Radeloff et al. 2012

Post date: Aug 14, 2012 4:29:13 PM

Volker Radeloff and huge team of ecologists, economists, GIS people and computer scientists did some modeling (subscription required). They wanted to know how different land use policy scenarios would affect land cover for the entire conterminous US. They did this at a 100 m spatial grain. That's quite an accomplishment.

I found this paper after searching for and reading John Withey's publications. I saw Withey give a really interesting talk at ESA. His talk focused on an integrated model of land use decisions, based on economic comparisons of net revenue to annualized conversion costs for a few land cover types - crop land, forests, urban areas. Withey is a bird guy (I think), so he was simulating bird habitat by forest and cropland types. He also simulated carbon storage (in soils and biomass) for each of the forest and cropland cover types. Like the Radeloff et al. paper, he did this for the conterminous US at a 100 m spatial grain. With this modeling he found that, no matter what, urbanization happens. Forest conservation scenarios can work, but they will essentially conserve forests at the expense of cropland. This could have obvious impacts on agricultural production, but he also showed that it significantly affects biodiversity; many bird species associate with cropland, so reducing cropland and increasing forests actually can reduce biodiversity, and some of the adversely affected species are rare or threatened. Forest conservation also creates huge increases in carbon storage, but, again, this comes at the expense of bird biodiversity. It's an interesting demonstration of how a fairly simple economic incentive can have strong, varied, and lasting outcomes.

What was missing from Withey's model? Well, first, the agents in the model weren't making decisions based on anything other than economics. One thing that the social scientists I worked with at UO identified - through surveys of local land owners - is that economic incentives were only marginally important for conservation of endagered ecosystems, when compared to guarantees of future autonomy over property rights and land use decisions. For example, people in these surveys said that they would only be willing to restore oak savanna, for example, if they knew that, in the future, they could use that land for timber production. So we know that economics aren't necessarily the most important factor when land owners make land use decisions. But more interesting to me is that this modeling really begs for some consideration of how climate change will affect available moisture and plant productivity. Lands that are marginally useful for agriculture or forestry now may become less so in the future, and that may also drive land use decisions Withey suggested that he would be very interested in this. It was hard not to jump up in the back of the room and shout "I know how to do that! Pick me!"

Back to the Radeloff paper. They didn't simulate carbon or vertebrate habitat for this paper - they were more interested in the contrasting land use policy scenarios. (See, this is why going to ESA is so cool. Withey's paper describing the impacts on carbon storage and bird habitat won't be published for... a while. Maybe a year? Who knows? Even Withey doesn't. But by going to the conference, one can get a more current glimpse into where the science is, and where it's going. That was one of my main motivations for recording all those audiocasts.) What they found was fairly surprising. It didn't really matter which land use policy scenario they used. The overall trends were basically the same among scenarios: "All four scenarios predicted substantial declines in crops, pasture, and range, and a substantial increase in urban area. The scenarios predicted medium to strong increases in forest land depending on the scenario." That should be a little surprising, since one of the scenarios subsidized agriculture, while another subsidized forestry, but there weren't big differences between those scenarios. This is a little alarming, because it suggested to me that we're deeply committed to our present land use course - there's so much social inertia built into our practices - that even alternative policies can't do much to change where the landscape is headed.

But here's my favorite part of the paper:

"If climate changes substantially over the next 50 years, the range of reasonable land-use choices available to landowners in various parts of the United States will be different than the reasonable choices landowners faced in the same region a generation earlier. A worthwhile extension of our model may thus be to add biophysical data to restrict tree growth where it is not possible."

Once again, it's hard not to jump up and down and shout "I know how to do that! Pick me!"