Quantifying biodiversity

I stumbled across an interesting paper in PNAS these days but I am afraid I have to complain a little:

Over a series of three- to six-month field sessions across 10 years, a research team hiked across the hilly tropical agricultural landscape of Coto Brus, an area in Costa Rica. In an ecological gradient from protected forest to treeless pasture, the team made a total of 67,737 observations of 908 species, comprised of understory plants, non-flying mammals, bats, birds, reptiles and amphibians.

They plotted their plant and animal observations on fine-scale maps from Google Earth aerial photographs. The perhaps not so surprising find was that tree cover is very important. More surprising was the fact that for four of the six species groups (plants, non-flying mammals, bats and birds), the colleagues saw a significant increase in the number of species with increasing tree cover visible on Google Earth maps. The models they developed were able to predict biodiversity in the region within a 30- to 70-meter radius and also demonstrated how the number and kind of species change as trees are added to plots of land.

So far so good but I think this study lacks a lot of species that make up the majority of the local diversity - the team did not include any invertebrate species although they make up most of the diversity and are of crucial importance for agriculture. I am not saying their idea was a bad one or the results are flawed. On the contrary, their analysis is great but I think the study is incomplete. One of the main messages "trees are good" could be even stronger if they had started to count  e.g. arthropod taxa as well which would have taken them much longer and they already worked for 10 years. 

My main problem with this study is not the study itself but the claim that this way of looking at diversity (remote sensing through aerial photography) could be used by policymakers to help protect biodiversity and endangered species. For such a statement they haven't done enough which might sound harsh given the years of work that were put into this. However, I am always critical of shortcuts which leave out a very large portion of live and its role in the respective ecosystems. I am afraid it ain't that easy. Most of our current estimates, assessments, policies are based on very few key indicator species selected for their alleged role in an ecosystem. This has been rightfully criticized by many scientists and even policy experts. If we as humans are really serious about protecting biodiversity we have to go through the pains of understanding it properly and in its entirety. Bad news for policymakers but who said that live is easy?