{"id":70655,"date":"2023-09-21T07:29:00","date_gmt":"2023-09-21T10:29:00","guid":{"rendered":"https:\/\/www.ufrgs.br\/jornal\/?p=70655"},"modified":"2023-12-06T11:00:32","modified_gmt":"2023-12-06T14:00:32","slug":"drones-can-help-in-population-estimation-and-monitoring-ofendangered-animals","status":"publish","type":"post","link":"https:\/\/www.ufrgs.br\/jornal\/drones-can-help-in-population-estimation-and-monitoring-ofendangered-animals\/","title":{"rendered":"Drones can help in population estimation and monitoring of endangered animals"},"content":{"rendered":"\n

Environment | In addition to assessing the distribution of the marsh deer using unmanned aerial vehicle (drones), a doctoral thesis discusses how drones can be used to aid research in the field of Ecology<\/h4>\n\n\n\n

*By Geovana Benites
*First published May 3th, 2023
*Photo: Personal File<\/em><\/p>\n\n\n\n

Assessing the abundance of animal groups is one of the essential aspects for the evaluation of extinction risks and detection of population declines of endangered species. In his doctoral disser<\/a>tation defended in the Graduate Program in Ecology of the Institute of Biosciences at UFRGS, researcher Ismael Verrastro Brack studied the use of drones to estimate populations and track endangered animals. In addition, the research evaluated the factors that affect the spatial distribution of the marsh deer, the largest cervid in South America.<\/p>\n\n\n\n

Drones have come up as a groundbreaking tool in the field of ecology, offering a fresh and promising approach to sampling and analyzing population data, emerging as an affordable and safe alternative compared, for example, to research done with manned aircraft. Conducted under the supervision of Professor Luiz Flamarion Barbosa de Oliveira, Brack’s doctoral dissertation has been released in essay-format publications that demonstrate how to plan a sampling with drones to estimate the abundance of wild animals and the applications of these techniques to assess factors affecting the spatial distribution of marsh deer.<\/a><\/p>\n\n\n\n

Observing from above<\/h5>\n\n\n\n

The utilization of drones is rapidly becoming a promising avenue for future wildlife tracking studies because they allow sampling with high spatial resolution and high frequency of flights. In addition, the tool allows for permanent records, in which researchers can review the images collected. “Drones initially emerged as a tool that could replace airplanes for counting, since many decades ago they were used to count various animal species, especially large mammals,” explains the researcher.<\/p>\n\n\n\n

At the beginning of his master’s studies, in 2014, Brack already had the intention to work with drones for sampling and counting animals, but the device was still new on the market and, therefore, had many limitations. “[In the Ecology Department] We had a lot of problems with the drones, because the one we bought performed poorly, it got damaged very easily, it was impossible to make long flights with it, limitations which made the process difficult,” he says. Later, when Brack was doing his PhD, the department bought a drone from another manufacturer, and so the equipment had already acquired a growing role when the opportunity for the scientist to work with the sampling of endangered cervids in open areas of South America arose.<\/p>\n\n\n\n

A species on the verge of extinction<\/h5>\n\n\n\n

In 2017, the marsh deer was the species chosen to estimate the abundance of the animal populations inhabiting the northern Pantanal. The animal is the largest cervid in South America, weighing up to 150kg (330lbs), and is a species that usually lives in wet areas, inhabiting regions of swamps, marshes, and floodplains. Lower densities of marsh deers can also be found in grasslands and savannah environments, where they feed on sprouts, new leaves, and flowers.<\/p>\n\n\n\n

Currently, the species is classified as endangered due to modifications of their habitat \u2013 mainly those related to agriculture, hydroelectric dams, illegal hunting, and diseases caused by domestic cattle. Even the largest deer populations, located in the Pantanal, have been threatened and, as a consequence of global climate change, even intensified by the increased frequency of extreme weather events \u2013 such as severe droughts and wildfires.<\/p>\n\n\n\n

\n

\u201cEstimating population size and assessing space requirements of animal populations is very important to predict possible population declines\u201d<\/em><\/p>\nIsmael Verrastro Brack<\/cite><\/blockquote>\n\n\n\n

With the use of drones, the researcher analyzed the images collected and proved that the local abundance of the marsh deer holds a strong connection to areas that are taken as \u2018high quality for the species\u2019: high levels of vegetation greenery, i.e., forage supply for the species, and with permanent water reservoirs, but not with the occurrence of its main predator, the jaguar. Core areas for the population were also identified in the remaining humid environments and even near the artificial water reservoirs of the driest region.<\/p>\n\n\n\n

Planning First<\/h5>\n\n\n\n

From this study, it was found that methodological questions about planning a sampling with drones to estimate animal abundance are of great importance for the outcome of research. “How can we use this tool to count animals and estimate the population size? How does this number of individuals vary in space and time?”, asks Brack. “So I have explored a lot of analytical tools to treat these counts so that we can estimate the errors that can occur in this process,” he adds.<\/p>\n\n\n\n

In the study, the researcher pointed out that counts made from drone sampling can be altered by four main types of error detection. The first is the availability error<\/strong>, that is, when there are animals present in the sampling area, but unavailable to be detected because they are hidden. The second is perception error<\/strong>, when the collected images, which are analyzed manually or with the help of an algorithm, fail to detect an available individual in the images. Double counting<\/strong> can also occur, due to animals moving between flight lines or overlapping images. In addition, misidentification<\/strong> is also common at the time of analysis, as the observer may mistake vegetation or an animal of another species for an individual of the species of interest. According to Ismael, the optimization of the sample design, i.e., the search planning taking these variables into account, is relevant and became, after the study with the marsh deer, a crucial step to improve the accuracy of the estimate.<\/p>\n\n\n\n

Brack explains that the tools and techniques explored are already serving as a basis for other studies in the Department, expanding the research to other populations and species. “Having established a comprehensive protocol and a clear understanding of the sampling methodology for species monitoring, we have now broadened our scope to encompass other endangered deer populations across South America,” he adds.<\/p>\n\n\n\n

Furthermore, one of the next steps related to the study is the monitoring \u2013 carried out in the Reserva Particular do Patrim\u00f4nio Natural <\/em>(RPPN) (Private Natural Heritage Reserve) Sesc Pantanal, in Mato Grosso \u2013 of the marsh deer after the large fires occurred in the region in 2020, when more than 90% of the reserve area was burned. “We have been doing annual population surveys to compare with the estimates we had prior to the fires and assess the impact of these catastrophic events for this endangered species,” says Brack.<\/p>\n\n\n\n

\n