lethal dose sub-lethal dose consumption (g) 1&4 2&3 5 Yes Yes No 3.97 + 1.84 2.09 + 1.02 0.63 + 0.43 Yes No No 30% 20% 0% 10 10 10 Effectiveness of 1080 gel as a means of overcoming 1080 bait shyness All remaining 1080 shy possums were then exposed to 0.13% 1080 gel for one night. This treatment killed 13 of 15 (80%) of the 1080 cereal bait shy possums (Table 3). There was no significant difference in bait consumption or mortality between the prefed and non-prefed 8 possums from the previous treatment (two sample t-test; t=1.77, d.f.=13, p=0.9; x2test of association; p=0.6). Mean consumption of the 1080 gel (12.4g ±4.7) was significantly higher than consumption of 1080 cereal bait by all groups (Table 2) in the previous treatment (rank sum test; U=380, p<O.OOl). Table 3: Daily consumption of 1080 gel. Means are presented + SE. Groups Mean 1080 Mortality No. consumption (g) 1&4 11.9+3.6 2&3 12.8 + 5.7 71% 88% 7 8 DISCUSSION Consumption of non-toxic prefeed by naIve possums was reasonably consistent over the 7 nights. All possums consumed some prefeed, so there was no evidence of a neophobic response to the novel food source. The presence or absence of dye had no statistically significant influence on prefeed consumption, although the possums seemed to prefer the dyed prefeed on 5 of the 7 nights. Prefeeding had a significant effect on the development of shyness, with the non-prefed groups producing a significantly higher numbers of shy animals on all re-exposure nights. To induce a measurable level of shyness amongst the prefed possums required four more re-exposures with increasing amounts of 1080 bait, and even then only 62% of the survivors were shy. The presence or absence of dye had no significant influence on the number of shy animals, although the group that ate the same coloured prefeed as their sub-lethal dose did produce the lowest number of shy animals on all re-exposure nights. All 1080 shy possums consumed some non-toxic postfeed when re-exposed for 7 nights however, consumption was significantly lower than the earlier prefeed treatment with naIve animals. Consumption remained low throughout the week. Postfeeding had little effect on the number of shy possums when the animals where re-exposed to 1080 bait later in the trial. 9 In contrast, the 1080 gel was highly effective at killing 1080 cereal bait shy possums. Most (13 of 15) shy individuals consumed a lethal dose of gel only seven days after surviving a reexposure to 20g of cereal 1080 bait. Shyness to the new bait base was minimal, with the mean consumption of gel only marginally lower than mean consumption of non-toxic postfeed by the 1080 shy animals in the earlier treatment. Consumption of 1080 gel was significantly higher than consumption of 1080 cereal bait (all groups). This result suggests that bait base recognition is more important cue than toxin detection. RECOMMENDATIONS The benefits of prefeeding appear to be twofold. Firstly, it attracts additional possums to feeding stations and improves the percentage kill (Thomas et al. 1996), even though trials with captive animals have shown that individual prefed possums eat less 1080 bait than non prefed possums on first exposure (Thomas et al. 1996; Ross et al. 1987). Secondly, the survivors are less likely to become 1080 bait shy (Moss 1997; this study) In contrast, postfeeding shy possums does not appear to be an effective method for reducing the number of 1080 shy animals. 1080 gel is an effective way of killing 1080 cereal bait shy possums. This result supports earlier trial work, which showed that 1080 shyness could be largely overcome by switching both the bait matrix and lure (Morgan et al. 1996; O'Connor 1996). This result also compares favourably to an earlier pen trial in which 64-73% of 1080 shy possums were killed using a sub-acute or chronic toxicant (cholecalciferol and brodifacoum) in cereal bait. The 1080 gel was not, however as effective as using these alternative toxicants in an unfamiliar bait (apple paste) which achieved a 100% kill in the same pen trial (Ross 1997). This trial suggests that 1080 gel is effective at killing 1080 cereal bait shy possums but this result requires field validation. Also, 1080 bait is unavailable to unlicensed operators. It could be advantageous to develop a commercial gel bait which could be used by unlicensed operators in areas with 1080 shy possums. Future research should: i) investigate the efficacy of the new gel bait in the field and ii) investigate the effectiveness of gel bait containing subacute or chronic toxicants. 10 REFERENCES Bhardwaj, D., and Prakash, I. 1982. Poison aversion in Rattus rattus through acclimatisation. Indian Journal of Experimental Biology 20: 396. Hickling, G. 1994. Behavioural resistance by possums to 1080 toxicant: implications for sustainable pest management in New Zealand. Proceedings of the Science Workshop on 1080. Pp. 151-158 in Seawright, A.A.; Eason, C.T. (eds.) The Royal Society of New Zealand, Miscellaneous Series 28. Moss, Z.N. 1997. (unpublished) Implications of prefeeding for the development of bait aversion in brushtail possums (Trichosurus vulpecula). M.Sc. thesis, Lincoln University, Canterbury. 87pp. Morgan, D.R., Morriss, G., and Hickling, G.J. 1996. Induced 1080 bait shyness in captive brushtail possums and implications for management. Wildlife Research 23: 207-11. Naheed, G., and Khan, J.A. 1989. Poison-shyness and bait-shyness developed by wild rats (Rattus rattus L.). Angewandte Zoologie 4: 469-484. O'Connor, C.B., and Matthews, L.R. 1995. Behavioural mechanisms of bait pOlson avoidance. In: Proceeding of a workshop on improving control of possums. The Possum and Bovine Tb Control National Science Strategic Committee: 51-53. Prakash, I. 1988. Bait shyness and pOlson averSIOn. In: I. Prakash, (ed.) Rodent pest management, pp. 321-340. CRC Press, Florida, 480 pp. Ross, J.G., Hickling, G.J., and Morgan, D.R. 1997. Use of subacute and chronic toxicants to control sodium monofiuoroacetate (1080) bait shy possums (Trichosurus vulpecula). Proceedings of the 50th New Zealand Plant Protection Conference, 397-400. Ross, W.D., Bell, J., and Robson, D.L. 1987. Preference and acceptability of commercially prepared cereal pellets and the effectiveness of sodium monofluoroacetate (1080) on caged possums. MAF Qual. Research Report. 25 pp. 11 Thomas, M.D., Henderson, R.J., and Hickling, GJ. 1996. Improving conventional control of possums. Proceeding of a workshop organized by the Possum and Bovine Tuberculosis Control National Science Strategy Committee., Miscellaneous series 35, The Royal Society of New Zealand, Wellington, 65-69. 12