While a lot of attention is paid to reporting on human warfare across the globe, the illegal war on wildlife that is carried out through trafficking, poaching, and bush meat trades is often given comparatively less coverage. (At least in U.S. mainstream media, to which I refer here.) I’ve been starting to pay more attention to wildlife trafficking issues for a few reasons. One being a story I wrote recently for Birder’s World magazine (you’ll have to wait until February to read it!) where I interviewed a Florida Fish and Wildlife Conservation Commission agent who told me about people smuggling Cuban bullfinches into Miami. He said the smugglers give the birds Valium to knock them out, then stuff them in shaving cream cans for the short flight to the U.S. Te visual that provided nearly made me retch in disgust. I grew up in Florida, but had no clue this was going on six hours south of my home. (The BW story is not about Cuban bullfinches, by the way).
One of the things that makes wildlife trafficking so hard to fight is law enforcement’s ability to detect it (well, that and political will). And because the trade is illicit, it’s often hard for researchers to pin a number on the problem and to be able to say that a species is declining by X percent because of poaching. So a recent paper on developing methods to model illegal wildlife trafficking published in Conservation Biology caught my eye. The article discusses applying occupancy models to illegal wildlife markets. Just as in ecological field work, where testing for the presence or absence of a species in a given range is vital to modeling that species population dynamic, so too is documenting the presence or absence of illegal wildlife (or their parts) in specific markets and shops in order to model the inputs and outputs of the trade cycle and its effect on the wild meta-population. Models that accurately describe the trade can give researchers and policymakers central clues as to how many animals are being siphoned off from the wild – which has obvious implications for their conservation, especially if the focal species is imperiled or protected.
The study author, Shannon Barber-Meyer of the World Wildlife Fund, points out a parallel between the problem of detection biases in ecological field work (finding evidence of a species in a given area… just because you can’t find sign of it does not mean it is not there), with detection biases in documenting illegal animals or their parts in shops and markets (just because you don’t see it on a shelf in the shop, does not mean the business owner is not selling it). Sometimes the dealer simply does not keep the goods on site, and the surveys demand the observer physically sees the illegal good in order to mark it as present. The cladestine nature of this trade makes it suitable for the detection biases that occupancy models address in wildlife studies. She writes,
A recent article stated, “Over the last year, TRAFFIC India recorded 27 seizures of Tiger skins from various parts of India, but given the clandestine nature of the trade, these seizures can only represent the tip of the iceberg” (Chhabra et al. 2008). Because traditional wildlife trade surveys have suffered from the same detection assumption as traditional wildlife field studies (i.e., the probability of detection is often incorrectly treated as close to one to allow for estimation of presence across a study site), the application of occupancy methods to wildlife-trade market surveys could have important wide-ranging impacts on the determination of more accurate trade estimates. These methods can even estimate trade that has gone almost entirely underground (i.e., low probability of detection) provided a sufficiently large number of repeat surveys are conducted (MacKenzie et al. 2006).
She uses a rough case study of analyzing Sumatran tiger markets as an example. Randomly selected towns with historic trading violations were surveyed in one year, and then an educational campaign to halt the illegal practice was conducted. Surveys were done in the same towns after the educational campaign. A slight decline was found in the follow-up surveys. But does this mean that the campaign worked? That is an important question both for the organizations that funded it, and for the focal species the campaign sought to protect. How are we supposed to know if the decline was a real result of the educational campaign, or simply a detection probability error? After all, wouldn’t it makes sense that the trade might be driven further underground after being exposed to a public awareness campaign telling people it was illegal to buy or sell tiger teeth, skins or body parts? The author appears to address this question by encouraging development of new sophisticated methods for measuring detection errors specific to the wildlife trade.
Despite questions of detection biases, she argues that occupancy models are good tools for estimating the fraction of shops participating in a particular trade in a given city, region or country. I’d be curious to know what professional academic ecologists think about the feasibility of using occupancy models to model illegal wildlife markets. Will this help to better define the problem? Would long term studies help to track whether the markets were causing or contributing to species declines?
BARBER-MEYER, S. (2010). Dealing with the Clandestine Nature of Wildlife-Trade Market Surveys Conservation Biology, 24 (4), 918-923 DOI: 10.1111/j.1523-1739.2010.01500.x