The Atmospheric Component of Biogeochemical Cycles in the Amazon Basin

Tropical forests, with their high biological activity, have the potential to emit large amounts of trace gases and aerosol particles to the atmosphere. The accelerated development and land clearing that is occurring in large areas of the Amazon basin suggest that anthropogenic effects on natural biogeochemical cycles are already occurring (Gash et al. 1996). The atmosphere plays a key role in this process. The tropics are the part of the globe with the most rapidly growing population, the most dramatic industrial expansion and the most rapid and pervasive change in land use and land cover. Also the tropics contain the largest standing stocks of terrestrial vegetation and have the highest rates of photosynthesis and respiration. It is likely that changes in tropical land use will have a profound impact on the global atmosphere (Andreae 1998, Andreae and Crutzen 1997). A significant fraction of nutrients are transported or dislocated through the atmosphere in the form of trace gases, aerosol particles, and rainwater (Keller et al. 1991). Also the global effects of carbon dioxide, methane, nitrous oxide, and other trace gases have in the forest ecosystems a key partner. The large emissions of isoprene, terpenes, and many other volatile organic compounds could impact carbon cycling and the production of secondary aerosol particles over the Amazon region. Vegetation is a natural source of many types of aerosol particles that play an important role in the radiation budget over large areas (Artaxo et al. 1998). There are 5 major reservoirs in the Earth system: atmosphere, biosphere (vegetation, animals), soils, hydrosphere (oceans, lakes, rivers, groundwater), and the lithosphere (Earth crust). Elemental cycles of carbon, oxygen, nitrogen, sulfur, phosphorus, and other elements interact with the different reservoirs of the Earth system. The carbon cycle has important aspects in tropical forests due to the large amount of carbon stored in the tropical forests and the high rate of tropical deforestation (Jacob 1999). In Amazonia there are two very different atmospheric conditions: the wet season (mostly from November to June) and the dry season (July-October) (see Marengo and Nobre, this volume). Biomass burning emissions dominate completely the atmospheric concentrations over large areas of the Amazon basin during the dry season (Artaxo et al. 1988).

General Characteristics and Variability of Climate in the Amazon Basin and its Links to the Global Climate System

The Amazon region is of particular interest because it represents a large source of heat in the tropics and has been shown to have a significant impact on extratropical circulation, and it is Earth’s largest and most intense land-based convective center. During the Southern Hemisphere summer when convection is best developed, the Amazon basin is one of the wettest regions on Earth. Amazonia is of course not isolated from the rest of the world, and a global perspective is needed to understand the nature and causes of climatological anomalies in Amazonia and how they feed back to influence the global climate system. The Amazon River system is the single, largest source of freshwater on Earth. The flow regime of this river system is relatively unimpacted by humans (Vörösmarty et al. 1997 a, b) and is subject to interannual variability in tropical precipitation that ultimately is translated into large variations in downstream hydrographs (Marengo et al. 1998a, Vörösmarty et al. 1996, Richey et al. 1989a, b). The recycling of local evaporation and precipitation by the forest accounts for a sizable portion of the regional water budget (Nobre et al. 1991, Eltahir 1996), and as large areas of the basin are subject to active deforestation there is grave concern about how such land surface disruptions may affect the water cycle in the tropics (see reviews in Lean et al. 1996). Previous studies have emphasized either how large-scale atmospheric circulation or land surface conditions can directly control the seasonal changes in rainfall producing mechanisms. Studies invoking controls of convection and rainfall by large-scale circulation emphasize the relationship between the establishment of upper-tropospheric circulation over Bolivia and moisture transport from the Atlantic ocean for initiation of the wet season and its intensity (see reviews in Marengo et al. 1999). On the other hand, Eltahir and Pal (1996) have shown that Amazon convection is closely related to land surface humidity and temperature, while Fu et al. (1999) indicate that the wet season in the Amazon basin is controlled by both changes in land surface temperature and the sea surface temperature (SST) in the adjacent oceans, depending if the region is north-equatorial or southern Amazonia.

