Our next speaker is Gerry Quinn, a marine ecologist who has
contributed to many textbooks in wide use today. We’ll discuss a recent paper
as a glimpse into how multivariate analysis is currently being applied to
ecological problems, in particular on the impact of agricultural land use on
invertebrates in intermittent streams. Here is a link to the article:
Hope to hear from you in the blogosphere in time for the meeting on Friday at 2.
First qn about causation - how do you get at what the processes are behind differences across land use types? Sounds like a design issue (would need to design an experiment to tease apart relative effects of salinity vs... I'm guessing there are studies on that)
ReplyDeleteLeaf packs tied together with string sound cool - would like to see a photo!
ReplyDeleteThis paper was to some extent a "null result" paper and we liked that they didn't try to hide the null results and oversell what they did find.
ReplyDeletewould have loved to see a simultaneous analysis across seasons. They say they didn't do it to keep the design balanced, but is balanced data really needed? (if using ML or REML estimation as for example in the lme4 package on R?)
ReplyDeleteMaybe just as well, the design looked complicated enough as it was...
Discussing the original hypotheses... why expect higher abundance in degraded streams? (because they have more algae?) Why expect less diversity? (because degradation removes more natives than it it introduces?)
ReplyDeleteDidn't see the detail on where these predictions came from