Wednesday, March 9, 2016

Declines in tree nutrient concentration over past 25 years


I've been trying to catch up on journals lately. Apparently, I hadn't read anything from Global Change Biology over the past 2 years. Must have been distracted. No time like the present...

Here's one that struck me as amazing.

Researchers in Europe resampled forest leaves from 1992 - 2009 across a large number of plots in Europe. At each site for a subset of species they assessed nutrient concentrations and leaf mass--a pretty simple and standard measurement. Doing this allowed them to examine the trajectory of nutrient concentrations (and contents). Nutrient concentrations in leaves are critical to determining tree productivity as well as interactions with herbivores, so knowing whether concentrations are going up or down is critical to modeling the future productivity of these forests.

Here's the simplified result: almost all nutrient concentrations were declining. 20 nutrients had declining concentrations. 2 were increasing.

Here's an example of the pattern for beech. white bars are concentrations, grey contents.




The authors focus on P nutrition the most, emphasizing the role of N deposition in promoting P limitation. Yet, even N concentrations were declining. These declines must be more than just N deposition causing imbalances, especially since N deposition has been declining over the time period. 

The authors suggest elevated atmospheric CO2 might also be playing a role, as well as droughts and warming, but this paper mostly describes the pattern, which is fine.

The big question is: What is causing this massive, continental decline in nutrient concentrations?




Monday, March 7, 2016

ASA statement on P-values



The American Statistical Associations statement on the use of p-values can be found here.

The short list is:

  1. P-values can indicate how incompatible the data are with a specified statistical model. 
  2. P-values do not measure the probability that the studied hypothesis is true, or the probability that the data were produced by random chance alone. 
  3. Scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold. 
  4. Proper inference requires full reporting and transparency. 
  5. A p-value, or statistical significance, does not measure the size of an effect or the importance of a result. 
  6. By itself, a p-value does not provide a good measure of evidence regarding a model or hypothesis. 
My personal take is that there are a few corrections in how p-values are used. 

1) p< 0.05 is arbitrary. report the exact p-values and think of them as a continuum. Don't think a paper should be accepted just because p < 0.05. Don't reject a paper just because p > 0.05. 

2) the p-value reported needs to be contextualized with the number of comparisons made. this is where p-hacking shows up. if you do 20 independent analyses, 1 is likely to have p-value < 0.05. You need to state that you did an additional 19 analyses if you are reporting the 20th. if you went and added more data or looked more carefully for outliers because a p-value wasn't low enough, this needs to be reported.

3) p-values and effect sizes must be reported together. an independent assessment of whether the measured effect is biologically relevant is needed. 

#2 on the list is the hardest to comprehend because it involves logical assumptions of the test. 

The manuscript's explanation of this is:

Researchers often wish to turn a p-value into a statement about the truth of a null hypothesis, or about the probability that random chance produced the observed data. The p-value is neither. It is a statement about data in relation to a specified hypothetical explanation, and is not a statement about the explanation itself.

At RetractionWatch, the author explains it this way:

Retraction Watch: Some of the principles seem straightforward, but I was curious about #2 – I often hear people describe the purpose of a p value as a way to estimate the probability the data were produced by random chance alone. Why is that a false belief? 
Ron Wasserstein: Let’s think about what that statement would mean for a simplistic example. Suppose a new treatment for a serious disease is alleged to work better than the current treatment. We test the claim by matching 5 pairs of similarly ill patients and randomly assigning one to the current and one to the new treatment in each pair. The null hypothesis is that the new treatment and the old each have a 50-50 chance of producing the better outcome for any pair. If that’s true, the probability the new treatment will win for all five pairs is (½)5 = 1/32, or about 0.03. If the data show that the new treatment does produce a better outcome for all 5 pairs, the p-value is 0.03. It represents the probability of that result, under the assumption that the new and old treatments are equally likely to win. It is not the probability the new treatment and the old treatment are equally likely to win.
This is perhaps subtle, but it is not quibbling.  It is a most basic logical fallacy to conclude something is true that you had to assume to be true in order to reach that conclusion.  If you fall for that fallacy, then you will conclude there is only a 3% chance that the treatments are equally likely to produce the better outcome, and assign a 97% chance that the new treatment is better. You will have committed, as Vizzini says in “The Princess Bride,” a classic (and serious) blunder.
**
I'm still looking for the right wording on this one, but it seems like the probability that the null hypothesis is true given the effect size observed. 

