Abstract
Investigation Overview: Giant skinks (Otago skinks Oligosoma otagense and grand skinks O. grande) are thought to be declining and are therefore classified by DoC as threatened. These two species are now only found in two 'small' and widely spaced areas of Otago (Macraes Flat/Middlemarch and the Lindis Pass; Whitaker & Loh 1995). All known populations live on schist outcrops, either in the form of bluffs or as free standing tors. The habitat around the tors is a variety of developed and undeveloped native tussock grassland. Sympatric with giant skinks are five other lizard species:- Hoplodactylus maculatus (common grey gecko), O. maccanni (common spotted skink), O. nigriplantare polychrome (common striped skink), O. chloronoton (green skink), and O. inconspicuum (speckled skink). The common skinks are not threatened but potentially affect the giant skinks. They may compete for food or affect predation pressure on giant species by being alternative prey for predators. Accordingly, both common and giant skink ecology was studied.
Predation was identified by the giant skink recovery plan (Whitaker & Loh 1995) as a potential cause of the decline of both giant skink species. Feral cats, ferrets, stoats, ship rats, Norway rats, mice, little owls, white backed magpies, falcons and humans are all known to prey on lizards and are all present in giant skink areas (Whitaker & Loh 1995). Feral cats definitely eat giant skinks (Baker 1989; Daugherty & Towns 1991). However, there was no conclusive evidence that any of the potential predators poses a significant threat to the continued survival of giant skinks. This study determined the abundance and diet of mammalian predators in giant skink habitats. It also investigated how biotic factors, such as rabbit abundance and habitat development, affect predator abundance and diet.
The abundance of grand skinks declines following the conversion of tussock into exotic grassland (Whitaker 1995). The mechanism for the decline is not understood but Whitaker (1995) believes the problem is occurring off the tors. Changes to the grassland may have an impact on skinks by reducing food availability. Habitat changes may alter the predator abundance or diet so more skinks are preyed on. This study investigated changes in predator abundance and diet following grassland development.
When this study started very little information was available on the impacts of predators on lizard popu1ations living in tussock grassland. This study aimed to produce baseline information on predator abundance as well as lizard predation rates, to enable informed decisions on the need for predator control to conserve giant skinks. Predator control is expensive and has potential unwanted side effects such as the poisoning of non target species (especially important it affects giant skinks), rabbit eruptions and possible changes in the predator guild. Therefore the decision on whether to control predators should be based on reliable knowledge. The study aimed to provide some of that knowledge and identify areas where information is still needed.
Objectives:
1. Describe the abundance of potential mammalian predators of lizards in tussock grassland.
2. Describe the summer diet of cats and ferrets in tussock grassland.
3. Describe intraspecific and interspecific differences in diet within and between predator species.
4. Describe the effect of rabbit abundance on cats and ferret abundance.
5. Describe the effect of grassland development on mammalian predator abundance and diet.
6. Measure the power of a scan-counting technique for diurnal tor dwelling lizards and advise on the sample size required to accurately assess the true population mean.
7. Develop recommendations on the implementation of a long term monitoring programme to detect changes in giant skink abundance.
8. Describe a method for determining relative abundance for ground dwelling skink species which can be used to compare with lizard predation rates as attained from stomach/scat analysis..
9. Describe the affects of grassland development on capture rates of common skinks (0. maccanni, O. n. polychroma).
10. Describe the affects of lizard sex and age on mammalian predation rates in common skinks (0. maccanni, O. n. polychroma).
11. Measure the number of giant skinks removed by cats and ferrets and scale these predation rates against population size and productivity.
12. Provide managers with advice on whether predator control is needed or a wasted effort.
Methods:
1. Predators were trapped from twenty five 2 km long trap lines in the Macraes Flat/Middlemarch are. The cats and ferrets were autopsied and their gut contents examined to determine the prey species eaten. Cat and ferret scats were also collected and analysed Lizard species were identified from scale colouration or subdigital lamellae counts. Lagomorphs, rodents and possoms were identified by examining the fur under the microscope.
2. Relative abundance of tor dwelling lizards was determined by scanning the rock tors on standardised transects.
3. The species and size/age distribution of common skink species in the grassland were assessed by pitfall trapping.
4. Risk assessments were carried out using a high and low risk scenario. The high risk scenario used the highest predicted predation rate from this study combined with the lowest giant skink population density (Patterson 1992). The low risk scenario used the lowest predicted predation rate with the highest population density (Whitaker & Loh 1995).
