Via Garry Peterson  I discovered this post by Simon Donner about forecasts of this year’s massive coral bleaching event in the Caribbean.

Donner and colleagues published a paper in PNAS in 2007 in which they calculated that heat waves that cause massive coral bleaching, like a previous event in 2005, had gone from being 1-in-1000-year events to a probability of once every 10-50 years during the 1990s, and by the 2030s will occur every 1-2 years. This year’s event, says Donner, is another fingerprint of human-driven climate change:

Keep in mind the predictions. This is what the scientific community predicted was likely to happen. An event which we calculated would be a once in a millennium occurrence without human impact on the climate, happened again five years later.

Donner’s post led me to NOAA’s Coral Reef Watch Program, which makes near-term forecasts of sea temperatures.  Our knowledge of  ocean climate allows us to make predictions several months out.  It turns out NOAA warned of this season’s bleaching at the beginning of June.

We are getting better at forecasting the future. As we learn more, we will get better at both Donner’s type of prediction (changes in long-term frequencies), and NOAA’s (changes in short-term likelihood).

My question is: Do we know enough to use this information?  I was actually surprised to learn from NOAA’s site the extent of short-term responses available in the case of corals.  Reefs can be shaded to prevent heat stress, and if possible certain stressors, like fishing or nearby pollution discharges, can be suspended.   But such actions are probably only marginally effective, and can only be carried out in a few systems. Besides minimizing the extent of climate change, building resilience into coral reefs and the social-ecological systems in which they are embedded will be more important in the long run.  

My own work, which I hope to write about here soon, involves similar problems in forests - we know that drought-driven die-offs will increase in the future, and (I hope) we may be able to predict them in the near term.   But how can we use our ability to forecast to build more resilient systems? Donner, S., Knutson, T., & Oppenheimer, M. (2007). Model-based assessment of the role of human-induced climate change in the 2005 Caribbean coral bleaching event Proceedings of the National Academy of Sciences, 104 (13), 5483-5488 DOI: 10.1073/pnas.0610122104

Normally I’m fairly skeptical of studies that attempt to put one big number around the value of a global ecosystem service.  In general, studies at such coarse spatial scales have more uncertainty and are not useful at the regional and local levels where decisions are generally made.  Nevertheless, I’m intrigued by this study in the latest Ecological Economics that attempts to put a value on marine genetic diversity’s contribution to the development of future pharmaceutical products:

….Here, we provide the first global estimate of the number, source and market value of undiscovered oncology drugs based on empirical data, industry statistics and conservative modelling assumptions. We report US$563 billion–5.69 trillion attributable to anti-cancer drugs of marine origin pending discovery, revealing a new and substantial at-risk ecosystem service value. Our model predicted 253,120–594,232 novel chemicals in marine organisms; 90.4–92.6% of these compounds remain undiscovered. A total of 55 to 214 new anti-cancer drugs were predicted to reach the market sourced primarily from animal phyla (Chordata, Mollusca, Porifera, and Byrozoa) and microbial phyla (Proteobacteria and Cyanobacteria). While no single aspect of extractive marine resource value should be relied upon to account for the opportunity costs of conservation initiatives, the application of valuation models to ecosystem services further reveals the true, irreversible economic cost of habitat degradation and biodiversity declines.

Of course, a ten-fold range of value leaves a lot to be desired, and it’s easy to pick out assumptions that could change these calculations.  However, the authors are fairly explicit in describing their research as primarily having demonstrative value:

Ecosystem service valuation aims to first demonstrate the existence of sufficient biodiversity value to promote conservation initiates, and second, to show how to capture and appropriate enough of the value to compensate for the opportunity costs of conservation in specific areas…The present study fulfils the former of these goals and promotes progression towards the second, where further studies focusing on case-specific scenarios and context-dependent variables will allow for the evaluation of ecosystem usage alternatives.

Perhaps this will get pharmaceutical companies to help with financing for marine protected areas.

One thing in the paper jumped out at me.   As shown in the figure below, there doesn’t seem be any any one group of marine creatures that yield a large number of potential drugs.  The number of potential drugs (shown below as Marine Natural Products, or MNPs) discovered has a strong relationship with number of species examined.  This means that these chemicals are found in most marine taxa - finding them just requires investigating more species.

However, according to the data in the chart below, the plurality of efforts at developing drugs from compounds found in marine life are focusing on one phylum: Porifera, or sea sponges.

This is curious.  Maybe I’m missing something, but if these compounds are found across phyla, it seems that we are missing opportunities by concentrating heavily on just a few. Erwin, P., López-Legentil, S., & Schuhmann, P. (2010). The pharmaceutical value of marine biodiversity for anti-cancer drug discovery Ecological Economics DOI: 10.1016/j.ecolecon.2010.09.030