In forested parts of the world, it is relatively easy to see human impacts on terrestrial vegetation and to discern them from natural impacts. In non-forested areas, such as arid and semi-arid regions, it is much more difficult to distinguish human from natural impacts.
In attempting to evaluate human impacts on vegetation, the area of study is quite large - the entire terrestrial surface of the globe. Therefore, satellite data are needed to obtain the necessary coverage of space and time. In addition, smaller scale process studies that can be extrapolated to larger areas are needed, as well as output from models from which predictive capacity can be gained. In pursuing these various methodologies, there is the problem of differing scales. Coarser scales involve more abstraction as well as the possibility of more errors. Some studies need to go all the way down to the 1 meter by 1 meter scale, if not to the individual leaf, while others go all the way up to hundreds of square kilometers. In situ measurements, satellite data and models, are all needed to gain an understanding of change in global vegetative cover.
The issue of deforestation has received spurts of interest from the popular press. There is a great deal of confusion and widely varying claims regarding the actual rates of deforestation in Amazonia and elsewhere. Consequently, new research is needed to provide improvement over the present inadequacy of knowledge. This subject can be approached incrementally and much can be learned from the process.
Figure 19.1 shows global sources and sinks of carbon and demonstrates the uncertainties that exist regarding the global carbon cycle. In some variables, there is a high degree of certainty, including the fossil fuel contribution and oceanic uptake. Other areas are not as well understood and need further refining, including the mid-latitude terrestrial carbon sink and the impacts of land use change.
There are naturally occurring variations in the climate system, such as the El Niño-Southern Oscillation (ENSO), which are closely correlated with regional climatic events. A good example is the direct correlation between an El Niño-induced rise of sea surface temperature (SST) in the Pacific and maize yield in Zimbabwe. Similarly, a satellite-derived vegetation index for Brazil is highly correlated with tropical Pacific SST. The alteration in the rainfall regime causes the vegetation index to rise with rising SST. These are examples of how natural phenomena, which may or may not be affected by humans, can affect rainfall and climate, and hence vegetation patterns on a regional scale. These stand in contrast to forest removal, where the role of humans is unequivocal.
The 1991 eruption of Mount Pinatubo injected a huge cloud of sulfate aerosols into the atmosphere, creating a significant climatic forcing, cooling the planet as a whole and possibly affecting vegetative response and patterns. This an example of a natural phenomenon which caused a global-scale perturbation of the system. There is so much interaction among the variables that much confusion exists with regard to causation.
African Sahel
In the Sub-Saharan African Sahel, a region dominated by seasonal semi-arid grassland, rainfall records show that from 1949 to 1970, conditions were wetter than usual. This was followed by a period of approximately 15 dry years, a drought which led to allegations of anthropogenic regional climate forcing. It was suggested that the principal mechanism for initiating this feedback was overgrazing and removal of trees for firewood. Because these areas are arid and marginal to begin with, they are most at risk for these kinds of changes.
Charney et al. of MIT hypothesized that in creasing human land use in the arid zone would remove litter and vegetation, thereby drying the land and increasing its albedo. This would reflect more sunlight away, causing the region to become cooler, and lessening convective activity. This would decrease rainfall, further drying the land and continuing a climate feedback that would make the region drier. After examining this region with satellite data, Tucker believes it is difficult to sort out the human impact from the natural variability.
Figure 19.2 shows the importance of taking a longer timescale look at data before drawing conclusions. If the 1980 to 1984 numbers were all the data available, they would suggest that the system had crashed. But the longer time scale reveals that there has been a recovery of the vegetation. It is important to avoid simplistic interpretations and recognize that there is large interannual variation. A baseline is needed to discern human-caused versus natural variation. Is this global change or is it a basically stable system with a high degree of variation?
Pacific Northwest Forests
Because of gross changes in surface conditions between land which is forested and that which is not, it is relatively easy to discern human impact in these regions. A checker board pattern shows up in remotely sensed images of forests that have been cut corresponding to a pattern of private holdings within public lands. Landsat Thematic Mapper mosaics of late succession coniferous forests of the U. S. Pacific Northwest show that in that region, the remaining old growth forest is almost exclusively at higher elevations and on Federal land. The scarcity of old growth forest on private land is largely due to the high economic value of old growth trees - a 200 to 300-year-old Douglas Fir can be worth $5000 to $7000 just for the felled tree.
The darker areas in the image are old growth forests which have more variation in tree height and hence more shade. Cut forests have less shade because new-growth trees are mostly the same height. The forest of the U. S. Pacific Northwest is tremendously fragmented due to the high degree of timbering on private land and the Forest Service policy of distributing timber cuts. This high degree of habitat fragmentation causes great problems for the Northern Spotted Owl and other species that require large contiguous blocks of undisturbed old growth forest.
Brazilian Tropical Rainforest
Tucker and colleague David Skole undertook a study of satellite data, coupled with several site visits, in an attempt to get definitive answers and repeatable results regarding deforestation in Brazil over the last 20 years.
There are large areas with little human activity. The impact of native indigenous people is usually quite small, with a few exceptions. There are large tracts of undisturbed forest as well as areas that are severely fragmented. The transportation network is the principal conduit for colonization of the forest, which leads to deforestation. Where there are no roads, there is little deforestation. Prior to road building, lake areas were the main ones colonized, so new growth forests are often found there.
The region is not all Amazonian high forest. There is also some cerrado (savanna) which has been converted to agriculture, since it is easier to use cerrado for agriculture and the soils can be better than tropical forest. Thus, there tends to be less impact on tropical forest in these areas.
Skole and Tucker estimated deforestation over of 5 million km2 in the Brazilian Amazon. They found that in 1978, approximately 80,000 km2 were deforested, and that by 1988, this had increased to about 230,000 km 2 In addition to outright deforestation, they also examined areas of isolated forest (those cut off from areas of contiguous forest), and the totality of "edge effect." In the edges of forests, hunting, predation by feral animals, drying due to microclimate conditions, and other biological and physical effects, greatly impact the biodiversity in a one kilometer area around the edge.
The results of these investigations revealed a much higher figure for habitat fragmentation than for deforestation, which is important with regard to biodiversity. When isolated forest and edge effects were included, they found that 600,000 km2 of tropical forest were affected with regard to impacts on biodiversity. The use of satellite data and selected ground visits help to break out results like this in an effort to reduce uncertainty in this area.
In Brazil, from 1990 on, the government instituted close monitoring and policing of fires. They actively enforced existing laws and put an end to subsidies that encouraged rainforest destruction. Through these policies, Brazil's deforestation was reduced to about 10,000 to 12,000 km2 of new deforestation per year, a rate of clearing of about 1/4% per year, equal to a 400-year rotation. By comparison, on private land in the U.S., there is a 50- to 75-year rotation; and on Federal land about a 100-year rotation or a 1% clearing rate per year.