Outcomes of the workshop: refining aerial surveys of kangaroos G. C. Grigg' and A. R. Poplel 'Department of Zoology, University of Queensland, Queensland 4072 INTRODUCTION and particularly the New South Wales study. For the first time, a correction factor specific What follows is an attempt to bring together for grey kangaroos could be proposed. The the discussions at the workshop which idea was to work out a particular correction was conducted separately from the formal factor for grey kangaroos rather than multiply- presentations. ing the existing correction factors which had been originally determined just for red ROLE OF AERIAL SURVEY IN kangaroos. Current practice by each state KANGAROO MANAGEMENT except New South Wales is to simply multiply the "corrected" population estimate for grey Aerial survey should be as precise and as kangaroos from fixed-wing surveys by 2 (South accurate as possible. However, precision was Australia and Queensland) or 2.5 (Western considered more important than accuracy if Australia). There was also discussion about one has to be sacrificed, because being able whether it was appropriate to have correction to track density provides the best tool for factors for individual habitats or regional management. Ideally, the monitoring method correction factors that would be applicable used must accommodate the trend to regional in all habitats (Pople 1999), with the latter management and allow local population receiving most support. modelling. In most states, aerial surveys provide estimates of kangaroo density for After a lot of discussion and review of extensive areas. Subdividing these areas to available information (Table 1, Cairns 1999), provide regional estimates wonld result in it was agreed that, pending the outcome of less precise estimates unless sample sizes further work to be completed in New South were increased or the sampling strategy Wales in the near future, the following changed. The costs in doing so are likely to correction factors for grey kangaroos would be be prohibitive. adopted in all habitats: Surveys should have broad applicability 4.8 for Western Greys, in the sense of allowing information to 3.5 for Eastern Greys and, be translated comparably between regions, 3.5 for both species in New South Wales regardless of the method by which the (because they cannot be readily distin- estimates of populations are made. This does guished from the air), representing a not necessarily mean that there should be conservative correction factor. more commonality of methods between states. The point was made strongly that regardless It was also agreed that existing correction of the method used in different states or factors wonld continue to be used for red different regions it was important to strive for kangaroos (i.e., 2.29 for open vegetation similar output. For example, a density of nine cover, 2.36 for light cover, 2.43 for medium red kangaroos per square kilometre in South cover and 2.57 for dense cover). The current Australia should be equivalent to a density of practice of multiplying counts of both red nine red kangaroos per square kilometre in and grey kangaroos by ll(1.474-0.0316T), Western Australia. where T is the temperature rC) at survey height, for temperatures >15'C, should also be continued. CORRECTION FACTORS STRIP WIDTH With a three year study beginning in New South Wales (Gilroy 1999) and further work There is some interest in moving from proposed in Queensland (Lundie-Jenkins et al. sampling in a 200 m strip to a 100 m strip. 1999), both directed at answering certain Some data exist on the bias and repeatability questions about aerial survey methodology, of the narrower strip (Pople et al. 1998a; it was too soon for the workshop to identify Pople 1999) which suggest that while bias is or recommend any permanent change in reduced, repeatability is poorer. Further work correction factors. However, interim changes on this is planned in New South Wales. (but conservative ones) could be recommended A concern that has frequently been raised is pending the outcome of the Queensland that to shift to a new method risks the June 1999 Ausfralian Zooiogist 31(1) 317 Tdlc I. Correction factors developed Frum previous studies at a range of locations in eastern Ausualia. For each location, the nearest major town is shown. Dashes indicate the species war either absent or at a density too low to be estimated. All correction factors have been developed for 200 m strip uansect counts using the same aerial survey method. Open teferr to grassland and open shrubland; wooded refers to vegetation ranging from low open woodland to woodland, and