The size of trees in tropical rainforests

This summer, at the start of my research at the Institute for Tropical Ecology and Conservation (ITEC) field station, I went to one of my field sites only to find that someone had cut down a really large tree.  The disconcerting thing was that I didn’t even remember there having been a big tree there despite my having passed by it many times.  I guess sometimes you can’t see the trees for the forest.

That I overlooked a relatively large tree might suggest to those who have never been to a lowland tropical rainforest that there must be a lot of large trees there.  That’s reasonable and there are indeed lots of large trees in the forests where I teach and do research.  These include amazingly straight, tall, large diameter trees whose first branches are nearly out-of-sight above the forest floor along with the occasional massive big-buttressed tree that has to be seen to be believed.

But these forests aren’t at all the cathedral forests of the Pacific Northwest where tree after tree is a giant. A figure in an interesting paper by Feldpausch et al. (2012) shows this. Researchers in this paper compiled tree height and diameter data from 327 relatively undisturbed tropical rainforest plots on four continents (Africa, Asia, Australia, and South America) to look at the contribution of trees of different diameter to the total above ground biomass (AGB in the below figure) in forests from different regions. Figure 3 in the paper (click on to enlarge) summarizes the part of their results relevant to this discussion:

The bars show the contribution of different diameter classes (diameter at breast height — 1.3 meters above the ground) to the above ground biomass for the different forests per hectare (the two bars for each diameter class reflect two models for estimating this with the second incorporating a correction based on regional tree heights).  The vertical dashed line is the midpoint for above ground biomass (half of the total above ground biomass is above this bin while the other half was below).  Finally, there are two curves (again without and with a correction based on regional tree heights) reflecting cumulative above ground biomass with the addition of each larger diameter class.

The figure shows something I have already mentioned.  There are large trees in a typical tropical rainforest.  That the largest diameter classes in this figure are collapsed into a single 160+ centimeter diameter class actually obscures just how giant some trees can be!

For me, though, the remarkable thing about this figure is how high the bars are on the left side of each plot. This means that despite their relatively small size, small diameter trees contribute significantly to above ground biomass. The only reasonable explanation for this is that there are lots of relatively small trees in these forests. Note too that the smallest diameter size class is 10-20 centimeters. These are small trees (in forestry terms these would be classified as pole timber): saplings and other smaller diameter vegetation aren’t included in this figure and don’t inflate the contribution of small trees to total above ground biomass.

The authors of the paper point out that, overall, trees with diameters of 40 centimeters or less (basically small sawtimber to pole timber) make up about a third of the above ground biomass across all of these forests. The most extreme forests in this regard are those of Western Amazon and the Brazilian Shield where trees with diameters of 30-40 centimeters or less make up half of the total above ground biomass.  I don’t know why this is the case for the forests of the Brazilian Shield but forests in the Western Amazon often sit on shallow, frequently waterlogged soils.  This results in frequent treefalls and lots of regeneration giving rise to small trees.

Not only do two South American regions have the largest fraction of total above ground biomass in small trees, the figure suggests that as a whole, South American forests have more small trees than other tropical rainforest regions. The authors don’t address why this is but I wonder if palms play some role. Palms are pantropical and are as diverse in Asian forests as in South American ones.  The difference is that they are much more likely to exist as small trees (diameters >10 cm) in the Neotropics than in African or Asian forests.  If so, the proportion of above ground biomass in small trees might actually be underestimated in Feldpausch et al.’s (2012) paper.  That’s because this paper uses a “dicot” model to relate tree height to diameter and the combined measures to biomass; palms tend to be taller than their dicot counterparts for a given diameter. Also if true, we might expect to see the “South American pattern” in Central America too.

Interestingly, I might have guessed that the Panamanian forests where I do research have their above ground biomass distributed among trees of different diameters like what’s seen in South America.  That’s interesting because I would have thought that that distribution is a result of logging that took place 40-60 years ago.  As for myself, I would have thought that a relatively undisturbed forest would have had its above ground biomass sorted out more like what’s seen in the figure for African rainforests.  Even here, though, I am still surprised by the contribution of small trees to above ground biomass.  I realize that there have to be small trees to give a tropical rainforest its multistoried aspect; I just didn’t realize how many there must be.

