VALUE OF INFORMATION (VOI):
My new paper (Fisher et al 2017) is essentially an analysis for the Camboriú water fund of how the choice of input data impacts the decision you'd make as a result. We compared a relatively quick analysis on free 30m resolution data to a more complex analysis using 1m data. I'd recommend most people skip most of the paper (which is quite technical) and just start with the two blogs I wrote about it (an overview at https://blog.nature.org/science/2017/08/24/camboriu-data-for-water-funds/, and a more technical one for people working with spatial data at https://rsecjournalblog.wordpress.com/2017/08/25/how-much-data-is-enough-investigating-how-spatial-data-resolution-impacts-conservation-decision-making/). In short, we found that the simpler analysis would have led us to the same decision in Brazil, but that for other water funds the choice of data could be critical (as the ROI was over 1 with 1m data, but below 1 with 30m data). Table 5 and the discussion have several guidelines to consider in how to select whether relatively low or high resolution data is most appropriate for a given context. I'm pretty excited about that part of the paper, and I'd really welcome feedback on it from anyone so inclined.
UPDATE: in December 2018 another science paper came out about the overall water fund project: https://www.sciencedirect.com/science/article/pii/S0048969718349611?via%3Dihub
Rodd Kelsey and his team just released a synopsis of the evidence synthesis they did for 20 different ag management practices and their effect on several ecosystem services (Shackelford et al 2017). It's a reference for finding information on a practice of interest, and a peer-reviewed version is forthcoming which will include expert assessment and scoring of the evidence as well. Everything in this book is also available and searchable online at http://www.conservationevidence.com under the Mediterranean Farmlands set of practices. I think this is a big step forward for TNC, especially for agriculture, but also as an example of what stepping up our evidence as part of CbD 2.0 can look like. Contact Rodd with any questions or comments you have.
Snyder 2017 is a useful reference with a lot of data on nutrient losses in the Mississippi River Basin and hypoxia in the Gulf of Mexico (as well as some global info, see Fig. 16). They show that overall the amount of nitrogen exported to the Gulf has trended down over the last 35 years (although with tons of annual variation, largely due to changes in precipitation) but phosphorous has trended up (Figs 10-13). Hypoxia in the Gulf is expected to lag way behind stream nutrient levels, and again is highly variable based on several climate variables each year, but Figure 14 shows that the average hypoxic zone from 2010-2015 is almost triple the size of the recently revised target for 2035 (<5,000 km2). So while it's not a surprise, this is more evidence that we really need to step up our game, especially given the projected impacts of climate change (see Sinha et al 2017 below) which will make our task significantly harder.
You've probably heard about "payment for ecosystem services" (PES) where a land owner / manager is paid to do something (e.g. change how they farm) or to NOT do something (e.g. not cutting down trees they would otherwise clear). Until now there hadn't been a robust, fully randomized experiment to test how well they work. Jayachandran et al 2017 is a study looking at 121 villages in Uganda, half of which were paid for two years to not cut trees (with payments tied to area of intact forest as measured via remote sensing). The good news is that overall it worked well: participating villages deforested half as much as control (4.2% forest loss vs 9.1% loss), and there didn't seem to be leakage (cutting down other neighboring forests). It also appeared to be cost-effective (based on assumptions about how villagers would respond after the 2-year program ended). Remaining questions: what would happen under a long-term version of this program (or if it was actually abandoned after 2 years), could the program be adjusted to reduce deforestation even more from participants, how can program overhead costs (1/2 of total) be cut, and could there be side-effects on biodiversity or humans? The bigger question is whether or not this would scale, the authors note that only 1/3 of people they approached agreed to participate, that if scaled up nationally it could impact timber prices which could cause some rebound, and that weaker enforcement or monitoring in a large-scale effort could impact efficacy. The calculations on costs and benefits in particular are a bit tricky, let me know if you'd like to discuss further. I'd recommend only people involved in PES schemes actually read the paper (and a longer version I can share), for others check out https://www.theatlantic.com/science/archive/2017/07/paying-people-to-preserve-their-trees/534351/?utm_source=atltw and/or http://www.nber.org/digest/aug16/w22378.html for a good overview of the paper.