includes mulga tall shrubland. Red Eastern grey Common Western grey Location (State) Habitat kangaroos kangaroos wallaroos kangaroos - Longreach (Q!' Open 1.7 10.2 3.8 - - W~ndarah( Q) Open 3.1 9.5 Blackall (Q)' Openlwooded 2.6 5.0 4.1 - - Charleville (Q)' Wooded 3.1 5.1 4.8 - - Bollon (Q)' Wooded 1.8 4.0 Roma (Q)' Wooded 0.7 3.4 - - - - - Goandiwindi (Q)' Wooded 7.6 Tiboohurra (NSW)P Open 1.8 - - - - - - Jerilderie (NSW)' Open 2.3 - - - Wilcannia (NSW)' Wooded 2.4 Menindee (NSW)' Open 1.8 - - 4.8 - - Menindee (NSW)' Wooded 2.8 16.7 Menindee (NSW)& Open 2.5 - 11.1 5.9 - - Quilpie (Q)6 Open 1.8 3.9 Quilpie (QY Wooded 4.2 - 23.3 - - - Cunnamulla (Q)' Wooded 3.8 3.5 Western Australian Ooen 2.3 - - - 'Pople 81 01. (1998h). ?Bailey (1971). 'Caughley cf nl. (1976). 'Short and Baylirs (1985). IShort and Hone (1988). 6Southwell d ol. (1986). 'Grice d a1 (1990). 8Southwell (1993). continuity of the long-term population data of aerial survey datasets, and there have for kangaroos that has been collected on already been two assessments. In the Queens- aerial surveys. If all methods are essentially land survey blocks, Pople et al. (1993) certainly striving to estimate the same, true population found counts on the southern side were more density, then having a more precise and accurate, hut not consistently more precise accurate method is surely desirable. Reworking than counts on the northern side of the ~reviouse stimates of population size with aircraft. Similarly, the bias of counts on the revised correction factors would be required southern side (determined from a comparison for grey kangaroos. So long as this was done, of fixed-wing and helicopter counts) was not either shifting to a new method or new set of less variable than the bias of counts on the correction factors, it should not compromise northern side. Across a much larger area in the accuracy of the long-term population trends. Western Australia, Southwell (1993) found the Obviously it is important to demonstrate that southern side counts were both more accurate any new technique results in improved and more precise than counts on the northern accuracy and particularly repeatability, before side. Because there was only a slight loss in bringing it into widespread use. Clancy (1999) precision in using only the southern side provided a strong case justifying the use of counts over using counts from both sides, helicopters to survey kangaroos in Queensland. Southwell (1993) argued that counting only on the southern side was worth adopting. SIDE-OF-AIRCRAFT Historical data could be reanalysed using only the counts from the southern side, because suggested change was restrict side-of-aircraft is always recorded, "Southern counting to the southern side of the aircraft side,, correction factors would need to be where visibility is generally better and the determined and have been reported by Pople required correction factor would he smaller. et al. (1993) and Southwell (1993). The other advantage is that only a single observer would be ;equired, provibing some saving in labour costs, although these costs INDIRECT MONITORING are likely to be small relative to the running A considerable amount of data relating to costs of the aircraft. Furthermore, using only the harvest are collected every year by all one observer effectively reduces the sampling states (e.g., Lundie-Jenkins et al. 1999). These intensity (in terms of area) by half, although databases are largely unexplored. They offer the actual number counted would he reduced the opportunity to determine, by correlation by less than half. with lone runs of aerial survev data. whether The effect of counting only on the southern or not ;hiller returns and other measures side could be tested with any one of a number of harvesting activity provide an alternative 318 Australian Zoologist 31(1) June 1999 means of monitoring the status of the a survey. Constant speed and height, f)opulation. \Vhe~ei nf&-marion on population accurate flying and efficient flight planning arc stnlctlrre is available. I)etrer inrerr)rctation of fixed transect lines, conduct of surveys 07 the data is possible.'Any change's to the in appropriate weather conditions and composition of the population resulting optimizing observer comfort will be best from food availability or harvesting may be achieved with pilots familiar with surveying reflected in the harvest data. Such data are wildlife populations. Finally, there is large supplementary to, rather than in place of, advantage in working with aircraft charter direct monitoring. They offer the potential of companies aware of the need for flexible continuous monitoring