Little has been published about the safety of darting wild primates, although this is a concern for many primatologists. Our report on primate darting (Cunningham et al. 2015) consisted of a literature review examining trends in the reporting of darting methods and results, and the results of two anonymous surveys of primatologists that aimed to gather information on darting methods and their effect on the primates involved. Our study concluded with a call for more detailed reporting of darting methods given the wide range of challenges primates present to a darting program due to their variation in size, habitat, and behavior (Cunningham et al. 2015). We are grateful to Fernandez-Duque and colleagues for continuing the discussion of this important issue with a commentary on our article (Fernandez-Duque et al. 2017) and are pleased that they also recognize the need for detailed reporting, echo our call for greater openness, join us in bringing attention to the challenges the diversity of primates bring to darting, and share our ultimate goal of making the capture of primates safer.

Although the findings of the article and the commentary are broadly similar, and some of the differences in methods are due to differences in the aims of the reports, other differences merit careful and thoughtful evaluation. Our comments focus on 1) likely bias in the data, 2) reporting of serious complications that arise during capture, 3) reporting results of darting medically compromised primates, and 4) reporting rates of primates falling to the ground. We also explain additional differences between our analyses (Cunningham et al. 2015) and those in the commentary of Fernandez-Duque et al. (2017).

Likely Bias in Data

The primary goal of our literature review (Cunningham et al. 2015) was to analyze trends in the reporting of darting methods and results. We conducted a literature search of complete texts of articles from 1940 to 2013 with words related to darting (Cunningham et al. 2015). The analysis reveals the scarcity of published data: only 18 of 111 articles provide comprehensive information on darting methods and results. An additional 11 articles provide the rate of fatalities and serious injuries but do not provide details of darting methods. Eighty-two articles (74%) do not indicate whether fatalities or injuries occurred. In addition to this primary analysis, we summarize the causes of fatalities and serious injuries and complications reported in the 18 articles that provide comprehensive information. We state: “The results of our literature review and surveys cannot give an accurate rate of overall injury rates and death, but suggest a 5% rate of serious and fatal injuries and complications” (Cunningham et al. 2015, pp. 923–924), thus acknowledging the limitations of our analysis.

Fernandez-Duque et al. (2017) reanalyzed the 18 articles in the literature review that provided comprehensive information and, in addition, analyzed 11 articles that indicate whether fatalities or serious injuries occurred but offer no details of methods. They found a mortality of 2.5%, an injury rate of 1.5%, and a combined rate of 4.0%. Fernandez Duque et al. then compared their mortality estimate alone with the combined mortality and injuries rate in Cunningham et al. (2015), and concluded that their reanalysis yielded lower percentages than the earlier study. However, comparing mortality alone (i.e., comparing like with like) reveals no substantial difference between the two studies, given the limitations of the data (2% in Cunningham et al. 2015, 2.5% in Fernandez-Duque et al. 2017).

Alhough there is no way to know the rate of injury in the 74% of papers that did not provide information on fatal and serious injuries, people tend to provide information that will be viewed positively by others (Zuber and Kaptein 2014), and are likely to withhold negative information when self-reporting is voluntary (Friesena and Gangadharan 2013). Extrinsic factors may also play a role in what information is reported (Sakamoto et al. 2013). For example, primatologists have spoken to us about their concern that revealing high rates of injury and mortality will make it more difficult for them to obtain the permits and approvals necessary to conduct their research. For these reasons, we believe that the rates reported in Fernandez-Duque et al. (2017) and Cunningham et al. (2015) are probably lower than the actual rates of serious injuries, complications, and mortality. Fernandez-Duque et al. (2017) do not discuss this likely bias in the data, but we believe acknowledging the limitations of the data is an important step in finding ways to encourage effective self-reporting.