Results, Conclusions and Recommendations:
1. The mammalian predator guild consists of cats, ferrets, mice and low densities of stoats and rats. Ferrets are present in high densities. Cats were the most frequent mammalian predator caught after hunting during the day. Recommendation: If predator control is carried out for giant skink conservation then cats should be targeted.
2. No stoats were caught during this study. Subsequently however, stoats have been caught and observed in giant skink habitat. The largely diurnal behaviour and the highest proportion of lizard in the diet of any mustelid potentially make stoats a major predator of giant skinks. There is an indication from previous studies that long term suppression of cat and ferret abundance may result in increase stoat abundance. Recommendation: Using a non removal technique, monitor all mammalian predators in giant skink habitats. Ensure that a mechanism is developed to prevent the establishment of large stoat populations.
3. Lagomorph was the main prey species for both cats and ferrets. However, no linear relationship was found between rabbit and predator abundance. Manipulating rabbit abundance to control predators may result in the predators including more giant skink in the diet. Not enough is known about rabbit/predator interactions to stop diet switches from occurring. Rodent was infrequent in the diet. This probably reflects low rodent abundance in mid-altitude tussock grassland. Recommendation: Given the lack of understanding of rabbit/predator interactions no attempt should be made to control predator numbers through manipulating rabbit abundance.
4. Predator abundance and diet were not significant different in developed and undeveloped grassland. The capture rates of common skinks species declined following grassland development. The capture rates of lizards within remnant tussock within developed sites did not differ significantly from those in undeveloped areas. The decline of ground dwelling skinks following grassland development may be explained by habitat destruction. Giant skink declines can not be explained in this manner since tors are not destroyed. Recommendation: Investigate whether giant skinks decline following grassland development is a direct result of vegetation change or an indirect result of changes in insect species composition and abundance.
5. Harriers are potentially important in tussock grassland as their main prey species (rabbits) is the same as ferret and cats. Harriers are also known to prey on lizards. Significantly more harriers were caught in developed pasture. Recommendation: Determine the amount of predation by harriers qf giant skinks, particularly in the inter-tor region of developed pasture.
6. Significantly more hedgehogs were found in developed pasture. Although nocturnal, hedgehogs may act as a giant skink competitor since both are insectivorous. Recommendation: Determine diet overlap of hedgehogs and giant skinks in developed grassland to predict their potential impact on giant skinks and advise on the need or otherwise for their control.
7. DoC has decided to remove grazing from land purchased for giant skink conservation. The expected result will be vegetation regrowth. This has potential harmful impacts on giant skinks since long grass is known to increase prey species such as rabbits and rodents. Predator abundance may increase as a result. If the increased prey abundance is insufficient to sustain the increased predator abundance then increased lizard predation may occur. Recommendation: Monitor rodent and rabbit abundance before and after grazing removal. Monitor predator species abundance and known giant skink populations after grazing removal to ensure no detrimental effects. Have in place, before grazing removal, a contingency plan for the protection of giant skinks if large scale increases in predation occur.
8. By using a variation of the standard scanning technique used for giant skink distribution surveys it may be possible to estimate giant skink relative abundance. Eight repeated surveys were needed on each trap line to estimate the population mean within 50%. It is vital to develop an accurate population estimate to measure the affects of future management practises. Recommendation: A single method of population estimate should be chosen which is used for all population estimates. Carry out a power analysis to determine the sampling effort to required to detect a 30% (or less) change in abundance over 5 years. Then develop a long term plan (and funding) for regular monitoring of populations. The monitoring should not occur only at the 'best' sites since these are likely to be the last sites affected by population collapse.
9. There were no conclusive examples of giant skink predation by ferrets found in this study. Difficulties in the identification of lizard species meant one lizard specimen was potentially from a giant skink. High densities of ferrets in giant skink habitat mean if ferrets only kill one giant skink annually predation rates may be as high as 19.54 giant skinks km-2 year-1. Therefore ferrets should not be excluded when considering reasons for giant skink decline. Recommendation: If the technique used for cat and ferret control is not the same and if funding is extremely limited then only cats should be targeted. However every effort should be made if carrying out predator control to include ferret control as a component.
10. Only one example of giant skink predation by a cat was found. All known and potential examples of giant skink predation were used to calculate a cat predation rate of between 4.24 - 42.47 giant skinks km-2 year-1. Even in high density populations such as the Redbank Conservation Area 43% of annual production by giant skinks may be removed by cat predation. Recommendation: The implementation of a structured experimental management programme is needed with replicates and non reatment controls so the potential positive and negative effects of wide scale predator control can be determined.