Environmental and physiological drivers of tree growth : a pan-tropical study of stable isotopes in tree rings

Research output: Thesisinternal PhD, WU



Forests in the wet tropics harbour an incredible biodiversity, provide many ecosystem services and regulate climatic conditions on regional scales. Tropical forests are also a major component of the global carbon cycle, storing 25% of the total terrestrial carbon and accounting for a third of net primary production. This means that changes in forest structure and forest cover in the wet tropics will not only affect biodiversity and ecosystem services, but also have implications for the global carbon cycle and – as a result – may speed up or slow down global warming. Deforestation rates are still high in the tropics and have profoundly affected the extent of forests in many countries. Additionally, there are indications that undisturbed and pristine tropical forests are changing. The most notable changes found by the monitoring of permanent forest plots are an increase of tree growth and forest biomass per unit of surface area over the last decades. If this is indeed the case, it would entail that the world’s tropical forests are potentially absorbing a significant fraction of human caused CO2emissions and as such are mitigating global warming. However, increased tree growth and forest biomass have not been found in all studies. Furthermore, it is unknown whether the observed changes in intact forests are part of a long-term change, or merely reflect decadal scale fluctuations. These uncertainties lead to an ongoing debate on whether tree growth and forest biomass have increased in tropical forests and – if so – to what extent. In addition, there is also a scientific discussion on the factor(s) that could underlie the potential changes in tree growth and forest biomass. Possibly, they are caused by an internal driver, like the lasting effect of large scale disturbances in the past, or by external drivers. Possible external factors affecting tropical forest dynamics are (1) climate change (temperature and precipitation), (2) increased nutrient depositions and (3) increased atmospheric CO2concentration.


In this thesis, I investigated the environmental changes that could have formed the basis for changes in tropical tree growth. I used two relatively new tools in tropical forest ecology: tree-ring measurements and stable isotope analyses. Tree-ring widths were measured to obtain long-term information on tree growth. Stable isotopes in the wood of tree rings were analysed to provide information on the environmental and physiological drivers of tree growth changes. This thesis is part of a larger project on the long-term changes in intact forests in the wet tropics (the TroFoClim project, led by Pieter Zuidema) and also includes the PhD theses of Mart Vlam and Peter Groenendijk. In this project, ~1400 trees of 15 species were examined that were collected in three forest sites distributed across the tropics (in Bolivia, Cameroon and Thailand).


For the assessment of long-term changes in growth and stable isotopes, it is important to understand shorter term fluctuations due to forest dynamics (i.e. gap formation), because these interfere with changes on a longer temporal scale. The formation of a gap in a closed canopy forest, after the death of a tree, can cause considerable environmental changes in the surrounding area, e.g. in light, nutrient and water availability. This can strongly affect the growth rates of the remaining trees. However, in most studies the environmental drivers of changes in tree growth after gap formation are not considered. In CHAPTER 2 I measured carbon isotope discrimination (Δ13C) in annual growth rings of Peltogyne cf.heterophylla, from a moist forest in North-eastern Bolivia, and evaluated the environmental drivers of growth responses after gap formation. Growth and Δ13C was compared between the seven years before and after gap formation. Forty-two trees of different sizes were studied, half of which grew close (<10m) to single tree-fall gaps; the other half grew more than 40 m away from gaps (control trees). I found that increased growth was mainly associated with decreased Δ13C suggesting that this response was driven by increased light availability and not by improved water availability. Interestingly, most small trees did not show a growth stimulation after gap formation. This result was hypothesized to be caused by an increased drought stress. However, the measurement of Δ13C showed that increased water stress is unlikely the cause for the absence of increased growth, but rather suggested that light conditions had not improved after gap formation. These results show that combining growth rates with changes in Δ13Cis a valuable tool to better understand the causes of temporal variation in tree growth.


