Ariel E. Lugo and Sandra Brown
ARIEL E. LUGO is Project Leader at the Institute of Tropical Forestry of the United States Forest Service, Río Piedras, Puerto Rico. SANDRA BROWN is a scientist at the Center for Wetlands of the University of Florida, Gainesville, Florida. This work was sponsored by the US Dept of Energy for the Center for Wetlands.
Tropical forests affect the carbon balance of the world's atmosphere by: storing large amounts of carbon, doing so at a rapid rate, releasing carbon (as carbon dioxide) to the atmosphere through respiration, exporting organic matter to deep aquifers or oceanic ecosystems, or by their response to human uses which include the extraction of wood, forest destruction, or forest management for net growth. Those who believe that tropical forests are sources of carbon to the atmosphere argue that rapid rates of deforestation reduce carbon storage and increase the amount of CO2 going into the atmosphere due to burning and decomposition of vegetation. However, believers that tropical forest ecosystems are sinks of carbon emphasize the rapid rates of succession and high net CO2 uptake of tropical forests.
A balanced view of the issue must consider both sides of the argument as depicted in Figure 1 which is a diagram of land uses and their implications for the carbon balance. Climax forests (depicted at top of Figure 1) are presumed to be in balance with respect to carbon uptake and release. More carbon is released to the atmosphere than taken up when human or natural forces convert the climax forest into any of three possible states. The least disturbed is termed a successional or secondary forest, next is a clear-cut or deforested area, and below it is an area dedicated to permanent non-forest use. Net car bon release to the atmosphere in creases downward along this pathway of changing land use. Some investigators have assumed that this is virtually the only pathway operating in tropical regions, that 95 percent of the carbon in biomass in areas being deforested reaches the atmosphere as CO2, and that succession is negligible.
The left side of the diagram shows the successional pathway from disturbed forest to a climax state. Net carbon uptake characterizes this pathway. The rate of uptake depends on environmental conditions and speed of succession. Arid environments exhibit slow rates of succession, while humid environments exhibit much faster rates. Obviously, the secondary forest recovers more quickly than the completely deforested site, while land being dedicated to non-forest uses has little opportunity to act as a net carbon sink. The balance between the downward vs upward changes in land use depicted in Figure 1 dictates whether a tropical region is a source or sink of carbon. The objective of this paper is to explore the forces and factors that may tilt the balance.
A stable steady state condition of atmospheric carbon occurs when humans are completely dependent on solar energy for survival. Under these conditions, humans are simple consumers and may use traditional shifting cultivation, hunting, or gathering as a means of procuring food. The carbon balance is maintained because the intensity of human activity is not high enough to offset the regrowth of vegetation. Forest succession easily takes up the CO2 produced by the periodic clearing of vegetation. In general, the vegetation is considered to be in a "climax" steady state and large areas of virgin forests buffer human activity. Are these forests also in a carbon balance with the atmosphere?
Whether climax forests are or are not in a perfect carbon balance with the atmosphere is a question that is yet to be answered conclusively. Usually, one assumes they are, but all ecosystems are open systems and thus they should not be expected to be perfectly balanced. Climax forest stands in the Luquillo Experimental Forest in Puerto Rico, for example, exhibited a small but measurable loss of dissolved organic matter to downstream ecosystems (Odum, 1970).
Odum measured up to 15 mg/1 of dissolved organic matter in run-off waters during periods of heavy rains, but he only made four direct determinations. In a one-year study of watersheds in the lowland forests of Guatemala, Brinson (1973) measured high dissolved organic matter concentrations in rivers (up to 36.8 mg/1) and found these to be proportional to stream discharge.
A MIXED TROPICAL FOREST IN SURINAME, the carbon question is just as complex
High concentrations of organic matter in waters are usually associated with wetland ecosystems. For exam pie, streams draining watersheds with extensive wetland systems exhibit dissolved organic matter concentrations of 3 to 72 mg/l while those that drain watersheds with few wetlands exhibit dissolved organic matter concentrations of 1.8 to 12.4 mg/l. Black mangrove forests in Florida export about 0.7 g organic matter/m²/day (Twilley, 1980).