Trace Elements in the Mainstem Amazon River

Measurements of trace metals in rivers are of substantial interest for researchers examining basic scientific questions related to geochemical weathering and transport and to scientists involved in pollution control evaluation. Trace metals in natural waters include essential elements such as cobalt, copper, zinc, manganese, iron, molybdenum, nickel, which may also be toxic at higher concentrations, and nonessential elements, which are toxic, such as cadmium, mercury and lead. Recent findings indicate that iron and, to a lesser extent, zinc and manganese play an important role in regulating the growth and ecology of phytoplankton (Martin et al. 1991), while in contrast, cadmium, arsenic, and mercury have long been recognized as poisonous to living organisms (see Pfeiffer et al. 1993, for a description of mercury problem in the Amazon basin). The release of potentially large quantities of these toxic metals, particularly in the river systems of industrialized countries, but also in tropical rivers, is an acute problem of great environmental concern. An understanding of the weathering and transport processes controlling the fate and flux of trace metals in pristine environments is important in evaluating the capacity of receiving waters to accommodate wastes without detrimental effects. The Amazon River system, which is relatively free of industrial and agricultural interference, represents an ideal case for the investigation of the origin and transport of trace metals. This understanding may also provide a scientific basis for the anticipated development of the Amazon basin. With regard to trace metals, Amazon River is still poorly documented. Martin and Meybeck (1979) and Martin and Gordeev (1986) presented a global tabulation of trace metal concentrations in particulate matter of major rivers including the Amazon, and Palmer and Edmond (1992) measured dissolved Fe, Al, and Sr concentrations in the Amazon mainstream and a number of its tributaries. Boyle et al. (1982) and Gordeev et al. (1990) published some data on Cu, Ni, Cd, and Ag dissolved concentrations at the mouth of the Amazon River and in its oceanic plume. Konhauser et al. (1994) reported the trace and rare earth elemental composition of sediments, soils and waters, mainly in the region of Manaus.

Soil versus Biological Controls on Nutrient Cycling in Terra Firme Forests

Terra firme forests are those that by definition are not permanently or seasonally flooded (terra firme meaning “firm terrain”). This type of forest encompasses the Amazon and Orinoco basins, stretching from the lower slopes of the Andes, east to the Guianas, and south to about 15°S in western Brazil and northern Bolivia (Richards 1996). Structural and compositional variability in these forests in the Amazon basin is very wide as a result of climate differences and geomorphological position. The region is not climatically uniform; the central and much of the southern parts have less and more seasonal rainfall than the eastern and western parts (Walsh 1996). These differences have direct and indirect ecological significance, as phenology and biological processes related to nutrient availability will be strongly influenced by both factors (Cuevas and Medina 1986, 1988, 1990, Medina and Cuevas 1989). Periods of two or more consecutive dry days are ecologically significant in a humid area such as San Carlos de Río Negro, in the northern part of the Amazon, because of low water retention capacity in the widespread sandy soils. In lower geomorphological positions, dry spells of 5-10 days may result in fluctuations of the water table from 0.4-1.0 m (Herrera 1977, Bongers et al. 1985). In areas with a more strongly seasonal climate, roots have been found extending to 18 m depth (Nepstad et al. 1995). This may explain the presence of evergreen forest in the seasonally dry eastern Amazon. Structure and physiognomy of terra firme forests is very similar throughout Amazonia, but floristically it is quite variable due to different compositions in the subbasins of the Amazon’s major tributaries. These subbasins are located within geochemical regions that can be differentiated based on the physicochemical properties of drainage waters (Sioli 1975, Fittkau 1971, Fittkau et al. 1975). Blackwater rivers, such as the Río Negro, drain mostly sandy podsolized soils low in most essential nutrients for plant growth. They are characterized by a high content of humic acids, which remain dissolved because of the predominant low concentrations of polyvalent cations, mainly Ca2+ and Mg2+.

