The first science blogger?

I nominate Johannes Kepler as the first science blogger. Dedre Gentner, in her paper Analogy in Scientific Discovery: The Case Johannes Kepler (2001) writes that

[Kepler] provided a running account of his feelings about the work, including the kind of emotional remarks that no modern scientist would consider publishing.

As an example she offers the following quote from Kepler's Astronomia novae

If I had embarked upon this path a little more thoughtfully, I might have immediately arrived at the truth of the matter. But since I was blind from desire I did not pay attention to each and every part [...] and thus entered into new labyrinths, from which we will have to extract ourselves. (Kepler 1609, pp. 455-456)

Gentner provides a few other choice quotes too - hence I think that if Kepler were around today, he'd be blogging.

Predicting the future

This is good.

Don't Transform

One of my pet peeves about my ecological colleagues is their tendency to transform binomial data using arcsine of the squareroot of the proportion in order to use a linear model. OK, once upon a time, it might have made sense to do this. But we have better tools now, honestly! Travis Hinkelman brought a great paper by David Warton and Francis Hui to my attention this morning. I'm just going to quote one line, which sort of says it all:

The most striking result in power simulations was that logistic regression and GLMM always had higher power than untransformed and arcsine transformed linear models ...

So, don't transform your binomial data. And please, if you are collecting proportion data, write down both the numerator and denominator! This will be required reading in my Ecological statistics class next fall.

Understanding Government

Here's a bit of a fun (or maybe disturbing) read. It reminds me of something I have to remind myself often - that I was NOT a representative example of an undergraduate student - critical to keep that in mind when evaluating student work! Brigitte Tenhumberg and I were having a conversation about what we should be trying to get our undergraduate students, especially non-biology majors, to understand about ecology. Like Pielke on politicians, we concluded that it isn't reasonable to expect undergraduate students to become experts in ecology, and therefore it isn't possible for them to determine the validity of claims made in the media (including on the internet), about ecological consequences of various events. It seems as though we ought to be teaching them to evaluate the credibility of the people making the claims - but wow! what a can of worms that idea turns out to be.
So, as an "expert" (not sure in what!) offering my opinions up on the internet via this blog, perhaps the most important thing I can do is provide access to evidence that allows readers to evaluate my credibility on a particular claim.
Hmmm, from my bio in the top right corner it takes 3 clicks to reach an (outdated! oops) copy of my CV - and probably only because I know where to look. Googling my name gives access to that same 2 page vitae (2nd hit) and also my Facebook, linkedin, Academia.edu, Mendeley and Flickr profiles. All that tells someone is that I'm addicted to social networking sites ...

Additivity in wolf harvest

Scott Creel and Jay Rotella conclude:

Examined across populations, human killing of wolves is generally not compensatory, as has been widely argued. Management policies should not assume that an increase in human-caused mortality will be offset by a decline in natural mortality.

Seems pretty cut and dried, and looking at the way they analysed their data, I can't find any reason to disagree with them. Given their result, this is a very balanced and fair statement; they also say that some level of wolf harvest probably is sustainable. However that sustainable harvest is probably lower than proposed in current Montana and Idaho management plans.

Dr. The Bird Man wondered if Adaptive Management could be used to resolve the Additive/Compensatory controversy, along the lines of the North American Waterfowl Harvest Management Plan. I don't think it would help in this instance. The wolf harvest is marked by sharp distinctions in how wolves are valued among various stakeholders. In addition, the institutions tasked with managing wolves are new at the game - for the past 30 years that job has been handled by the USFWS. This means that everyone - for or against harvest of wolves - is learning a new set of skills, interacting with new people, and old people in new ways. In contrast, when the AM plan for waterfowl harvest was adopted in 1995, the institutions managing the harvest had been doing so for decades, using the same types of data, and the value diversity was (and still is) much lower than in the wolf case. It's worth noting that even after 12 years of analyzing data, the waterfowl AM process still couldn't distinguish between compensatory and additive harvest (Nichols et al 2007). I wonder if a meta-analytic approach similar to what Creel and Rotella used wouldn't be more helpful.

This does not mean that careful analysis and thinking about wolf populations won't be useful. The risk is that parties on both sides of the debate substitute arguments about the quality of the science for the real debate about how many wolves we want, or are prepared to live with. That's a value based question, and until the debate turns away from the science and focuses on the emotional, subjective, icky stuff, it'll be hard to resolve anything.

More on the science and politics of wolves.

The High Country News Range blog posted on the wolf controversy, and stated:

what’s become clear in the cacophony regarding wolves in the West is that where emotion rules, research should.

which is interesting, because the conclusion of social scientists who study the science-policy interface is exactly the opposite. It would be all too easy for scientists to fall into the "stealth issue advocacy" trap in controversy over wolves. The issue is a highly polarizing one - people seem to love wolves or hate them. A scientist wishing to connect their science to policy can easily find themselves arguing for a particular position "only based on objective science", ignoring that their values inevitably influence what they research, how they research it, and what conclusions they draw.
A better role for a scientist, albeit more difficult, is to use science to evaluate a range of policy options. This is in fact what Creel and Rotella have provided for the wolf case: based on 21 studies of wolves, what is the relationship between human offtake (harvest or culling), and total mortality? With this relationship in hand, it is possible to evaluate different harvest quotas in terms of the future wolf population size. That may or may not be used by policy makers in Montana, but it certainly should be taken seriously.

Politics and science

I haven't read (yet) Scott Creel and Jay Rotella's article that is at the heart of this controversy, but I can feel for them. It is necessary to make assumptions when constructing a population model - and if you make different assumptions you'll get different results. The fact that the paper is peer-reviewed increases my confidence that their assumptions are defensible. Unfortunately, the results are not politically palatable! I'm looking forward to digging into their study in detail.