of a population and timing of surveys, largely because of weather monitoring in areas not surveyed directly. A restrictions, and prepared to provide recently number of assumptions would need to be serviced aircraft equipped with clear windows satisfied for any harvest parameter to be an and a radar altimeter. effective index and Southwell (1989) discusses possible violations of these assumptions. These SAFETY ISSUES include standardizing harvesting equipment, Communications towers, electric and other efficiency and conditions, and that the wires represent a threat. It would be possible harvesting of each animal is independent and to have all of these identified in a Global that animals do not learn to avoid shooters. Positioning Receiver (GPS) database which is regularly updated, and which would alert a OBSERVER TRAINING pilot when there are hazards in the vicinity. Beard (1999) provided the first extensive, This might he something for the future. quantitative look at variation among experi- Survey managers and observers should check enced and novice observers. Until now, most with their pilot that he/she is taking guidelines for training observers had been appropriate steps in relation to Search and based on anecdotal reports. Consideration of Rescue (SAR) requirements. This usually takes observer differences had been restricted the form of a flight plan submitted to the largely to single survey comparisons of the aviation authorities or, these days, can be total counts of left and right side observers. as simple as leaving a "Flightnote" which After some discussion it was agreed that all includes flight details and an estimated time teams conducting aerial surveys of kangaroos of return with a person on the ground. The should adopt the concept of observers having need to check that a pilot has a low flying a substantial period as a trainee, including endorsement was identified. There was also tracking of their performance against an discussion about the limitations posed by experienced (i.e., trained) observer. This has restrictions on pilots in the number of hours the opportunity to build confidence in an they can fly on a day, the number of days observer by the survey manager and will give without a break, and this has implications for more confidence in the value of the data. survey planning. There seemed to be agreement that about There have been a lot of new develop- fifty hours in the aircraft was necessary until ments in safety-related items of dress for you could reasonably expect to use data from people carrying out aerial surveys from a particular observer and that this should helicopters, including helmets, survival gear, comprise at least ten survey sessions. There is and suits in bright orange colours. All also probably a need for checking trained aerial work conducted by Conservation and observers and this can be done by comparing Land Management (CALM) in Western left and right observers on a total survey, Australia is covered by occupational Health by comparing their totals. It was agreed that and Safety regulations and this is the case in it was appropriate to continue the present other states as well. Most of the regulations practice of recording the names of the are requirements of the Air Navigation observers in reports and publications. Main- Regulations. taining and being aware of a pool of trained observers was important in keeping the MINIMUM WEATHER CONDITIONS technique truly standardized. FOR SURVEY A related issue is the importance of using Work by Graeme Caughley, David Grice and pilots familiar with broad-scale, low level others showed the possible need to correct aerial survey and working with aviation aerial survey data for air temperature, so this companies sympathetic to the needs of such should be recorded routinely on all surveys surveys. Essentially, pilots are part of the and data corrected where appropriate. The aerial survey team. It is inefficient to regularly question about the likelihood that kangaroos have to train naive pilots at the start of undergo metabolic acclimatization, however, Australian Zoologist 31(1) 319 was raised. Different temperature correction REFERENCES factors may need to be developed both Bailey, P. T.. 1971. The red kangaroo, Mcgaliin mfn regionally and seasonally. There was some (Desmarest), in northwestern New South Wales. I. discussion about the width of the seasonal Movements. CSIRO Wildl. Re*. 