School of Veterinary Science, The University of Queensland, Gatton, Queensland, 4343, Australia.
Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Parkville, Victoria, 3010, Australia.
Sci Rep. 2019 Jul 11;9(1):10037. doi: 10.1038/s41598-019-46376-5.
The active collection of wildlife sighting data by trained observers is expensive, restricted to small geographical areas and conducted infrequently. Reporting of wildlife sightings by members of the public provides an opportunity to collect wildlife data continuously over wider geographical areas, at lower cost. We used individual koala sightings reported by members of the public between 1997 and 2013 in South-East Queensland, Australia (n = 14,076 koala sightings) to describe spatial and temporal trends in koala presence, to estimate koala sighting density and to identify biases associated with sightings. Temporal trends in sightings mirrored the breeding season of koalas. Sightings were high in residential areas (63%), followed by agricultural (15%), and parkland (12%). The study area was divided into 57,780 one-square kilometer grid cells and grid cells with no sightings of koalas decreased over time (from 35% to 21%) indicative of a greater level of spatial overlap of koala home ranges and human activity areas over time. The density of reported koala sightings decreased as distance from primary and secondary roads increased, indicative of a higher search effort near roads. Our results show that koala sighting data can be used to refine koala distribution and population estimates derived from active surveying, on the condition that appropriate bias correction techniques are applied. Collecting koala absence and search effort information and conducting repeated searches for koalas in the same areas are useful approaches to improve the quality of sighting data in citizen science programs.
专业观察员主动收集野生动物目击数据的成本很高，只能在小地理区域进行，且频率较低。公众报告野生动物目击事件为在更广泛的地理区域内持续收集野生动物数据提供了机会，成本更低。我们使用澳大利亚东南部昆士兰州（1997 年至 2013 年）公众报告的 14076 只考拉目击事件，描述了考拉存在的时空趋势，估算了考拉目击密度，并确定了与目击事件相关的偏差。目击事件的时间趋势与考拉的繁殖季节相吻合。在居民区（63%）的目击事件较多，其次是农业区（15%）和公园区（12%）。研究区域被划分为 57780 个一平方公里的网格单元，没有考拉目击事件的网格单元随着时间的推移而减少（从 35%降至 21%），这表明考拉的栖息地和人类活动区域的空间重叠程度随着时间的推移而增加。报告的考拉目击密度随着与主要和次要道路的距离增加而降低，这表明道路附近的搜索工作更为密集。我们的研究结果表明，在应用适当的偏差修正技术的前提下，可以利用考拉目击数据来改进基于主动调查的考拉分布和种群估计。收集考拉缺失和搜索努力信息，并在同一地区对考拉进行重复搜索，是提高公民科学项目中目击数据质量的有效方法。
