Hi, Kevin,
I think statistics are the least of your problems. Depending on the
forest type you choose, your tree crowns are going to be 40 to 200 feet in
the air. How are you going climb hundreds of trees, get collecting sheets
under representative branches where their foliage is (i.e., out on thin
twigs far from the trunk), whack them with a stick, and get the critters
rounded up and into your cages before the end of time? I hope you have
access to an all-terrain vehicle with a very tall cherry-picker arm. In
addition to these logistical difficulties, your comparisons will only be
meaningful if the samples are collected at the same time of day, at about
the same time of year, and in about the same weather conditions. Do you
have 45 of those cherry pickers and an army of volunteers so you can get all
of your samples in a short period of time? And what about insects that hang
on tight, fly away, or resent having their branch whacked? And while you're
dealing with these issues, who's going to be feeding the perhaps many
thousands of specimens in their individual containers and recording the
their behavior, growth, and mortality?
Martin Meiss
2010/5/9 Etienne Laliberté <etiennelalibe...@gmail.com>
> Dear Kevin,
>
> In the future, r-sig-eco
> (https://stat.ethz.ch/mailman/listinfo/r-sig-ecology) might be a better
> list for R-related ecological questions, although you didn't really have
> questions on bipartite itself (but if that's what you want to know, yes,
> it's the package you want to use for this kind of stuff).
>
> Some hints:
>
> Question 1)
> have a look at:
>
> Nielsen, A. and J. Bascompte. 2007. Ecological networks, nestedness, and
> sampling effort. Journal of Ecology 95:1134-1141.
>
> and this one also discusses this sampling issue:
>
> Tylianakis, J. M., E. Laliberté, A. Nielsen, and J. Bascompte. 2010.
> Conservation of species interaction networks. Biological Conservation
> (in press; see www.elaliberte.info/publications for a PDF).
>
> Webs are often pooled in analyses. But bear in mind webs from different
> habitats can be more spatially / temporally variable than others, and
> pooling will obscure this. For example see:
>
> Laliberté, E. and J. M. Tylianakis. 2010. Deforestation homogenizes
> tropical parasitoid-host networks. Ecology (in press; see
> www.elaliberte.info/publications for a PDF).
>
> Question 2)
> Field constraints are always a tough call. A rule-of-thumb is that you
> should have at least 10 replicates per level, so maybe you could drop a
> few replicates. Then again, a 20m X 50m also plot seems pretty big to
> me. A small pilot study could help you decide how to optimize sampling
> effort.
>
> Question 3)
> Go for abundance. You can always convert back abundance to
> presence-absence, but not the opposite. It can be a good idea to look at
> differences in both binary and quantitative webs, because you can get
> different answers, e.g.
>
> Tylianakis, J. M., T. Tscharntke, and O. T. Lewis. 2007. Habitat
> modification alters the structure of tropical host-parasitoid food webs.
> Nature 445:202-205.
>
> or, sometimes, similar answers, such as Laliberté and Tylianakis (2010)
> (see above reference).
>
> Hope that helps,
>
> Etienne
>
>
> --
> Etienne Laliberté
> ================================
> School of Forestry
> University of Canterbury
> Private Bag 4800
> Christchurch 8140, New Zealand
> Phone: +64 3 366 7001 ext. 8365
> Fax: +64 3 364 2124
> www.elaliberte.info
>
