The words Amazonian Rain forest immediately conjure up images of the exotic, of high biodiversity, of the unexplored, and of the poorly understood. While an obvious target for conservation, the very mystery of these systems means that conservation efforts must be based on unusually incomplete data. Forests around major airports and institutions may be relatively well known (Nelson et al., 1990), but with increasing distance there is a commensurate decline in data. Lomolino (2004) proposed the terms Linnean shortfall to describe gaps in our taxonomic knowledge, and Wallacean shortfall regarding our inability to map species’ ranges accurately. Nowhere are these shortfalls more evident than in Amazonia. Indeed, Silman (2007) observed that no full Amazonian distribution of any plant species is completely known, much less its fundamental niche. Understanding the biases in our knowledge, and identifying key gaps and filling them, have to be priorities if effective conservation strategies for Amazonia are to be established.
The familiar threats to Amazonian diversity formed by a rapidly swelling human population, major road and hydroelectric works, the perverse economics of subsidies, and nationalism that demands lands be altered and occupied are even more dangerous now than previously (Laurance et al., 2001). The most recent of threats may be the increased demand for ethanol and biodiesel, and the consequent demand for soy beans and palm oil, respectively. Ironically, the efforts of industrialized nations to wean themselves from oil, partially justified in conservation terms, may have major adverse impacts on Amazonian ecosystems. The probability is that highly diverse caatinga and forest-savanna ecotonal areas of the southern Amazon will be lost to biofuel agriculture before the full effects of anticipated climate change are felt.
The reality that it is impossible to protect all species in all settings has led to many insightful analyses based on economics, endemic diversity, and threat. Myers et al. (2000) proposed prioritization for protection based on a combination of endemic plant diversity and habitat loss. They estimate that, globally, 50% of the rarest plant species occur in just 2% of the Earth’s land area. The hotspot prioritization (now modified to include 34 areas; http://www.conservation.org/xp/CIWEB/regions/priorityareas/hotspots.xml) had a distinct goal of identifying imminently threatened, highest-biodiversity settings. Critics pointed out that large wilderness areas not meeting such explicit hotspot criteria were nevertheless integral to larger-scale conservation (Bates & Demos, 2001). Furthermore, without those larger landscapes and their ecological services the hotspots would fail to meet their conservation objectives (Jepson & Canney, 2001; Whittaker et al., 2005). For example, the hydrological role of vast forest areas such as Amazonia clearly determines the sustainability of the hotspot identified on the eastern flank of the Andes.
As early as the 1980s two further realizations influenced our thinking about conserving the biodiversity of Amazonia. The first of these is that the human occupation of Amazonia is older, and possibly much more influential on past and present ecosystems, than envisioned by the early explorers and most later ecologists (Balée, 1989; Roosevelt et al., 1996). The second realization is the observation that conservation must incorporate climate-change strategies if the goals of maintaining broad biodiversity or even specific populations are to be met (Hannah et al., 2002). Overwhelming data indicate that present and projected rates of climate change, even in the supposedly stable tropics, are without Holocene parallel (Bush et al., 2004; Lovejoy & Hannah, 2005; IPCC, 2007).
The case for human modification of Amazonia is based on archaeological data for distinct locations, a broad scatter of modified soils along major tributary rivers (and possibly elsewhere), and inferences regarding pre-Columbian populations and their distribution (for a colourful review, see Mann, 2005). The claims regarding the extent of these modifications range from most of Amazonia to 10% of the upland area (Balée, 1989). Disease drastically reduced and reshaped the human population of the Americas in the first century after European contact, leading to widespread abandonment and succession. If the claims of broad-scale impacts by pre-Columbian populations are true, and if it is also true that the spectacular biodiversity has survived millennia of hunting, forest alteration and management, the system is possibly much more resilient than hitherto believed (Meggers, 1971). Alternatively, if these claims are overstatements, the potential for misguided policy that expects such ecological resilience as a response to development - particularly of much of the current type - is made more real. Establishing the historical context of these forests and trajectories of change is one topic where biogeographers can serve conservation.
Determining how species will respond to ongoing climate change is surely one of the most important and understudied aspects of tropical ecology. Almost nothing is known of the ecophysiology, resource allocation, or interactions between plants, their competitors, predators, and parasites under conditions of elevated atmospheric concentrations of CO2. Similarly, fine-resolution climate models that can depict temperature, precipitation change, and, as importantly, cloud formation, on the Andean flank are lacking. In the absence of good predictive data, palaeoecological insights into past conditions provide potential proxies for future changes. In Amazonia, much as elsewhere, species responded individualistically, vegetation structure was altered, and migration on millenial scales took place in response to climate change (Colinvaux, 1989; Bush, 2002). The palaeoecological data also demonstrate that the Holocene climate has not been steady and uneventful, but has featured a series of wet and dry events that have caused changes in community composition and distribution (Mayle et al., 2000). These droughts did not cause massive changes in the distribution of savanna and forest systems, but they probably did have many subtle influences on successional pathways, on the probability of fire spreading (whether natural or human-induced), and on sites occupied by humans.
In this special issue we feature papers that deal with some key uncertainties in establishing effective conservation strategies. The themes of determining the timing of past speciation events, the effects of climate change on forest history. The potential for long-distance migration and maintenance of the evolutionary potential of forests in the future are addressed from a variety of viewpoints in these papers (Quijada-Mascareñas et al., 2007; Rull & Nogué, 2007; de Toledo & Bush, 2007). The western Amazon receives special consideration in two papers, one dealing with geomorphic influences on modern climatic patterns (Killeen et al., 1990) and the other a case study on the extent of pre-Columbian human activity in a southern Peruvian landscape (Bush et al., 2007). The scale and mapping of Amazonian wetland habitats is addressed in a paper by Toivonen et al. (2007), and two papers discuss the shortcomings of our knowledge of Amazonian plant distributions and the implications for protecting endemics or sites of highest diversity (Hopkins, 2007; Schulman et al., 2007).
The overwhelming message from the papers as a group is that Amazonia is heterogeneous in terms of landscape, climate, and history. These three factors contribute significantly to the poorly understood or documented patterns of regional biodiversity, and underline that conservation efforts must perforce proceed with imperfect knowledge and that they will need to be refined as fresh insights are acquired. One important insight that emerges from the palaeoecological studies is that Amazonian forests have withstood substantial climate change in the past. These forests are resilient systems, but what they have not withstood is fire.
The projected climatic changes for the next century (IPCC, 2007) are faster and more profound than any experienced in the last 40,000 years (Bush et al., 2004), and probably in the last 100,000 years. The synergism induced in Amazonia through simplifying ecological structure from the complexity of forest to the simplicity of soybean fields or ranchland, the increased probability of human-set fires, and complex ecological interactions mediated by climate, e.g. bacterial diseases (Pounds et al., 2006), takes us into unknown bioclimatic territory. In addition, we now know that the hydrological cycle, which produces a significant fraction of Amazon rainfall, is also responsible for 40% of the rain south of the forest in Brazil and northern Argentina. With 20% of the Brazilian Amazon deforested (and considerable further areas modified in various ways), the point at which an irreversible drying trend is triggered cannot be far away. The immediate consequence of that alteration to the modern hydrological cycle, forest fragmentation, and human activity is the increased probability of Amazonian wildfire. Policies to limit anthropogenic forcing of the climate are absolutely necessary, and inextricably linked to measures that control wildfire. What is needed is an Amazon-wide management plan – consisting of a mosaic of protected areas and forest used in other ways – so that the hydrological cycle is robust in the face of stresses from El Niño and the Atlantic circulation patterns that triggered the great drought of 2005.