Several of you sent me articles about Harwatt et al 2017, which calculated how much impact replacing all beef consumed in the US with beans would have on climate change. Note that this paper doesn't model a real world scenario, rather it performs a very simple calculation by first calculating the GHG impact of switching from beef to beans (they ran it two ways, keeping total calories the same, and keeping protein intake the same), and then comparing that to the US 2020 GHG reduction targets under the Paris agreement. They found that this switch could meet between 46-75% of the US obligations (which is a lot), based almost entirely on Nijdam et al 2012 which provided the data on emissions. I have a few concerns about the methods of this paper; I don't see the US-specific data in the Nijdam paper they cite for it, and this paper's assertion that emissions in the US per kg of beef are almost double a global average appears contrary to the underlying paper's findings that intensive systems have much lower emissions. I'm guessing this may be due to inappropriately weighting culled dairy in Europe but I can't tell b/c they don't provide the detail. So while the general idea (we should eat more beans and less beef to fight climate change) is sound, I wouldn't trust these specific numbers.
CLIMATE CHANGE & AGRICULTURE:
Kim et al 2017 argues that especially warm weather in the Arctic has led to reduced vegetation growth (from forests to crops) in Canada and some of the U.S., primarily via colder temperatures (as well as less rain in South-central U.S.). In the U.S. crop yields were 1-4% lower on average as a result, up to 20% lower for corn yields in Texas (but with the majority of states unaffected, and only a few showing a very strong relationship). As with much of climatology, this is more about concerning patterns than ironclad proof of trouble ahead. But it makes a good point about some of the complex and unexpected impacts of climate change for us to watch out for.
There's a news article about the paper here: https://www.washingtonpost.com/news/energy-environment/wp/2017/07/10/the-stubbornly-persistent-idea-about-climate-change-that-just-wont-go-away/?utm_term=.9a374e5ce552 and you can read the full paper here.
Zhao et al 2017 also looks at how climate change may reduce crop yields, although through the lens of how global temperature increases will affect wheat, rice, maize, and soy yields. They draw on and summarize four independent analytical methods (historic data, field trial data, and both global and local crop models), which is a cool trick to increase confidence in the findings. On average, they predict each degree C increase will drop wheat yields roughly 6%, rice by 3%, maize by 7%, and soy by 3%. As you'd expect, results are quite spatially heterogeneous (including a few isolated positive effects), see Fig 3 for details. There are a lot of somewhat simplistic assumptions necessary to make these estimates work but they make a good case for temperature increases causing yields to drop on existing farms. Note that they did not account for shifting cultivation (e.g. moving plantings north to reflect new conditions) or other forms of adaptation.
One concern about climate change is the shift to more intense rain (causing more runoff, erosion, and flooding than steadier weaker rain), as well as increased rain in some areas (including the US). Sinha et al 2017 does some modeling based on climate projections to predict global changes in nitrogen loads in rivers (which leads to eutrophication in coastal waters, e.g. the dead zone in the Gulf), finding that they will increase substantially in 2070-2100 (with some increase 2031-2060). There are a lot of scenarios in the paper, but under "business as usual" for climate change they predict an overall increase in N loading of 19% for 20170-2100 (driven primarily by the Northeast, Upper Mississippi, and Great Lakes regions (see Fig 1 for details, Fig 2 is less useful since it groups areas with opposing trends). They note that simply to offset that increase, we would need to reduce nitrogen inputs to farms by 33%; to actually make progress on reducing eutrophication we would have to do substantially more. They also show other countries at risk of increasing N loading, especially India, parts of China, and SE Asia. It's worth noting there are a lot of assumptions in this paper, but the overall trend that moving to flashier rain is likely to make the problem with nutrient runoff from agriculture worse is something we need to be thinking about, especially if we are unsuccessful in limiting climate change. There's a news article about the paper at https://www.nytimes.com/2017/07/27/climate/nitrogen-fertilizers-climate-change-pollution-waterways-global-warming.html