[caption id="attachment_44" align="alignleft" width="200" caption="Bear is unimpressed. Photo CC by Flickr user Marshmallow."][/caption]

It was summer of 2008 and the rhetoric was getting as hot as a globally warmed hood on a '91 Chevy Camaro RS (my 2nd car, with t-tops of course). While you might fry an egg on the hood, you could broil a few cornish hens on the hot tin roof that encapsulated the election year debate. In true political posturing, Senator John McCain (R-AZ), the GOP presidential candidate, thumbed his nose at wasteful government spending on science research, "I don't know if it was a paternity issue or criminal, but it was a waste of money." This was in regards to a $3 million grant to what McCain called "bear paternity tests" throughout his campaign.

The bear, once a rallying symbol of the Reagan Administration, is now a symbol of everything that is wrong in American politics. Of course, rhetoric is meaningless unless it is backed by action and Senator McCain failed to act on his shining example of government waste. In 2003, as part of the Appropriations Committee, he ended up voting for the final bill without seeking to remove the bear study, deriding the research much as he would continue to do in the next 5 years:

"Approach a bear: That bear cub over there claims you are his father, and we need to take your DNA. Approach another bear: Two hikers had their food stolen by a bear, and we think it is you. We have to get the DNA. The DNA doesn't fit, you got to acquit, if I might."-McCain, Senate floor, 02/13/03 via Factcheck.org


The Bear DNA research, more formally known as the Northern Divide Grizzly Bear Project - headed by Dr. Katherine Kendall of the United States Geologic Survey at Glacier Field Station in Montana, used wire traps and video surveillance cameras to capture hair from bears all over the Rockies. It is an ambitious and highly successful program. In its operation has identified nearly every individual bear in the area, estimated to be ~765 (Kendall et al. 2009). The focus of this study was to determine the population size of an integral component of the great northwestern ecology, to evaluate the success of a Grizzly bear conservation measure that dates back to 1975, and to understand how individuals (and groups) of bears move throughout the terrain. For instance, are there barriers that prevent sub-populations of bears from coming into contact with each other and doing what.. you know... bears do when they find each other... out in the wild... you know... The authors concluded in their study the following:

  1. Bear population was at a healthy size
  2. In 2004, humans caused around 5% of bear deaths, slightly above what they consider to be sustainable
  3. Bears occupied a range that included over 10,000 square kilometers outside of the recovery zone
  4. There is evidence for populations being fragmented over major transportation corridors, such as interstates

In any measure of a conservation initiative this would be considered a successful study. They sampled thoroughly, had solid statistical methods, used existing tools innovatively, made robust conclusions about their data and unbiasedly assessed the conservation status of the Grizzly bear. Still, even if you are convinced of the scientific integrity you could still fairly ask was it worth $3 million? Its a fair question and its a question that scientists typically scoff at answering. Why should THEY after all, have to answer to Joe the Plumber about their research's integrity when it had already passed judgement through the exalted channels of traditional scientific scrutiny.

The easy answer, of course, is that the import of any publicly funded research must be communicated to the public. As taxpayers in the American economy, we are funding all research from government agencies, NSF and NIH whether we know, or like, it. There is a often a disconnect between the research proposal and the research implementation due in part to the back channels of science funding. Government agencies, rightly, place the job of evaluating research proposals in the hands of the experts. The experts have a vested interest in others' research for a variety reasons, some nefarious but most want to see their fields progress under the guidelines of good science. Once a panel of experts has concluded the proposal is of merit, the budget is appropriate, and there exist enough funds the research may begin. This is a condensed, simplified version of the proposal process.

***For a more thorough breakdown of the actual science of "bear paternity tests" see this excellent post at Southern Fried Science: Effective Size of a Population in Flux***


One of the reasons misleading statement such as "bear paternity tests" can be shocking to the general citizen, is because of the behind-the-scenes nature of submitting and evaluating research proposals. This isn't a bad thing, but merely an observation. To Joe the Plumber the costs of doing research seem frivolous and inciting. He, after all, is working hard to make ends meet, breaks his back every day supporting his family and never asks for nothing from the government! All the while, these eggheads up in their ivory towers play chess with his hard earned tax dollars. $3 million is a lot of money.

Let's take a closer look at where taxpayer funded research grants go, starting with a shiny, new $1 million National Science Foundation grant. The university takes a rather large cut off the top, somewhere near 40% (but can be much higher), this covers all the costs of housing the researcher and their staff such as the electrical bills, support staff salaries (secretaries, accounting office, maintenance, etc.), water and sewer, any shared facilities, the amount of toilet paper you and your research staff will go through - among other things - and on and on. This would leave the project $600,000. Given that the research was to occur over 3 years that leaves about $200,000 a year for the project. Many projects have several investigators, often from different institutions, and the amount each receives depend on their involvement. Let's say on average there are 3 investigators and for the sake of argument each are equally involved so that the money is spread equally, about $66,000 each.

