Thursday, March 29, 2007

This and that...

I'm finally back, after a several-week blogging hiatus. The last few weeks have been incredibly hectic, between working out the final details of my experimental design, writing 3 proposals, and generally preparing for my field season which will begin May 7 (getting permits, equipment, etc etc). Oh yeah, then I took a few days off to go to Big Bend NP in Texas - a much-needed break in an incredibly beautiful spot!!!

I'm beginning the process of learning Bayesian analysis techniques. I contacted Jackie Mohan, a candidate for the EBIO dept's Global Change position, regarding good intro texts, and she made several recommendations. Several of the papers she recommended were helpful and provided a good introduction, especially "Alternatives to statistical hypothesis testing in ecology: A guide to self teaching" by Hobbs and Hilborn (Ecological Applications 16(1): 5-19; 2006).

I also purchased several Bayesian Analysis textbooks (thank goodness for cheap used textbook websites - how did I ever manage to get through undergrad without the internet?!?) and am starting to work through them. Interestingly, we've been discussing Bayesian techniques quite a bit in my Phylogenetics class as well, and we even have an "expert" in the use of likelihood and Bayesian techniques for phylogenetic studies speaking to our class next week.

Tuesday, March 6, 2007

Losing Ground

If you haven't already read the 3-part series "Losing Ground" featured in the Times-Picayune, do so now! There is an excellent interactive graphic as well, explaining the issues clearly and concisely for those who have little time for extensive, non-scholarly reading (e.g., students).

The bottom line is, wetlands have been and are currently being lost at a rate far more rapid than most people had realized. Those maps of Louisiana's coastline that we're all familiar with, showing miles of intricately webbed marshes protecting New Orleans from storm surge? Turns out those are based on 1930s data, and the modern-day picture is disturbingly worse.

Think we still have a long time - maybe 30-50 years, beyond most of our lifespans - to solve the problem? Think again - turns out we have 10 years to restore wetlands before they're too far gone to repair.

Think solving the problem is too expensive? Well, yes, $45 billion is a big number, but here are a few points to consider to put this number into perspective:
> $45 billion = 5-8 months in Iraq at recent (pre-surge) spending rates of ~$6-9 billion/MONTH (source:
> Louisiana is critical to the nation's oil and gas industry infrastructure, with 19 active refineries (15% of total refining capacity in the country), thousands of miles of pipelines, and the closest access to the huge offshore oil-drilling industry (source: Whatever your opinion may be re: our dependence on oil, currently the nation's economy is inextricably linked to petroleum prices.
> Estuaries are critical to the seafood industry. Louisiana is the largest producer of crawfish in the world, and also exports huge quantities of oysters, catfish, and other seafood nationally and globally (source:
> The Mississippi River is an important shipping channel and New Orleans is a crucial access point and port, with ~80 million tons/year passing through (source:
> This isn't even getting into all the qualitative, difficult-to-quantify reasons why New Orleans and southern Louisiana are crucial to the United States - culturally, musically, et al. - nor does it touch on the fact that we are the United States of America, and we as a country didn't shirk at spending huge sums to rebuild Manhattan post-9/11, San Francisco post-Loma Prieta (or post-1906), Chicago post-fire, etc.

OK, stepping off the soapbox - what does this have to do with biostats or science in general? Well, the need for "proper" science, for thorough studies and full understanding of all potential impacts of action has hindered our actions to this date, and brought us to the point we are at today. While it's true that as scientists we need to be cautious and objective and understand impacts of potential conservation and restoration measures, when do the dangers of inaction or postponement of action outweigh the need for urgent action? How many pilot studies need to be done before we accept the data and act on it? I'm torn, because as a scientist I say we need to fully understand the processes before acting, but as an environmentalist and concerned citizen I say that we need to do something about this, and fast. And I'm concerned that science and the lengthy process of conducting studies and writing reports is all too commonly abused by politicians who benefit from retaining the status quo (see this book for an excellent summary of this abuse, particularly as perfected by the current administration)...while meanwhile opportunities are slipping out of our hands. See this quote in the first TP article for just one example:

"In a convoluted process, restoration projects proposed by a local sponsor must undergo a series of studies of their economic and ecological benefits and detriments. Congress must approve the studies, which can take three to five years. Then, in a separate process, the corps must seek money in its annual budget to conduct them."

In all, it can take over 10 years for any - even small - restoration action to occur, by the time all studies and bureaucratic processes have been completed. We don't have 10 years...

So, where do we draw the line? When - if ever - should the slow, cautious steps prescribed by the scientific method be overruled by the need for urgent action? Who should make the decisions regarding when or how - scientists (ecologists, engineers, wetland biologists?)? Politicians? Committees?

Friday, March 2, 2007

Independent Project

After speaking with Mike Guill re: Bayesian stats earlier this week, I've decided to change my independent project. I'm really interested in learning more about Bayesian analysis - I like the philosophical concepts behind the techniques (directly testing the hypothesis rather than the data, using prior information to come to a conclusion). I'm curious to see for myself how the analysis works (I can't really fully understand a technique like this until I've done it myself), and just how influential informative priors are on the outcomes. So, rather than analyze Tom's dataset of niche breadths for competitive exclusion, I'm going to revisit an old dataset of mine comparing arthropod abundance and defenses and avian insectivore foraging success in two Costa Rican forests.

This data was collected to test one component of an avian life-history theory which my lab has been working on. Birds - like many other organisms - have a distinct latitudinal gradient in demographic rates (most notably clutch size, but also survival et al.) wherein tropical birds have "slow" life-histories (low reproduction, high survival, low metabolic rates) and temperate birds have "fast" life-histories (high reproduction, low survival, high metabolic rates). Hypotheses have been proposed to explain this since the early 1900s, but none fully explain the observed patterns.

We suspect that food limitation is at the root of this gradient. Specifically, we cite references showing that tropical insectivorous birds face strong food limitation - contrary to popular belief - due to the low seasonality (lack of a seasonal "flush" of resources), low population densities of arthropods, and strong predator pressure (due to many species of insectivorous birds feeding on arthropod prey) leading to development of strong anti-predator defenses (physical, chemical, and behavioral) by arthropods. These forces should be strongest in so-called "perhumid" moist tropical rainforests with limited seasonality, moderate in dry-seasonal tropical forests, and weakest in temperate areas.

I sought to test this by comparing arthropod population densities and defenses in a moist tropical forest (La Selva, go figure) and a dry-seasonal forest (Palo Verde), both in Costa Rica. I also observed insectivorous birds and recorded various measures of foraging success & movement rates. I originally analyzed this data using univariate parametric and non-parametric statistics (t-tests, ANOVA, Chi-square, Kolmogorov-Smirnov) and also ran it through a multivariate MANOVA, but found few significant results.

I'm very curious to see if Bayesian analysis will produce similar results. I also have several references upon which I based my hypotheses from which I can draw data for prior distributions, and I want to test the influence of several different informative and non-informative priors.