Carbon Storage in Biomass and Soils

Carbon dioxide and methane integrate biogeochemical cycles of C and constitute, together with nitrous oxide, the main trace gases responsible for the greenhouse effect. Increasing interest in the global consequences of climate change has prompted the global scientific community to deepen their studies about the global C stocks and the interrelations among its different compartments. As main compartments, soils and phytomass (living and nonliving) have received special attention. Many authors proposed a quantification of C stored in soils and proposed to study their role as both a source and sink of carbon (Post et al. 1982, Eswaran et al. 1993, Sombroek et al. 1993, Batjes 1996). The world’s mineral soils are estimated to contain about 1500 Pg C (Post et al. 1982, Eswaran et al. 1993, Batjes 1996), while the biomass of plants is estimated to be comprised between 560 and 835 Pg C (Whittaker and Likens 1975, Bouwman 1990). Tropical forests account for between 20 and 25% of the world terrestrial C (Brown and Lugo 1982, Dixon et al. 1994). The Amazon contains the largest expanse of native tropical ecosystems and has a direct influence on global biogeochemical cycles, especially the C cycle. The C stored in phytomass is of importance because of its quantity and its potential to be released easily. Carbon in soil is proved to be important because soil organic carbon (SOC) is intimately involved in virtually all biological processes, and organic matter (OM), even when present in small amounts, is an extremely important soil constituent. Two Brazilian soil classes, Latossolos and Podzólicos, make up 73% of the total area of the Legal Amazon Basin of Brazil (Prado 1996, Jacomine and Camargo 1996). More precisely, only three dystrofic soil types, Podzólico Vermelho Amarelo (Acrisol), Latossolo Amarelo (xanthic Ferralsol), and Latossolo Vermelho Amarelo (orthic Ferralsol) cover approximately 60% of the total, and are therefore of prime interest. The remainder is distributed between 13 additional classes. Only 6, however, represent more than one percent, and only 2 of which are more than 5%: Plintossolos (Inceptisols, Oxisols, and Alfisols) and Gleissolos (Entisols and Inceptisols).

The Interface Between Economics and Nutrient Cycling in Amazon Land Development

Most of the terra firme soils in the Amazon are highly weathered, highly leached, have low capacity for retaining nutrients against the continual leaching and weathering of the tropical climate, and are classified as Oxisols and Ultisols, soil types with extremely low fertility (see Cuevas, this volume). The naturally occurring forests of the region maintain a high production of wood and leaves through very efficient recycling of nutrients from decomposing litter to roots in a root-humus layer on top of the mineral soil or near its surface. The decomposing litter is important not only as a source of nutrients, but as a source of organic acids which prevent phosphorus fixation in the iron- and aluminium-rich soils of the Amazon. When forests on Amazonian terra firme soils are cut and burned, and the soils used for agriculture, litter, and humus are rapidly oxidized and destroyed. As a result, the potassium remaining from the original forest is quickly leached, the nitrogen is volatilized, and the phosphorus is immobilized in the mineral soil. This is one of the most important reasons that crop production can be carried out for only a few years under shifting cultivation. It is not just small scale agriculture that is limited by the low fertility of Amazonian soils. In the past, almost all types of development that destroy the nutrient conserving mechanisms of the forest have suffered financially. Two examples are given here to illustrate. In 1967, one of the largest conversions of tropical forest to pulp plantation began near the junction of the Jarí and Amazon rivers, in the state of Pará, Brazil (Time, 1976). The “Jarí” project was initiated and financed by Daniel K. Ludwig, one of the world’s richest men, and owner of numerous international corporations. Ludwig had anticipated a global shortage of wood fiber for pulp, and to meet this shortage, he and his advisors selected a site that they believed had high potential for pulp production (Time 1979, Kinkead 1981). By 1981, the total investment in the 12,000 km2 tract of land was approximately $1 billion (Kinkead 1981). Ludwig’s advisors recommended melina (Gmelina arborea) as the best species to plant.