Forest areas and their carbon storage in Puerto Rico at different times in history
|
Date |
Forest area (10²) |
Carbon storage equivalence¹ (106 mt) |
Source |
|
At the time of discovery (1493) |
28.5 |
93.8 |
Zon and Sparhawk, 1923 |
|
1916 |
1.787 |
19.7 |
Zon and Sparhawk, 1923 |
|
1950 |
2.352 |
26.0 |
Department of Natural Resources Inventory Program |
|
1973 |
3.758 |
41.5 |
Department of Natural Resources Inventory Program |
¹Estimate for 1493 is based on Life Zone distribution of Puerto Rico and biomass storage in plants and litter from Brown and Lugo (1980). For other years, estimates assumed equal rates of deforestation and growth for all Life Zones. - 2Assuming 95 percent of island forested.
These data usually yield low absolute amounts of carbon when extrapolated to a global scale. The Amazon basin, for example, has been reported to export 0.2 × 1014 g carbon/yr based on dry season measurements of dissolved carbon concentrations and annual river discharge (Williams, 1968). But, recently this estimate was revised upward to about 1014 g carbon/yr (Richey et al., 1980).
These data suggest that even "balanced" ecosystems export organic matter to downstream aquatic ecosystems where carbon may be sequestered for long periods. The global significance of these exports still awaits elucidation because the data base does not permit a reliable calculation. For example, the calculation for the Amazon river of Williams (1968) is based on 12 determinations, none at high flood stage. Richey et al.'s (1980) estimate was based on sampling during the beginning of the rainy season when river waters were rising and during peak flooding. For those rivers for which data are available, it is clear that export of organic matter is proportional to discharge. Of more significance, however, are the unusual events that escape sampling. In Figure 2 such an event is depicted for a river in Puerto Rico which discharged 70 percent of its annual sediment load and 10 percent of its annual water discharge in one day (27 November 1968). Until these events are taken into consideration, particularly in the wet tropics, it will be difficult to disprove the hypothesis that a significant global sink of carbon exists in the form of organic matter that is produced in terrestrial ecosystems and flows to the ocean as dissolved carbon through aquifers and rivers. While it is true that this organic input is partially respired by marine organisms, it is equally possible that CO2 thus produced does not reach the atmosphere.
If humans have additional energy sources available, population densities can increase above natural carrying capacity and humans can exploit natural systems more fully (Figure 3). We argue that this leads to long-term oscillations in terms of net. carbon uptake or release by the landscape. Examples from two countries illustrate the point.
In Costa Rica, a country that is entering a phase of rapid development of its lands because of increasing fossil fuel use, deforestation is proceeding at fast rates and the country may be a net carbon yielder to the atmosphere. In Puerto Rico, deforestation occurred much earlier, and fossil fuels long ago replaced fuelwood as the main source of energy. Now Puerto Rico has more forest resources than it did at the turn of the century (see table). Apparently, the use of fossil fuels and the increasing dependence on imported food have allowed agricultural lands to go fallow and forests to return through rapid succession. The forest lands of Puerto Rico are now probably net CO2 sinks as are those of the temperate zone of the world where similar phenomena have occurred.
Land-use changes are the critical element in understanding the exchange of carbon between a landscape and the atmosphere. Puerto Rico and Costa Rica are examples of the two extreme points in a long-term oscillation in land use that is a product of how humans use their fossil fuel energy and other natural resources.
Under extreme rates of forest exploitation (Figure 4) such as is happening in certain areas in the wet tropics, large amounts of forest biomass may be converted to CO2 but this condition may paradoxically lead to a sink of carbon because not all the biomass is converted to CO2 and forest succession may be feast enough to support a vigorous young ecosystem with a rapid rate of net carbon uptake.
The fate of wood cut from tropical forests determines how fast that organic matter is returned to the atmosphere as CO2. The current production of roundwood by the world's forest is considerable (about 1.8 thousand million metric tons and increasing) and tropical production accounts for about 40 percent of the total. However, over 70 percent of the world's charcoal and fuelwood production originates in the tropics. In terms of roundwood production within the tropics, it is estimated that 90 percent is allocated for fuelwood and charcoal production-amounting to 0.6 thousand million metric tons and rising. These data suggest that tropical woods are predominantly used as energy sources and thus this biomass is rapidly returned to the atmosphere as CO2 rather than remaining for long-term storage in structures or in other uses. Furthermore, the use of wood energy in the tropics may be more significant than these statistics reflect because of the large amount of wood that is used at the local level and not reported.
During traditional slash-and-burn activities, however, large quantities of wood remain on the ground where the decomposition rate is extremely slow. For example, Ewel et al. (1980) reported that in Costa Rica 40 percent of the original biomass remained on site mostly as soil organic matter and wood.
In environments under natural stress, such as in very dry or cold regions, human populations subsidized by external energy sources could deforest the land irreversibly and thus create conditions that are clearly net sources of CO2.
