Skip to main content

How to steal food from predators – behavioural strategies of a scavenger

Periodic Reporting for period 1 - FFP-BSS (How to steal food from predators – behavioural strategies of a scavenger)

Reporting period: 2019-06-01 to 2021-05-31

The ecological relationships between scavengers, predators and their prey can substantially influence their behaviour and population dynamics. Still, the impact of scavenging, i.e. the consumption of carrion, has been largely underestimated in ecosystems. In this project we studied one of the Northern hemisphere’s most widespread predator-scavenger systems: wolves (Canis lupus) and ravens (Corvus corax), which likely have a co-evolutionary relationship.
Our research has been carried out in Yellowstone National Park, where wolves have been reintroduced in a rewilding project 25 years ago. Since then, these top-predators have been intensively monitored with numerous studies on behaviour and ecology of this top-predator but also on its effects on other species in the ecosystem. Previous studies on ravens, the most abundant scavengers in this area, were based on the observations and counts of unmarked individuals at carcasses and anthropogenic food sources revealing processes on the population but not on the individual level. However, both, population level effects as well as individual strategies, are important to fully understand the role of each species in an ecosystem. Particularly now, where human pressure threatens all ecosystems around the world, it is highly important to better understand such relationships for nature protection and restoration.
The main aim of our study was to identify which foraging and scrounging strategies ravens follow to locate and exploit wolf kills as well as other food sources. Different hypotheses such as opportunistic foraging, a scout strategy (ravens continuously follow predators), a turn-taking strategy (ravens switch between scout and opportunistic strategies) and effects of experience (mostly by age) have been tested. For this we GPS-tagged more than 70 ravens within or close to Yellowstone National Park since October 2019 and compared their GPS-tracks with the movements of GPS-collared wolf packs but also as comparison with GPS-collared pumas and the respective kill sites of both species. Further, we started an extensive resource mapping to include other food sources such as winter-killed animals and in particular anthropogenic food sources.
Surprisingly, the most important resources were of anthropogenic origin such as landfills, sewage ponds, hunting areas that are often more than 100km distant to the National Park, indicating that these resources also have a strong influence on the protected ecosystem. Analysing the exploitation of natural carcasses indicates age effects with adults using this food more often than juveniles, but we could not confirm any of the other hypotheses. Even more surprising, ravens frequently approached fresh carcasses from distances far beyond their perception range which challenge current hypotheses for the transfer of information about the location of food sources. We suspect that ravens do not only learn and remember the location of predictable anthropogenic food sources, but also the areas where they can frequently find carcasses such as wolf territories.
A major effort in this project was trapping a large number of ravens. However, during several months of field work we were successful in trapping and GPS-tagging more than 70 individuals. GPS-data are collected from sunrise to sunset at 30min intervals or hourly when the solar powered battery is not fully charged. In addition every 15 min acceleration (ACC) data are collected with 33Hz over 4 seconds. So far, more than half a million GPS positions and more than 1.5 million ACC data have been collected and downloaded via base stations in the field or cell network where available. Further, we observed the behaviour of ravens at different food sources and identified around 3000 foraging sites (resource mapping). Together this leads to a highly valuable and unique dataset for a series of behavioural and ecological questions.
During a period of 14 months GPS-tagged ravens exploited 43% of the 201 known wolf kills and 25% of the 91 known cougar kills. For their first visit, ravens approached the kill sites on average from 15 km distance to their previous night roost, in some cases even from up to 155 km on a highly directional trajectory. Such distances are likely far beyond ravens’ perception range and challenge current hypotheses for the transfer of information about the location of food sources. Surprisingly, however, the Yellowstone ravens used primarily anthropogenic food sources often 50-100 km away from their territory or trapping site. Overall ravens moved much more than previously known, showing a far larger scale of connectivity of resources than previously known. Two non-breeders even moved from Yellowstone National Park to Canada, which resulted in the longest recorded dispersal distances for ravens (757 km and 745 km). The analysis is still ongoing and we expect a series of new findings over the next months.
The results have been presented to scientific audiences at conferences and invited seminars in Germany, Switzerland, USA, (upcoming talks this fall in UK and USA) and to public audiences in Germany and USA (upcoming talk this fall in Austria). A first perspective piece on raven management is under review and will be of high interest to wildlife management authorities in North America and beyond. Two papers are planned to be submitted to peer-reviewed journals in fall 2021 followed by at least two other paper submissions until summer 2022. For the public audience we published one article in Yellowstone Science, a journal devoted to sharing information about Yellowstone's natural and cultural resources, and another one in the journal/blog Closer by Swarowski Optik. Further there will be a chapter in the first book about birds of Yellowstone National Park which is expected to be published in 2022.
This project is one of the first combining GPS-tracking of predators and scavengers within the same area. Further, we added resource mapping and behavioural observations and thereby created a unique multidimensional dataset that will improve our understanding in food-web and community ecology. The analysis of how ravens adjust their movements to those of terrestrial mammals also leads to analytical challenges and thereby advances the emerging field of movement ecology.
Our results about how ravens exploit different food sources open up new questions about their cognitive abilities regarding spatial memory of predictable but also less predictable food sources (wolf kill sites). The large scale movements across hundreds of kilometres mostly between anthropogenic food sources show a great potential for transmission of information but also of seeds and possibly diseases, which need further investigation. With almost 9000 km² Yellowstone National Park is a relatively large protected area, however, the intensive use of anthropogenic food sources by ravens often far away from the park, likely leads to a higher ravens density compared to natural conditions and shows how much the area surrounding protected areas affects its ecosystems. Another consequence of these long distance movements is that local management of ravens by culling to reduce their numbers (as often carried out in many parts of the world) is relatively useless, ethically highly questionable and instead a reduction of anthropogenic food should be considered.
Ravens and wolves feeding on a wolf-killed elk
GPS-tracks of 70 ravens; red crosses indicate trapping sites, green crosses their latest position