Saturday, March 5, 2016

Biogeochemical Planetary Boundary: Beyond the zone of uncertainty? (Part II)


I think of scientists as having two jobs.

One is to create intellectual tension.

The other is to resolve it.

Creating intellectual tension is generating hypotheses. Hypotheses that we do not know whether they are true or false represents intellectual tension. Competing hypotheses about how the world works are also intellectual tension. We do not know which is true. This is the tension.

Resolving intellectual tension can sometimes occur by identifying logical flaws in one hypothesis. Generally, intellectual tension is resolved by collecting data. It is a fair question about whether a hypothesis can ever be proven or disproven and therefore whether intellectual tension is ever fully resolved, but the process of science works to reduce intellectual by favoring hypotheses.

In the previous post, I identified some important intellectual tension in the scientific world.

There is the hypothesis that the planet has exceeded a biogeochemical "planetary boundary". Too much nitrogen is being fixed and entering ecosystems. This is the hypothesis.

Yet, it is unclear whether this is causing planetary-scale eutrophication of terrestrial ecosystems or  aquatic ecosystems.

On the one hand, we have a hypothesis where the world is awash in nitrogen. We fix more nitrogen than ever and apply it to ecosystems on a massive scale. As a result, nitrogen is leaking out into waterways creating dead zones in the oceans. Nitrogen is also entering the atmosphere and raining down on even the most remote ecosystems on earth. As a result, terrestrial ecosystems are becoming eutrophied. Species adapted to low nitrogen availability are being crowded out by faster growing plants. Biodiversity is plummeting. Productivity is increasing unsustainably. With all this extra nitrogen, we have exceeded a biogeochemical planetary boundary. Civilization as we know it is threatened.

Yet, the intellectual tension on this hypothesis actually takes the form of a competing hypothesis. It is possible that not only have we not exceeded a planetary boundary for nitrogen, but ecosystems might be becoming more nitrogen limited over time. As temperatures warm and atmospheric CO2 builds up, this might stimulate the demand for N more than it is being supplied. Plants and microbes become more limited by N. Plant N concentrations decline. Photosynthesis declines. Plants that compete well for N become more dominant. Less N leaks out of ecosystems into streams. Productivity becomes more and more constrained by the lack of nitrogen. Vegetation sequesters less and less carbon than they could be, all because there is not enough nitrogen. As a result, more CO2 accumulates in the atmosphere than could be if forests had more nitrogen. Climates warm even faster. Civilization as we know it is threatened.

Intellectual tension like this could not be as stark.

If you reduce the world to one pixel, there is either too much nitrogen. Or there is too little.

Resolving this tension requires data. On the one hand, we know that N is being fixed in ever greater amounts. On the other hand, CO2 continues to increase which shifts demand for N even higher. Back again, N is raining down on ecosystems still at an elevated rate. Yet, the NO3- concentrations of water in streams is so low, stream water is approaching the NO3- concentrations of distilled water.

The only way to resolve this tension is to collect data on N availability.

Yet we need long-term measurements of N availability to know for sure whether N is becoming more or less limiting.

We don't have these.

We could use the species composition of plant communities in conjunction with indices of what plants represent low or high N availability, but again we have not invested in long-term monitoring of our plant communities.

The tension of whether the world is becoming more eutrophic or more oligotrophic has existed for a long time now.