This amount of $66,000 is used to for expenses of the research as well as to pay research and technical staff in the lab of the investigator. There needs to be someone to actually do the research. A graduate student is the cheapest labor you can find to carry out the field studies or experiments, engage in the intellectual work and bring the research eventually to publication. A typical graduate student salary is about $18,000 and again with overhead costs and tuition, brings this amount something close to $40,000. As an aside, a technician with an annual salary of $35,000 actually costs the investigator probably somewhere around $70,000. Hiring a technician in this scenario would put the investigator over budget. This leaves the investigator now with $26,000, some of which will pay the summer salary of the investigator (academics are paid on 9-month scales, so another 3 months of support might be around $10,000). The rest of the remaining $16,000, is used for travel to field sites (air tickets for 4 researchers? $2400. Food and lodging on government per diem rate for 4 weeks of field work? estimate $80/day for total of $2,200 x four researchers=$8,800), laboratory analyses and supplies, conference travel and registrations for lab members, publishing costs, and preliminary experiments to help get the next grant. Very little is left over in the end.

Clearly, government grants are not a get rich quick scheme for scientists. On the contrary, scientists, technicians, graduate students and other researchers are obliged to stay within the salary guidelines of their institutions, which are on par with typical middle class salaries. But the disconnect for the public comes when the message they receive politicians is that scientists are misusing their tax money when their elected officials have little to do with gets funded. Much like citizens entrust in politicians to represent their interests to the governmental body, politicians entrust the administrators and volunteer scientists to determine the best research proposals worthy of government funding.


Along the journey to uncover the mysteries of bear conservation genetics, Kendall and her colleagues had to develop a number of tools to get to the point where they could publish their ultimate results. some of these are noninvasive field methods that captured hair from bears in the wilderness without any human intervention. It is quite amazing that with these hair traps all over the northwest, they could use this as indicators or bear presence over an enormous range. The same individuals were tracked thousands of miles apart over the years.

One of the values of this study was the intensive sampling. Nearly every bear in the upper Rockies has been accounted for. They are able to match siblings and reconstruct families. This is important in being able to test ideas of ecology using genetic data. One of the more fundamental questions in population genetics is how many genetic markers does one need to be able to assign individuals to any particular population. A very basic methodological question that is fairly complex.

To assign individuals to populations we need to know the probabilities that full siblings share identical genotypes and that two individuals have identical genotypes (Woods et al. 1999; Waits et al. 2001). To work this out one really needs a full dataset of individuals with known relationships and a lot of genetic markers. This is exactly what the bear DNA research has accomplished. With these statistics in hand now conservation biologists can now evaluate whether the set of markers has the statistical power to assign to detect identical individuals in their datasets. This is enormously important because it allows us to find the minimum number of markers needed to accurately assign an individual to the correct population.

This basic statistical tool is the result of bear paternity testing and has uses far outside the realm of bears into any conservation genetics problem. When you spending thousands of dollars managing endangered species, you want to make sure you know who is in each population, how much mating is going on among populations and how populations are moving around. I've used these tools myself to understand if a microsatellite marker set I developed had the resolution to detect identical genotype and correctly assign individuals to populations of deep-sea hydrothermal vent shrimp (Zelnio et al. 2010).

Research on bears has had a major impact on an entire field and has contributed to spur innovative studies to even the deepest ocean basins. What is often missed between the sound bites and political posturing of the qualities of government-funded research is real purpose of doing research. Rarely is basic research ever without applied results. Development of new tools and methods during the journey to answer research objectives can be far-reaching and set new bars for entire fields of study. While you might not care for bears, perhaps you care about mountain lions, birds or butterflies. The tools developed for one species may be quite applicable across the breadth of animal life.

The big challenge is to make research relevant to everyone. This will require extensive collaboration with scientists, funding agencies, politicians, communicators and marketers.  We need better ways to show that basic research isn't frivolous. Scientists and science communicators should seize the moment when research is being used for political posturing to turn it into a teachable moment. But this matters most when the media are watching. In order to take full advantage of taking charge of misleading situations, action needs to be prompt and decisive. Waiting to gather your thoughts in the political landscape is political suicide. It is a reactionary medium, one that scientists in particular are not accustomed to. But if we want to continue to be leaders in science we must get our feet wet and join the conversation.


Kendall, K., Stetz, J., Boulanger, J., Macleod, A., Paetkau, D., & White, G. (2009). Demography and Genetic Structure of a Recovering Grizzly Bear Population Journal of Wildlife Management, 73 (1), 3-17 DOI: 10.2193/2008-330

Woods, J.G., Paetkau, D., Lewis, D., McLellan, B.N., Proctor, M., & Strobeck, C. (1999). Genetic Tagging of Free-Ranging Black and Brown Bears Wildlife Society Bulletin, 27 (3), 616-627 JSTOR

Waits, L., Luikart, G., & Taberlet, P. (2001). Estimating the probability of identity among genotypes in natural populations: cautions and guidelines Molecular Ecology, 10 (1), 249-256 DOI: 10.1046/j.1365-294X.2001.01185.x

Zelnio, K., Thaler, A., Jones, R., Saleu, W., Schultz, T., Dover, C., & Carlsson, J. (2010). Characterization of nine polymorphic microsatellite loci in Chorocaris sp. (Crustacea, Caridea, Alvinocarididae) from deep-sea hydrothermal vents Conservation Genetics Resources, 2 (1), 223-226 DOI: 10.1007/s12686-010-9243-0