Saturday, January 23, 2010

Biogeography and traits

A map of climates in the Cape Floristic region and how they relate to shoot design for 88 species of Leucadendron.

Research on species traits generally follows one of three avenues: traits are related to other traits, environmental variables where the plants were measured, or phylogenetic relationships.
Although we've made good strides over the past decade in understanding trait relationships, relationships with environmental gradients, and understanding the distribution of traits over phylogenetic space, we have progressed little in understanding how traits affect the distribution and abundance of species. A fourth road is just as important, but less frequently traveled: biogeographic distributions of species.

As far as I can tell, there is really only one paper to attempt this. Thuiller et al. (2004) measured traits on 88 species of Leucadendron (proteas), and then derived the biogeographic distribution of each species. With these data, they generated mean climate parameters for their occurrence.

The simple pairings led to some novel insights. First, species in drier, more Mediterranean regions had smaller leaves and smaller stems, but also smaller cones. This might not be a surprise to many, but sometimes we need ways to quantify the obvious. More interesting, species in more arid regions had narrower niche breadths and were more likely to have seeds dispersed by wind, rather than by ants, which were more prevalent in the more subtropical, continental climates.

In all, I'd have to add this paper to my list of needle-in-haystack papers, but it's actually been cited relatively well (42 cites since 2004), more than my arbitrary cap of 5 citations a year. That said, the specific approach hasn't penetrated trait research. As a consequence, research on trait distributions have little mechanistic underpinnings. The Thuiller paper is a good model for future research that hasn't caught on yet.

Thuiller, W., S. Lavorel, G. Midgley, S. Lavergne, and T. Rebelo. 2004. Relating plant traits and species distributions along bioclimatic gradients for 88 Leucadendron taxa. Ecology 85:1688-1699.

Saturday, January 16, 2010

The “Nitrogen Throttle” hypothesis of primary productivity

Imagine a global change scenario. Models predict precipitation will decline for a region. Other things are likely to happen, too. For example, N deposition is likely to increase. But, with soil water determining productivity, the model returns the prediction that productivity will decline, too. As a result, ranchers will fail. Forestry will be diminished. But what if the model was structured wrong? What if soil moisture only indirectly determined productivity and N was actually more limiting? The decline in precipitation might be less important that the increase in N. The original prediction might be 100% wrong.

Globally, nitrogen is the nutrient that limits primary productivity the most. Also, secondary productivity, since protein is often more limiting to herbivores than energy. Yet, whether nitrogen or water limits productivity is still open. Places with little precipitation have little plant biomass. Yet, even vegetation in dry places still responds to N addition.

Separating between water and nitrogen limitation is complex enough that for all intents and purposes, ecologists have ceased to try. Not that it doesn’t matter any more—those that were interested in the topic have likely just stopped hitting intellectual walls. And some simple scenarios show how important it is to get the fundamentals right.

There are two major hypotheses regarding water and N limitation. First, the “Water Stress” hypothesis states that water directly limits production. As soil moisture declines, water supply to plants declines, water stress increases, and plants reduce stomatal conductance to match supply and demand and/or limit cavitation risk.

The other hypothesis is the “Nitrogen Throttle” hypothesis. As soil moisture declines, water stress increases for plants, but this is not what limits photosynthesis. Instead, microbial N mineralization declines causing N supply to plants to decline. Here, low moisture ‘throttles’ N mineralization. As a result of the lowered N supply, plants decrease their stomatal conductance in order to match C and N supplies. Under more extreme soil moisture stress, plants begin to senesce their leaves to maintain minimum growth requirements if not store N for when favorable growing conditions return.

How important is each hypothesis in explaining patterns of productivity? Hard to know, but as I’ve described before there is evidence that nitrogen throttling happens and is important. For example, why would there be soil moisture left deeper in the soil profile if plants were limited by water? Why is it that plants that can photosynthesize at -9 MPa water pressure begin to senesce at -3 MPa?

Like other competing hypotheses, proving one over the other is all but impossible. Factorial resource additions can’t solve the problem when adding water also increases N supply. In the end, it’ll be parsimony that will be tested as lines of evidence are compared.

Thursday, January 14, 2010

A new Whittaker biome diagram


Whittaker biome diagram from Chapin, Matson, Mooney Ecosystem Ecology text.

Whittaker long ago attempted to explain the major patterns of vegetation in the world with combinations of temperature and precipitation. The Whittaker biome diagram is a fundamental starting point for understanding the vegetation of the world.

There are general questions about the overarching role of climate in determining biomes vs. other state and interactive factors, as well as what the boundaries should be and how much to subdivide biomes.

With recent advances in our understanding of the distribution of climate across the globe, we can now see that some of the patterns were not detailed initially correctly. Andrew Elmore and I redrew the Whittaker biome diagram to also include the actual distribution of land area for each combination of temperature and precipitation.