It probably is not a bad thing to think that civilization is threatened. But we should at least know whether it is because there is too much nitrogen or too little before we try to fix it. Or else our remedies might exacerbate the situation.

Without the right data, we cannot resolve this tension. That means we start monitoring key indices like N availability and species composition now and try to answer the question in 10 years.

Or we find a different dataset that allows us to reconstruct N availability on broad spatial scales far enough back in time to discern the trajectory of N availability.

Do we have the data to resolve this tension?

I think we might...

Let's see what reviewers say.









Biogeochemical Planetary Boundary: Beyond the zone of uncertainty? (Part I)


The cycling of nitrogen in a terrestrial ecosystem determines its primary (and secondary) productivity, its diversity, and how much (and how) nitrogen is lost to the atmosphere and waters. In general, plant productivity is limited by the availability of nitrogen. Add a little more nitrogen, and not much changes. Productivity increases, but qualitatively, the ecosystem functions the same. Add a little more, and the ecosystem changes quantitatively, but not qualitatively. Productivity increases. N concentrations increase a bit, but it still is qualitatively similar to the unfertilized ecosystem.

Keep fertilizing the ecosystem with N, and eventually the ecosystem reaches a threshold. Not only does productivity increase, but a lot of other things change. Suddenly, plant N concentrations increase a lot. The plant community shifts towards plants that thrive under higher N. They have high N concentrations, they use alkaloids instead of tannins to defend themselves, their leaves are built to capture as much light as possible, rather than avoid capturing too much light. In the soil, the soil microbial community shifts and the richness of N causes N to start leaving the soils in ways it hadn't before. More NO3- comes out in the waters. More gaseous N is lost to the atmosphere.

This threshold has been repeated experimentally in individual ecosystems throughout the world. And we've seen it when we non-experimentally add a lot of N to pastures or croplands or even forests.

What we see at the plot level or even at the level of the stand or region could potentially have analogs at the planetary level. As humans fix more and more N and more and more N is added to the ecosystems, could the whole planet flip states and autocatalyze from a oligotrophic world to a eutrophic world? Could N limitation become the exception, rather than the rule.

In 2009, Rockstrom et al. published their summary of the state of the earth in respect to Planetary Boundaries (see my 2012 post on the issue here). These planetary boundaries are planet-wide environmental boundaries or ‘tipping points’. Exceed these thresholds, and humanity is at risk.

That paper was updated last year by Steffen et al. As before, the authors state that for climate change, we have entered a "zone of uncertainty" with "increasing risk". Despite all the warming, the sea level rise, the collapsing ice sheets, the potential for a shutdown of the thermohaline circulation, losses of coral reefs, thawing of permafrost, and climatic reorganization underway, their summary is that humanity is still in a safe operating space climatically.

In contrast, for the global nitrogen cycle, the status is the same as in 2009. We are apparently beyond the zone of uncertainty, and humanity is currently at high risk of exceeding a planetary threshold.

That sounds pretty dire.

But are we?

The basis for this assessment is from a recent paper by de Vries et al. 2013.

Reading the paper, apparently, for the planet to have exceeded a planetary boundary for N requires that one of the following (according to the authors) has exceeded safe operating space:

1) eutrophication of terrestrial ecosystems
2) eutrophication of marine ecosystems
3) acidification of soils and fresh waters
4) NOx, a greenhouse gas
5) ozone formation
6) groundwater contamination
7) stratospheric ozone depletion

There is really  no evidence of too much tropospheric ozone or too much groundwater contamination for humans to safely inhabit planet. Soils do not appear to be becoming acidified due to N deposition and fertilization globally. NOx levels are not deathly high. Stratospheric ozone levels are still recovering from CFC phase-outs.

Therefore, if humanity has exceeded a biogeochemical planetary boundary, then there must be evidence of planetary-scale eutrophication of terrestrial or marine ecosystems.

In a future post, I'll examine the intellectual tension about this idea...