The Recovery of Biomass, Nutrient Stocks, and Deep-Soil Functions in Secondary Forests

Secondary forests cover approximately one third of the 0.5 million km2 of the Brazilian Amazon that have been cleared for agriculture (Houghton et al. 2000, Fearnside and Guimarães 1996). These forests counteract many of the deleterious impacts of forest conversion to agriculture and cattle pasture. They absorb carbon from the atmosphere, they reestablish hydrological functions performed by mature forests, and they reduce the flammability of agricultural landscapes. Secondary forests transfer nutrients from the soil to living biomass, thereby reducing the potential losses of nutrients from the land through leaching and erosion. They also allow the expansion of native plant and animal populations from mature forest remnants back into agricultural landscapes. The study of forest recovery has focused on aboveground processes, primarily biomass accumulation. The few studies that have examined the recovery of belowground functions in Amazon secondary forests have been restricted to the upper meter or less of soil (e.g. Buschbacher et al. 1988). A review of our knowledge of secondary forest recovery is needed that incorporates accumulating evidence that approximately half of the region’s forests rely upon root systems extending to depths of several meters to maintain evapotranspiration during prolonged seasonal drought (Nepstad et al. 1994, Jipp et al. 1998, Nepstad et al. 1999a, Hodnett et al. 1997; see also Richter and Markewitz 1995). This discovery demands a conceptual shift in our approach to forest recovery on abandoned land. Are secondary forests capable of regrowing deep root systems, thereby recovering hydrologic functions and fire resistance of the mature forest? At what rate does this recovery take place? How does this ability to tap a large soil volume change our thinking about the role that nutrient shortages play in restricting secondary forest recovery? In this chapter, we begin to address these questions with the goal of furthering a mechanistic understanding of forest recovery on abandoned Amazonian lands. Our analysis focuses on three measures of secondary forest development: biomass accumulation, nutrient accumulation, and hydrological recovery. We choose biomass accumulation, because it is the best integrative measure of secondary forest development, it is the basis for estimates of carbon sequestration by secondary forests, and it is the most frequently measured secondary forest parameter.

The Relevance of Biogeochemistry to Amazon Development and Conservation

To read the press of recent years, one might imagine that the fate of the world rests in the hands of those who would develop the Amazon basin. Waves of incoming colonists are blamed for the bulk of the deforestation and development (Schomberg 1998), but Asian logging firms, multinational oil companies, and gold miners are also portrayed as destructive agents hacking down the forest, systematically undermining its biodiversity, and severely contaminating its myriad ecosystems (Althaus 1996, Ferreira 1996, James 1998). The effects of these varied threats are regularly broadcast in alarming tones. Rueters News Service warned in January 1998 that “Brazil’s Amazon rain forest, the world’s richest trove of biological diversity and source of much of the Earth’s oxygen, continues to be ravaged” (Craig 1998). And, in April 1999, a writer for the Associated Press communicated the “fear” of unspecified scientists that “damage to the rain forest... could throw the Earth’s climate out of balance” (Donn 1999). Clearly, the fate of the Amazon and the implications of its fate to the overall Earth system are topics of enormous scientific and popular interest. While there is little disagreement that the complete destruction of Amazon forests would be catastrophic, what about partial deforestation of the region? How much, and which parts, of the Amazon can be converted to sustainable human land uses without compromising the ecological integrity of the conserved areas? How might this development impact regional climate, adjoining coastal systems, and overall global processes? Answers to these volatile questions remain elusive and seemingly endless strands of controversy swirl about them. At the heart of the matter, yet largely beyond the public discussion, are biogeochemical cycles that support and regulate the functioning of the Amazonia’s biological systems. Moreover, it is the incomplete understanding of these cycles that promotes uncertainty and feeds the controversy. The purpose of this book is to present a coherent assessment of our current understanding of the biogeochemical functioning of the Amazon basin. Although it is surely presumptuous to assume that this presentation will shed sufficient light on the uncertainties to eliminate the current controversies, we hope that it will provide a basis for lifting the discussion to a higher level.

