[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Statistics and other stuff



Hi Folks,

Sorry to ramble, but...

I gained my BS in Wildlife Biology at Umass, and watched the funding shift from wildlife to the lab sciences (since they pay more $$$ in terms of results). However I have to take issue with the idea that statistics are somehow evil and the anti-scientific stance it encourages.

I too only have a BS and my entire family are math midgets, yet I have taken the time to read scientific papers and understand research design (Wilson Bulletin and Condor are on line now, so everyone can do this to some degree). I understand that it?s frustrating to read rather simple, descriptive ornithological works by the likes of Wilson, Alfred Russel Wallace and even later works like Odum and wonder why things can?t be the same now as they were then. But things have changed because of those peoples work, willingness and desire to spend long hours observing their environment and the animals in it, not just tick it off on a list and move on (don?t get me wrong. A bit of listing can go a long way, esp. in the tropics). It does not take much for someone to go out to the local forest tract, beach or other favorite birding spot and come up with a publishable paper. It only takes time. Ted Parker (another guy with a BS) published many papers and presented valuable data on neotropical birds simply by walking survey transects over and over (Again, this was in the tropics where not much is known about distributions, and Ted WAS the best at ID, but the point is still valid). Had Wilson, Audubon, Gould etc. etc. collected their data in a more hardened scientific manner, we would all be much better off. A paper published in 2001 regarding Ammodramus sparrows in SC begs the point. The authors examined museum collections of Grasshopper Le Contes and Henslow's sparrows collected 100 years ago in SC. They came up with some interesting items regarding populations and distributions in the state. But one problem with the method is that it assumes the birds were shot an equal rate (i.e. shooters decided to take ANY grassland sparrow equally with no preferences to color, flight ability etc.) 100 Years ago, had the ornithologists developed a hardened scientific method, we'd be better off today. Heck, Darwin didn?t even indicate on his specimens which islands they came from. It took a letter from Wallace to suggest there might be differences across the Galapagos. Even today we run into sampling problems with BBS, CBC, GBBC (but they are fun as heck aren't they?).

The point is that instead of listing hundreds of species in your state for a given year, go out and observe 1 or 5 or 10 species for that year, and do it intensively. Simple statistical programs and help are available all over the net, and you may just learn some valuable information about bird species distribution, reproductive success, population trends or who knows what? Science isn't for everyone, but it certainly should be for a lot of people. It hones ones day-to-day decision-making process.

Example: While I DO work on an avian research project here in southern South Carolina, I have taken my own time to pursue an added project. I have taken the time to go out on power line transects, Carolina bays and pine savanna (with is extremely rare I have discovered? research shows <5% is left. I believe it). I walked numerous transects with a ten meter long pole, brushing the grass stalks with it. I surveyed grassland birds in this manner from late Nov through now. Now not only did I find out some interesting things about Bachman?s, Henslow?s and Grasshopper Sparrow winter populations in SC, I also discovered some interesting things about habitat selection. The process was rather simple, and the result is going to be journal papers and a poster presentation at the Partners In Flight meeting in March. Now, If I were retired, I?d have been able to put even more time into it.

Here?s a title for you: ?Norway Spruce As Roost Habitat For Birds of Prey?. Go through your neighborhood and locate spruces. Shake each one (with permission for those on private property) once-a-week (?or month or whatever you can manage). As you approach a target spruce, make sure you don?t scare off any quarry that might be there, and if you do record how many you scare off. Shake each for 30 seconds. Your assumption is that this will scare out any raptor that might be in the spruce. Now as a control, you do the same to randomly selected, similar sized trees of a different species (or several species). Do the exact same to them, varying the order so you don?t survey any tree at the same time every visit (raptors may keep schedules and you don?t want to miss it). Simple t-test programs are available on the net. So now, instead of walking by spruces for the ?several hundredth time and wondering about them?, you have not only provided information for other people, but you have hardened your decision-making method about them (and perhaps other things). Is that so bad?

Paul Champlin (science geek)
New Ellenton, SC

_________________________________________________________________
The new MSN 8: smart spam protection and 2 months FREE* http://join.msn.com/?page=features/junkmail