A Wrinkled Landscape

sam_piroWant to know why hills matter to how a tropical forest works?  Ask Dr. Samantha Weintraub (at right), who successfully defended her PhD thesis today.  Here’s the short story on what she found.

Everyone who comes to my lab is subjected to my occasional (ok, sometimes more than occasional) rants on how generalizations about ecosystem function are particularly dangerous in the tropics.  The very wonder of these systems emerges from their complexity, which plays out at so many scales.  Soils vary, species vary, climate varies, landform varies….hell, it all varies at scales from smaller than you can see to longer than you can fly even in a couple of hours.  You get hundreds of different tree species in a single hectare, some with lots of nutrients in their leaves, some not, some that grow fast, some that don’t.  Here’s just one example, in an image of canopy nitrogen from our study area, generated by Greg Asner using Carnegie Airborne Observatory data:


All those different colors are all different amounts of N in the leaves of different species — big differences on a level known to matter for ecosystem function — and all happening right next to each other.  And it’s not just about the trees. Some places are flat, some are steep.  Some are really dry part of the year, some are frog-choking wet.  Some are near the ocean and its rainfall full of essential nutrients; some are not.  And on it goes.

So the kinds of generalizations you often see (tropical forests are nitrogen-rich; they have poor soils; they are X, they are Y, they are Z…) may apply in some places, or at some times, but almost certainly don’t in others.  And yeah, you can make the complexity argument anywhere for any ecosystem, but the gap between generalizations and the reality (and importance) of heterogeneity is particularly big in tropical forests.  And that importance is not just academic – if you want to know how these systems may respond to a given global change, you need to get much of this right, from trees that differ over less than a stone’s throw, to the entire Amazon and beyond.

The nice thing is, with emerging new technologies like airborne remote sensing (among others), we can finally start pulling that off.  Put another way, there was a time when broad generalizations were almost unavoidable – it was just too damn hard to actually see and quantify the variation with basic boots-on-the-ground methods.  But now we can do better.

It was into this new landscape that Sam walked.  We had already learned that the nitrogen cycle of our study region in Costa Rica didn’t seem to behave like the textbook generalization of a lowland tropical forest – in other words, there was not a lot of free N floating around.  But….we also had seen some spots in the forests that seemed to have more N than others.  Was that a species effect?   Was it something else?  We didn’t yet know.

Sam’s new insight was that maybe part of the explanation was about the shape of the landscape.  Like a lot of Central American (and other tropical) forests, those on the Osa Peninsula are not flat.  Or more accurately, some aren’t flat.  Strip the vegetation away, and what’s underneath looks like this:


In other words, there are ridges, valleys, and lots of slopes of varying steepness.  In tropical regions with this so-called “complex terrain”, it reflects the work of climate and time (and yes, life too) on the legacy of past landscape formation.  E.g., a volcano forms, stops erupting, and then water, heat and time cut it all to pieces, slowly but surely.  Or sometimes, as on the Osa, those processes that are ultimately trying to suck the land back into the sea are working against the collision of continental plates, such that uplift is taking things one way while erosion is fighting back.  In between is life, which can exert its own influence on how the landscape is formed.   For example, the roots of trees can help keep a hill from sliding away….but when a tree falls over the ripping of those roots from below can also set off a pulse of soil loss.

Put it all together, and you get places that are, well, wrinkled.  Sometimes those wrinkles are densely packed and deep; sometimes those signs of age are less ever-present, with expanses of relatively smooth landscapes occasionally interrupted by an incutting cleft that foreshadows greater things to come.

And those wrinkles matter.  The trees change in form, size and identity.  The soils can look different….and indeed they are in ways that matter to how the ecosystem works.  For example, past work by Peter Vitousek, Stephen Porder, Greg Asner and others had shown that erosion effects how much phosphorus might be available in a tropical forests. Here’s an example from their work on Kauai, showing P in the leaves of Hawaiian forests:

Screen Shot 2014-04-04 at 5.55.56 PM

On stable ridges that are the remnants of a long dead volcano (purple colors), where the soils have been able to sit in place for a looooong time and just chemically weather away, most of the P is locked up or gone.  That’s the standard mantra for lowland tropical forests – old soils, highly weathered, P in short supply.  But move just a few hundred meters away onto a steep slope (yellow shades), and holy crap, lots more P!  Why?  Because those steep slopes erode….and when they do, the rock from below comes back up in closer contact with the surface soil and vegetation, and presto, new bank account of P.  This effect of erosion can be a big deal – the differences in the image above are on par with those created across millions of years of soil development.

OK fine, but does this matter for nitrogen?  Because, for the most part that doesn’t come from the rocks below, but from the air above.  It’s fixed or rained out from the atmosphere.  But still…could geomorphology matter?  That was Sam’s essential question.  She had wandered about the wrinkled landscape of the Osa, she had read an influential paper by Ron Amundson and others which contained some discussion of hillslopes and the nitrogen cycle, and she began to be convinced this could matter in tropical forests.

And turns out it does….but in the opposite way that erosion affects phosphorus.  In the Osa soils beneath that colorful canopy nitrogen image above, flat, broad ridgetops are rich in N….and the steep slopes next to them are much more N poor.  Here’s a whole bunch of her work boiled down into one graph, in which 8 different metrics of the nitrogen cycle are combined into a principal components analysis, with the three different shades being the flat ridges (RDG), the initial transition to steeper slopes (RST) and the steep slopes themselves (SLP).  For those of you who don’t speak-a-da PCA, this just says that on the slopes nitrogen is in much lower supply than on the ridges…with the transition zone falling in between.


Why?  Erosion again, as Sam showed in follow up work.  But this time, while the loss of that surface soil during erosion may expose new rock from below to enrich the P cycle….it also removes organic matter from above, and by doing so prevents N from accumulating as the ecosystem develops.   It’s like a cruel game of keep away — N comes in from above in biological fixation or deposition, starts to build up for a while, but wooosh suckah, a good dose of erosion sends it right down the hill and out of the system and everybody’s gotta start over again.  In effect, soil development starts over – or is at least chronically counter-acted by erosion – such that steep slopes are younger in their time course of development….and right in line with classic ecosystem theory, they are more N poor.

Figuring this out took Sam a hell of a lot of work, doing things like digging big holes in the ground right during the month when a meter of rain was falling on the joint:

sam october soil pit

But it paid off.  A really great piece of work that opened up whole new lines of inquiry in our lab.  In a collaboration with Bob Anderson here at CU, Cory Cleveland at Montana, Stephen Porder at Brown and Greg Asner at Carnegie – and all of their respective teams – we are trying to quantify how variation in species, landform and climate all interact across the region to set up a pastiche of different mini-ecosystems that in turn have distinct biogeochemical characteristics.  Ultimately, the hope is to take that knowledge and be able to generalize again….but in ways that account for the real variation these remarkable forests contain, at scales from individual trees up to hundreds of miles.

Congrats to the now Dr. Sam!  Well earned.


2 thoughts on “A Wrinkled Landscape

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s