2. Discharge and sediment export of Rio Tanamá in Puerto Rico for the period of November 1968.
In places like Haiti and the Sahel this must be happening. In these environments, forest succession is not fast enough to make up for human exploitation which is fuelled by a growing population responding to external energy subsidies. Because of land devastation these areas hold little possibility of sustaining human populations in the future when cheap fossil fuels will disappear completely.
It is easier to find mechanisms and reasons why tropical ecosystems are sinks rather than sources of carbon. Paradoxically, even deforestation may create carbon sinks if succession is fast enough and if a portion of the initial biomass remains on site. Climax ecosystems may also be slow sinks of carbon via export of organic matter to downstream aquatic ecosystems. This pathway probably becomes more important in wet environ meets and in disturbed areas where the ecosystem may be more susceptible to leaching. An analysis of the Mauna Loa data that we made also suggests that the biosphere acts as a sink of carbon as does the large carbon accumulations that have occurred in the past.
In experiments with closed microcosms, Odum and Lugo (1970) found that each microcosm reached and maintained a different "atmospheric" CO2 concentration
This "atmospheric" balance depended on the relative proportions of plants and consumers in the microcosm. This leads to speculation about the potential role of the biota in affecting such climatic phenomena as ice ages via the regulation of atmospheric CO2. The current intense activity of humans, powered by abundant fossil fuels, is again tipping the world's carbon balance toward the net production of carbon and possibly toward a warming trend in the world. However, there appears to be enough carbon uptake capacity in the terrestrial biota to slowly counteract this trend. In fact, one could hypothesize that, prior to the fossil fuel era, CO2 in the atmosphere must have been decreasing because our analysis of the Mauna Loa data reflects a net rate of carbon uptake by the biosphere.
If our analysis is correct, it would appear that as long as we manage tropical forests for net yield or as long as these forests are allowed to regenerate naturally, and provided that the area cut is a small fraction of the total forest area, the carbon balance will be maintained. Difficulties, however, will develop at the local level where subsidized human populations can overwhelm the forest ecosystem. Proof of these ideas is hard to come by. The need for the missing information required to reach that state of understanding is behind our own interest in this issue.
BRINSON, M.M. 1973. The organic matter and energy flow of a tropical lowland aquatic ecosystem. Ph.D. thesis. University of Florida, Gainesville. 251 p.
BROWN, S. & A.E. LUGO. 1980. Preliminary estimate of the storage of organic carbon in tropical forest ecosystems. In S. Brown, A.E. Lugo & B. Liegel, eds, The role of tropical forests in the world carbon cycle, p. 65117. Proc. Symposium at the Institute of Tropical Forestry, Rio Piedras, Puerto Rico. US Department of Energy, Washington, D.C. (In press)
EWEL, J., C. BERISH, B. BROWN, N. PRICE & J. RAICH. 1980. Slash and burn impact on a Costa Rican wet forest site. Ecology. (In press)
ODUM, H.T. 1970. Summary: An emerging view of the ecological system at El Verde. In H.T. Odum & R.F. Pigeon, eds, A Tropical rain forest, Ch. I-10. US Atomic Energy Commission. NTIS, Springfield, Va.
ODUM, H.T. & A. LUGO. 1970. Metabolism of forest-floor microcosms. In H.T. Odum & R.F. Pigeon, eds, A tropical rain forest, Ch. I-3. US Atomic Energy Commission. NTIS, Springfield, Va.
RICHEY, J.E., J.T. BROCK, R.J. NAIMAN, R.C. WISSMAR & R.F. STALLARD. 1980. Organic carbon: oxidation and transport in the Amazon River. Science, 207: 1348-1351.
TWILLEY, R. 1980. Organic matter exports from black mangrove forests in south Florida. Ph.D. dissertation. Univ. of Florida, Department of Botany. (In preparation)
WILLIAMS, P.M. 1968. Organic and inorganic constituents of the Amazon River. Nature, 218: 937-938.
ZON, R. & W.N. SPARHAWK. 1923. Forest Resources of the World. McGraw-Hill Book Co., New York. 2 vols. 997 p.
Figure 3. Flow diagram of human use of natural forests when humans have access to fossil fuels. The additional source of energy allows an increased rate of forest exploitation or a substitution of forest products for fossil-fuel fuel derived products. If environmental conditions are favourable, natural succession is rapid after disturbance.
Figure 4. Flow diagram illustrating the use of fossil fuels to rapidly exploit natural forests