An important potential driver of long-term changes in tree growth is climate change, e.g. global warming and altered annual precipitation. To understand the effect of climate change on tree growth, the availability of reliable data on historical climate is off course crucial. For the study areas in Bolivia and Thailand, previous studies have investigated the occurrence of temporal trends in temperature and precipitation. For the study area in Cameroon however, as well as for West and Central Africa in general, the availability of instrumental climate data is very restricted. This limits the possibility to relate past climatic variation to changes in tree growth and calls for proxies that allow reconstruction of past climatic conditions. In CHAPTER 3 I assessed the potential use of stable isotopes of oxygen (δ18O) in tree rings as a tool for the reconstruction of precipitation in tropical Africa. I measured δ18O in tree rings of five large Entandrophragmautiletreesfrom North-western Cameroon. A significant negative correlation was found between annual tree-ring δ18O values (averaged over the five individuals) and annual precipitation amount during 1930-2009 in large areas of West and Central Africa. I also found tree-ring δ18O to track sea surface temperatures (SST) in the Gulf of Guinea (1930-2009). These two results are related because rainfall variabilityin West and Central Africa is profoundly influenced by the SST of the tropical AtlanticOcean. Thus a high SST in the Gulf of Guinea is associated with high precipitation over large parts of West and Central Africa and recorded in tree rings by a relatively low δ18O value. On the other hand, dry years when SST is low, are recorded by relatively high tree-ring δ18O values. I also found a significant long-term increase of tree-ring δ18O values. This trend seems to be caused by lowered precipitation from 1970 to 1990 (the Sahel drought period). From 1860 to 1970, no significant long-term trend was observed in tree-ring δ18O values, suggesting no substantial change in precipitation amount occurred over this period.


Another potential driver of altered tree growth and biomass in intact tropical forests is the increase of anthropogenic nutrient depositions, especially nitrogen. The deposition of nitrogen has likely risen due to an increased industrialization and use of artificial N fertilizers in most tropical countries. Nitrogen can stimulate plant growth, as is well known from the positive effect of N fertilizer application on crop yields. Previous studies have shown that the stable isotope ratio of nitrogen (δ15N) increased during the last decennia in the wood of trees from Brazil and Thailand as well as in tree leaves from Panama. This increased δ15N has been interpreted as a signal that tropical nitrogen cycles have become more ‘open’ and ‘leaky’ during the last decades in response to increased anthropogenic nitrogen depositions. The underlying mechanism is that high rates of nitrogen deposition and high ambient nitrogen availability lead to an increased nitrification. This process can cause a gradual 15N-enrichment of soil nitrogen. In CHAPTER 4 I analysed changes in tree-ring δ15N values of 400 trees of six species from the three study sites. In the trees from Cameroon no long-term change of tree-ring δ15N values was found (1850-2005), even though NH3and NOxemissions seem to have increased strongly around the study area since 1970. Possibly, the very high precipitation at that site causes the local nitrogen cycle to be already very ‘leaky’, limiting the effect of additional nitrogen input on the δ15N signature of soil nitrogen. Alternatively, nitrogen input in this forest might be much lower than reconstructed NH3and NOxemissions suggest. For the study site in Bolivia, no significant change of tree-ring δ15N values was found (1875-2005), which is in line with the expected result for areas with a low anthropogenic nitrogen input. I found a marginally significant increased of δ15N values since 1950 in trees from Thailand, which confirms previous observations. This points to an effect of increased anthropogenic nitrogen deposition, which could have stimulated photosynthetic rates, if indeed nitrogen was limiting tree growth.


