Response to Roberts, Stenglein, Wydeven, and others
2024
Treves, Adrian | Agan, Suzanne W. | Langenberg, Julia A. | López-Bao, José V. | Louchouarn, Naomi X. | Parsons, Dave R. | Rabenhorst, Mark F. | Santiago-Ávila, Francisco J. | Treves, Adrian [0000-0002-3052-4708]
Human-caused mortality has been the major cause of death among wolves worldwide. In 2017, we summarized a 33-year data set of >933 Gray Wolf deaths from Wisconsin, United States, and estimated that poaching was the major source of mortality. Roberts et al. (2024; hereafter, just Roberts et al. unless using a direct quote from that paper) challenge our reinterpretation of data on causes of death and disappearances and urge us to use standard known-fates survival models rather than the combined time-to-event and total accounting methods that we used. They do not cite subsequent time-to-event and competing risk and incidence models that we published, raising an issue of selective citation of only their own work. Regarding reinterpretations, Roberts et al. neither present evidence for their claims nor revisit records of cause of death to argue their claims. Regarding traditional known-fate survival models, we review the violation of a critical assumption of such models. Namely, causes of death were not independent of censoring among Wisconsin collared wolves. Rates of disappearance approximating 42% of all collared animals are incompatible with the assumption that unknown-fate collared wolves died of the same causes as known-fate animals. We demonstrate that Roberts et al. made an erroneous claim that wolves frequently outlive the operational lives of their VHF collars. We present evidence of undisclosed competing interests among Roberts et al.'s coauthors. In scientific debates, the most transparent assumptions, methods, and data prevail because outside reviewers can judge for themselves. We stand by the conclusions of our combined analyses from 2017 to 2023.
显示更多 [+] 显示较少 [-]Peer reviewed
显示更多 [+] 显示较少 [-]