Reporting of Serious Complications during Capture

Our analysis included fatalities, serious injuries, and serious darting-related complications. We considered serious complications and injuries those as that had the potential to affect survival or reproduction or required expert care (Cunningham et al. 2015). In contrast to our analyses, Fernandez-Duque et al. (2017) excluded darting-related complications, in particular 13 cases in which primates needed artificial respiration for up to 45 min (Scott et al. 1976). Neither did they include these cases in their calculation of the injury rate, although they referred to them in their discussion of the injury rate and they included two primates in the same report that died while under artificial respiration. We believe that analysis of serious complications will help us improve the safety of darting methods.

Reporting Results of Darting Medically Compromised Primates

Fernandez-Duque et al. (2017) excluded two papers (Hyeroba et al. 2011 and Sleeman et al. 2000) from their analysis because compromised health was a “conflicting factor in analyzing the injury and mortality risk of darting” (p. 5). As neither article reported injuries or fatalities due to darting, the health of the primates in these papers did not bias the data in Cunningham et al. (2015).

We agree, however, with Fernandez-Duque et al. (2017) that factors such as poor health, age, and reproductive status may increase complications, injuries, and mortality (Cunningham et al. 2015). One of the principal challenges of field anesthesia is the preanesthetic assessment of the animal. Conditions such as cardiac disease or liver disease, which are not readily observed in a free-ranging primate, can exacerbate stressors common to chemical immobilization (sites in Cunningham et al. 2015). Primates with unknown preexisting medical conditions should be included in calculations of injury and mortality, as they were in Cunningham et al. (2015) and Fernandez-Duque et al. (2017). We believe it is important that they continue to be included in calculations of serious complications and injuries and fatalities. Cases in which a primate is captured to provide help for a known preexisting injury or medical condition should also be reported. They could be treated, however, as a special circumstance and reported separately from darting in which the medical condition of a primate is unknown, or the primate is darted for reasons other than providing help.

Reporting Rates of Primates Falling to the Ground

Fernandez-Duque et al. (2017) included fall-related injuries in their rate of serious and fatal injuries. However, they stated: “We considered that the risk of darting is best assessed by quantifying the number of animals injured due to actual darting, rather than using falling as a metric for risk” (p. 4). In contrast, we believe information on the rates of catching falling animals can help primatologists assess risk associated with a darting protocol, as injuries from falls were the main cause of death in the results of our surveys and one of the leading causes of fatal and serious injury in our literature review (Cunningham et al. 2015). The risk of injuries from falls is greater for arboreal primates (Cunningham et al. 2015), and analysis of injuries from falls should distinguish between arboreal and terrestrial primates.

Reporting the Same Darting Events

The analysis of Fernandez-Duque et al. (2017) also differs from ours in that they excluded articles that reported on the same darting events. This is appropriate for their aim to calculate the rate of serious injuries and fatalities due to darting. Our primary aim was to identify trends in the reporting of darting methods and results, so we did not exclude papers that might describe the same darting events because we aimed to consider the level of detail offered in every paper. We stated, “As we are interested in general trends, rather than precise numbers, we did not attempt to exclude papers that might be describing the same darting events” (Cunningham et al. 2015, p. 213). We were aware that two articles included in our estimation of injury and death rates (Fernandez-Duque and Erkert 2006 and Fernandez-Duque and Rotundo 2003) reported the same darting events. Including data from both papers did not change the rate we reported, which was estimated to the nearest percent. Additional articles that Fernandez-Duque et al. (2017) excluded because they reported on the same events were not included in the Cunningham et al. (2015) estimates of serious and fatal injuries because we judged that they did not provide enough information. We agree with Fernandez-Duque et al. (2017) that calculations of injury and death rates should not include multiple reports of the same events.

Conclusion

We appreciate the opportunity presented by Fernandez-Duque et al. (2017) to further the discussion of issues related to primate capture. Such discussion is an essential step in refining methods. To further this goal, the International Primatological Society has created an ad hoc committee to address issues concerning the capture and release of wild primates (Setchell 2017). We serve with E. Fernandez-Duque on the committee and are working together to make primate capture safer.