Finally, what if I am right about the relatively recently disturbed forests where I do my research having its biomass distributed among trees like what’s seen in undisturbed forests?  It certainly could be just happenstance based on how trees were selected when logging took place. On the other hand, many forest processes occur surprisingly fast in tropical rainforests.  On relatively fertile soil sites tree growth can be quite rapid and living biomass can completely turn over in several decades. Could it be also that the way that biomass is distributed among tree diameter classes also recovers quickly?

At any rate, the next time you find yourself admiring some massive canopy emergent in a virgin tropical rainforest, make sure to turn around and take a look at all the little trees that really are a significant part of the forest.



Brown, S. and A.E. Lugo. 1990. Tropical secondary forests.  Journal of Tropical Ecology 6: 1-32.

Goodman, R.C., O.L. Phillips, D. Del Castillo Torres, L. Freitas, S.T. Cortese, A. Monteagudo, and T.R. Baker. 2013. Amazon palm biomass and allometry. Forest Ecology and Management 310: 994 – 1004.

Emmons, L.H. and A. H. Gentry. 1983. Tropical forest structure and the distribution of gliding and prehensile-tailed vertebrates. The American Naturalist 121: 513-524.

Feldpausch, T.R., J. Lloyd, S.L. Lewis, R.J.W. Brienen, M. Gloor, A. Monteagudo Mendoza, et al. 2012. Tree height integrated into pantropical forest biomass estimates. Biogeosciences 9: 3381–3403.

Finegan, B.  1996.  Pattern and process in tropical secondary rain forests:  the first 100 years of succession.  Trends in Ecology and Evolution 11:  119-124.

Pan,Y.,  R.A. Birdsey, O.L. Phillips, and R.B. Jackson.  2013. The structure, distribution,
and biomass of the world’s forests.  Annual Review of Ecology and Systematics 44: 593-622.




A tropical cyclone in Panama and the unusual case of Typhoon Vamei

At the Institute for Tropical Ecology and Conservation (ITEC) we occasionally take our students into Bocas Town on a Friday or Saturday afternoon.  While there, every now and then a student will notice that buildings everywhere are along the water’s edge and will wonder out loud if the area might be susceptible to obliteration by a hurricane.

It’s a more interesting question than you might think.  What makes it interesting is a certain Hurricane Martha.

Continue reading

Why it rains in the ITCZ

The tropical rain belt is responsible for a significant portion of the rainfall that occurs in the tropics. How much so can be seen in the following photograph showing 30 day rainfall averages:

Formerly, meteorologists referred to the entire tropical rain belt as the intertropical convergence zone (ITCZ).  Today, more and more meteorologists are restricting use of this term to refer to that part of the tropical rain belt found over oceanic regions (see below for why) and it is in this sense that I’ll be talking about it.

From space, the ITCZ is often visible as a rough band of cloudiness (up to 300-500 nautical miles wide) that is generally tropical in location and roughly parallels the equator.  Below is a photograph showing part of the ITCZ in the eastern Pacific:

The ITCZ is often associated with rainfall in the form of showers or atmospheric convection.  Atmospheric convection? On land, atmospheric convection is basically synonymous with thunderstorm. That’s not the case over the open ocean.  Meteorologists have recently come to realize that the amount of lightning a storm produces is related to the strength of the storm’s updrafts and downdrafts.  Apparently, updrafts and downdrafts in oceanic cumulonimbus clouds aren’t as strong as they are in cumulonimbus clouds on or near land. Consequently, despite towering as high and often higher, oceanic cumulonimbus clouds aren’t nearly as apt to produce lightning and associated thunder.  Since the ITCZ is now being thought of as a climatic feature largely limited to oceanic areas,  it wouldn’t be entirely accurate to say that it is frequently associated with thunderstorms.

Continue reading