A few major things change.

1) Tropical forests exist in areas much wetter than originally detailed. Much forest exists between 4.5 and 7 m of rain.

2) Most of the world's temperate wet forests are at about 4°C. Whittaker would have lopped off much of them.

3) There are scattered high precipitation areas between what was considered temperate and tropical wet forests. This happens to largely be Hawaii. These have never been classified into temperate vs. tropical biomes.


Tuesday, January 12, 2010

Why the world is +5


Average del15N of all the soil and vegetation in the world.

Ben Houlton and Edith Bai have a new paper in PNAS about the nitrogen cycle that has some good sleuthing in it.

Examining global patterns of soil and plant 15N, it turns out the 15N signature of all the N in non-managed ecosystems is about +5‰. But why +5 if naturally fixed N is about 0‰? They rule out enrichment from N deposited as NO3-, which averages about -2 to 0. And terrestrial systems aren’t becoming enriched because of fractionation during leaching out the bottom-- N being leached into streams scales with soil 15N pretty well.

What’s left to explain the enrichment? Just gaseous N loss. They argue its unlikely NH3 volatilization since its too reactive and gets redeposited too quick. With that ruled out, all that’s left is denitrification, which here includes losses during nitrification.

From here, its some simple back of the envelope calculations to estimate the total gaseous N loss on a global scale, which happens to be a lot.

The authors aren’t clear about where the largest uncertainties in their calculations are, but global patterns of soil 15N and fixation likely are pretty high up there. Also, considering that soil 15N is pretty close to zero in cold ecosystems, that leaves the warm ecosystems as the major players in the global N fluxes, which happen to be the places that we know the least about patterns of 15N.

Sounds like now is the hard part. Because if the world turns out to be +6 or +4, the budget changes a lot.

Houlton, B. Z. and E. Bai. 2009. Imprint of denitrifying bacteria on the global terrestrial biosphere. Proceedings of the National Academy of Sciences of the United States of America 106:21713-21716.

Wednesday, January 6, 2010

Whither elevated CO2 research?

Biocon FACE site in Minnesota.

The list of major global changes is short. The world is becoming warmer, CO2 concentrations are rising, N deposition is increasing, agriculture continues to increase, and the world’s flora is becoming homogenized.

The list of major global change research efforts is even shorter. For all intents and purposes its now all about climate. Our research into understanding the extent and implications of N deposition is greatly diminished and comparably small. Land use we largely quantify with remote sensing but do little else. Invasive species are approached piecemeal, but not with any major initiatives.

The surprising fall from grace has been research into elevated CO2. The politics behind this are one thing, but there is no indication that this remains a major research question. Do we understand the effects of elevated CO2 on ecosystems? Not even close. Does it seem like we do? For some, yes.

The reason the energy into elevated CO2 research has waned derives from the early focus of the research. Early research centered less on testing mechanistic hypotheses than on quantifying response ratios. The modeling community wanted to know how much will photosynthesis and NPP increase. For all intents and purposes, the Ainsworth and Long (2005) meta-analysis of responses of photosynthesis to elevated CO2 was, for some, the last important paper on the topic. I’m oversimplifying, but--with increases in elevated CO2, photosynthesis goes up 25%, stomatal conductance goes down about the same amount.

With that synthesis, with those response ratios, so withered the major impetus into elevated CO2. And with that, whither the research?

Without a major, overarching question, it is next to impossible to generate the funding necessary to sustain the research. The previous search for response ratios has fractured research into components that seek out other response ratios, but none of the remaining response ratios generate that much excitement.

In short, to reinvigorate CO2 research, look at the fundamental unanswered questions and question whether the current response ratios can be trusted. (This is largely what fueled a large round of research shifting from chambers to FACE technology.)

Here’s what we know:

Elevated CO2 increases water use efficiency.

Elevated CO2 increases nitrogen use efficiency.

Given these two points, here’s what we don’t know:

Elevated CO2 increases/decreases the relative limitation of water vs. nitrogen.

This is the major unanswered question for elevated CO2 research, because it’s one of the major question we have left for modern CO2 concentrations. If we do not understand the relative limitation of water and N to productivity and C storage now, and we know less about how CO2 will alter the balance, then we really can’t trust the average response from modern experiments if the relative availability of water and N will change in the future. And right now, from our experiments, we just do not know whether CO2 increases photosynthesis because more water is left in the soil or less N is needed to grow. Considering that if more water is left in the soil more N should be made available to plants, we might know what the average response is today, but not what it will be tomorrow. Without knowing why, we can't know what.

Surely, there are other major questions to go forward with, but without something as large as this, elevated CO2 research will continue to wither.