Biogeochemistry of Amazon Floodplain Lakes and Associated Wetlands

Floodplains and associated lakes are important components of the biogeochemistry, ecology, and hydrology of the Amazon basin. Amazon floodplains contain thousands of lakes and associated wetlands linked to each other and to the many rivers of the immense basin. These floodplain lakes modify the passage of flood waves (Richey et al. 1989a), increase nutrient retention and recycling (Melack and Fisher 1990), and influence the chemistry of the rivers (Devol et al. 1995). The mosaic of flooded forests, open water, and floating macrophytes in the central Amazon floodplain makes a significant contribution of methane to the troposphere (Bartlett et al. 1988, Devol et al. 1990). The fishery potential of the large river systems is closely tied to the area of floodplain and the magnitude and duration of inundation (Welcomme 1979, Bayley and Petrere 1989). The majority of fishes harvested in the Amazon basin obtain nutrition in flooded forests (Goulding 1980) or from organic matter derived from floodplain algae (Araujo-Lima et al. 1986, Forsberg et al. 1993). Much progress has been made during the last fifty years toward understanding the lakes of the Amazon floodplain. Still, the vast size of the Amazon basin poses challenges to limnologists working in the region. Recent research has been enhanced by the maintenance of functional floating laboratories in several areas, use of modern ships capable of regional surveys and equipped for hydrographic studies, and applications of remote sensing. Our objective in this chapter is to examine the role of lakes in the hydrology of the floodplain and in the biogeochemistry of carbon, nitrogen, and phosphorous within the central Amazon basin. Particular emphasis is placed on how inundation patterns interplay with carbon balance and nutrient limitation. By combining numerous measurements of primary productivity with recent results from studies using isotopes of carbon, we will examine the contribution of the major plant groups to aquatic foodwebs, and offer a new paradigm for the processing of organic carbon on the Amazon floodplain. The interplay between the Amazon River and local catchments as sources of nutrients to the floodplain indicates the potential sensitivity of the lakes to basin-wide and local disturbances.

Geoecological Controls on Elemental Fluxes in Communities of Higher Plants in Amazonian floodplains

Navigators visiting the Amazon during the fifteenth century provided the earliest descriptions of aquatic systems in the region, but it was not until mid twentieth century that systematic studies of the limnology of Amazon waters began (Sioli 1984). The inclusion of vegetation as an important part of the aquatic biota was only possible after the relatively recent change from traditional potamic limnology to wetlands limnology (Sioli 1975). The first studies of the vegetation of Amazon wetlands consisted mainly of species descriptions, and it is only recently that studies of floodplain vegetation have attained a level of importance equivalent to that of studies dealing with water chemistry, phytoplankton, zooplankton, and fishes. For the past thirty years, fertile floodplain systems along Whitewater rivers (várzea) have been focal areas of colonization. This fertility also supports high rates of primary production within the higher plants, especially of the herbaceous vegetation (Piedade et al. 1991, Junk and Piedade 1993a, 1997). Quickly turning-over pools of nutrients (Junk and Furch 1991, Furch and Junk 1992, 1997a) and direct connections with contiguous terra firme forest and river channels (Alves 1993, Furch and Junk 1997b) are also characteristic of these floodplain systems. As a consequence of the annual floodpulse (Junk et al. 1989), floodplain vegetation is subjected to aquatic and terrestrial phases, which hold important ecological implications for both the plant populations and related aquatic and terrestrial biota. Life cycles of the species and the time available for growing depend upon the duration of inundation and drought periods and the habit of the species. During the year, pulses of growth and dormancy occur and herbaceous vegetation changes its species composition according to the phase of the hydrological cycle. In this chapter we discuss the distribution and the development of plant communities in floodplain areas, mainly of the big Whitewater rivers, focusing on factors such as diversity, species composition, biomass and primary production. Based upon these factors, we also discuss the annual dynamics of bioelements stocks and their turnover through herbaceous and floodplain forest communities. Finally, we examine the implications of such nutrient dynamics and turnover for the aquatic biota.