16: 11-28. window in which it was appropriate to do Beard, L. A,, 1999. Training observers. Aust. Zool. 31: aerial surveys. The window is wider in the 287-91. southern parts of the continent than in the Cairns, S. C., 1999. Accuracy and consistency in the north. aerial survey of kangaroo populations. Aut. Zool. 31: 275-79. Concerning conditions for sighting kangaroos, the view was offered that cloud cover was less Caughley, G., Sinclair, R. and Scott-Kemmis, D., 1976. relevant than brightness (i.e., light level), and Experiments in aerial survey, J. Wildl. Manage. 40: 290-300. it was therefore appropriate for people to avoid conducting surveys in dull conditions Clancy, T F., 1999. Choice of survey platforms and and to record information about the weather technique for hroad-scale monitoring of kangaroo populations. Aurt. Zool. 31: 267-74. on data sheets. It was agreed that there would always be an element of judgement Gilroy, J., 1999. Kangaroo monitoring in relation ta in the decision to survey or not, but, like- kangaroo management in New South Wales. Awl. Zool. 31: 30608. wise, agreement that surveys should not be undertaken in conditions of poor visibility. Grice, D., Caughley, G. and Brown, B., 1990. Accuracy of aerial surveys. Unpublished report to the The issue of window haze was also discussed Australian National Parks and Wildlife Service, briefly. It is a common pmhlem in old aircraft, Canberra. 32 Pp. but windows can be replaced. Lundie-Jenkins, G., Hoolihan, D. W. and Maag, G. W., 1999. An overview of the Queenrland macropod CONCLUSION monitoring programme. Aust. Zool. 31: 301-05. No monitoring technique will ever be perfect, McCallum, H. I., 1999. How to count kangaroos. Aust. Zool. 31: 309-16. hence the continued refinement of monitorhg techniques for kangaroos. McCallum (1999) Pople, A. R., 1999. Repeatability of aerial surveys. Awl. ZooL 31: 280-86. offered a number of suggested directions - for further improvement. Invariably, what Pople. T, Cairns, S., Clancy, T., Grigg. G. and Beard, L., ~. methods are used represent a balance between 1993. Accuracy of kangaroo surveys in Queensland using fixed-wing aircraft. Unpublished report to the cost, accuracy and precision. In the past, Australian National Parks and Wildlife Service, the costs of surveys have been borne mostly Canberra. 75 Pp. by the relevant state government department, Pople, A. R.. Cairns, S. C., Clancy, T F., Grigg, but an increasing proportion of these costs G. C., Beard, L. A. and Southwell, C. J., 1998a. An are now being met by the kangaroo industry, assessment of the accuracy of kangaroo surveys using usually in the form of a tag levy. The need fixed-wing aircraft. Wildl. Rcr. 45: 31S26. to put a figure on the size of Australia's Pople, A. R., Cairns, S. C., Clancy, T. F., Grigg, kangaroo population ( i . , an accurate G. C., Beard, L. A. and Sou-th wcll, C. -J ., 1998b. estimate), at least that proportion that is Comoarison of survevs of Lanearoos in Oucensland exposed to commercial harvesting, and to using helicopters and fixed-wing aircraft. Rangel. J. 40: 92-103. track it through time (i.e., a precise estimate) is important for effective management. This Short, J. and Bayliss, P., 1985. Bias in aerial survey estimates of kangaroos. J. Appl. Ed. 43: 559-71. is particularly so, as quotas, which are set as a proportion of the population, are the Short, J. and Hone, J.. 1988. Calibrating aerial surveys main form of regulation. As quotas are now of kangaroos by comparison with drive counts. Aurt. Wildl. Rex. 15: 277-84. being reached in many regions of Australia, there is increased pressure on managers from Southwell, C. J., Fletcher, M. and Scotney, T., 1986. the kangaroo industry, the grazing industry Improving the aerial survey method for monitoring kangaroo populations in Queenrland: results of and conservation groups to have accurate preliminary investigations. Unpublished report and precise estimates of population size. prepared for the Queenrland National Parks and However, the need for regular, direct Wildlife Service. Brisbane. 60 Pp. monitoring is also political. Being one of Southwell, C., 1989. Techniques for monitoring the largest harvests of terrestrial wildlife the abundance of kangaroo and wallaby populations. in the world and with a growing export Pp. 659-93 in Kangaroos, Wallabies and Rnt- trade, much attention is directed at the Kangaroos, Vol. 2 ed by G. Grigg, P. Jarman and I. Hume. Surrey Beatty & Sons: Chipping Norton, industly. It is important then, that it is well Sydney. managed. This also makes kangaroo manage- ment a benchmark for the growing number of Southwell, C., 1995. Report on the 1993 aerial survey of kangaroos in Western Australia. Unpublished report sustainable wildlife use operations elsewhere to the Australian Nature Conservation Agency, in the world. Canberra. 77 Pp. June 1999