The most often hypothesized factor to cause a long-term increase of tree growth is the rise of atmospheric CO2. Since the onset of the industrial revolution (~1850) global atmospheric CO2concentration has increased by 40%. Elevated CO2can directly affect plants by increasing the activity, as well as the efficiency, of the CO2fixing enzyme rubisco and thereby increase photosynthetic rates. Potentially more important in plant communities subjected to periods of limited water availability (like a dry season) is that elevated CO2 can cause a reduction of stomatal conductance, which lowers evapo-transpiration and hence reduces water losses.This increases water-use efficiency (i.e. the amount of carbon gained through photosynthesis divided by the amount of water loss through transpiration) and might allow plants to extend their growth season and/or increase their photosynthetic activity during the hottest hours of the day when water-stress might be severe. Elevated atmospheric CO2is thus a very likely candidate to have stimulated tropical tree growth (also referred to as CO2fertilization), provided at least that plant growth was either carbon or water limited. In CHAPTER 5 I tested the CO2fertilization hypothesis by analysing growth-ring data of 1100 trees from the three study sites. The measurement of tree-ring widths allowed an assessment of historical growth rates, whereas stable carbon isotopes (δ13C) in the wood of the trees were used to obtain an estimate of the CO2concentration in the intercellular spaces in leaves (Ci) and of water-use efficiency (intrinsic water-use efficiency; iWUE). I used a sampling method that controls for ontogenetic (i.e. size developmental) changes in growth and δ13C. With this method, trees were compared across a fixed diameter (i.e. same ontogenetic stage). I chose two diameters: 8 cm (referring to small understorey trees) and 27 cm (referring to larger canopy trees). A mixed-effect model revealeda highly significant and exponential increase of Ciat each of the three sites, and in both understorey and canopy trees. Over the last 150 years Ciincreased by 43% and 53% for understorey and canopy trees respectively. Yet, the rate of increase in Ciwas consistently lower than that of atmospheric CO2. This ‘active’ response to elevated atmospheric CO2resulted in a significant and large increase of iWUE. Over the last 150 years, iWUE increased by 30% and 35% for understorey and canopy trees.A long-term increase of iWUE indicates either a proportional increase of net photosynthesis and/or a decrease of stomatal conductance and thus transpiration, both of which could have stimulated biomass growth. However, I found no increase of tree growth over the last 150 years in any of the sites. Although there are several possible explanations for these findings, I argue that it is most likely that tropical tree growth is generally not limited by water and carbon, but by a persistent nutrient limitation (e.g. of phosphates) and that this has prevented tropical trees to use the extra CO2for an acceleration of growth.


In this thesis I have studied environmental and physiological drivers of tree growth changes. I found evidence of decreased precipitation over the last decades at the study site in Cameroon (CHAPTER 3), a changed nitrogen cycle at the study site in Thailand (CHAPTER 4) and an overall change in the physiology of all tree species studied (increased iWUE; CHAPTER 5). One of the main findings of this thesis is however, that these changes have not led to a net change of tree growth over the last 150 years (CHAPTER 5). This is an important finding that could have two major implications. Firstly, the absence of a long-term growth stimulation suggests that the increase of iWUE is mainly driven by a reduced stomatal conductance, which likely leads to a reduced evaporative water loss. If trees across the tropics are reducing evapo-transpiration, this will change affect hydrological cycles, e.g. leading to a lower humidity, higher air temperatures and a reduced precipitation. Secondly, the absence of a growth stimulation over the last 150 years suggests that the carbon sink capacity of tropical forests is currently overestimated, e.g. by Dynamic Global Vegetation Models, which assume strong CO2fertilization effects and as such a high capacity of tropical forests to mitigate global warming. I anticipate that the planned Free Air Concentration Enrichment (FACE) experiments in the tropics will shed light on the reasons why increased CO2does not stimulate the growth rates of tropical trees. Furthermore, I argue that combining tree-ring measurements and stable isotope analyses together with permanent plot research is the most promising way to increase our understanding of the changes in tropical forests.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
  • Zuidema, Pieter, Promotor
  • Bongers, Frans, Promotor
  • Anten, Niels, Promotor
Award date30 Jun 2014
Place of PublicationWageningen
Print ISBNs9789461739544
Publication statusPublished - 2014


  • trees
  • growth
  • plant physiology
  • growth rings
  • environmental impact
  • isotopes
  • tropical forests
  • tropics


Dive into the research topics of 'Environmental and physiological drivers of tree growth : a pan-tropical study of stable isotopes in tree rings'. Together they form a unique fingerprint.

Cite this