Long-term forest dynamics in Peruvian Amazonia

Tropical forests are changing rapidly. Studies based on monitoring forest plots across the tropics show that, over the past 30 years, forests have accumulated additional biomass, possibly as a result of fertilization by carbon dioxide. Other changes have included an increasing abundance of certain taxa such as lianas (ref. [25] in the Case for Support) and faster-growing canopy trees [26]. These changes are important globally because (a) tropical forests play a major role in the global carbon cycle [27]; (b) increased growth rates may mean that they are acting as a short-term carbon sink (e.g. [8]); (c) some studies (e.g. [27]) suggest that tropical forests may themselves be vulnerable to climate change, and could either stop sequestering new carbon dioxide or actually start releasing it if mortality rates increase, which would have consequences for the climate system and economies on regional and global scales. One possible explanation for the observed change is that forests are not in equilibrium but are still responding to disturbances which took place at some time in the past. This could include natural disturbances, such as droughts [33] or storms [34], or human disturbances caused by, for example, clearing forest to grow crops and actively fertilizing the soils [35]. This latter idea has gained ground in recent years [37] on the basis of recent archaeological discoveries in South America [36] which suggest that, on the eve of the European colonization, indigenous populations were much larger than previously thought, and thus may have had a significant impact on the forests. One way to test whether or not forests have been disturbed in the distant past is to use the palaeoecological record. Peat sequences trap pollen grains from vegetation and preserve them in stratigraphic layers; analysing this pollen allows us to reconstruct past vegetation communities and the way they have changed over time. If the trends recorded over the last 30-40 years are only the most recent part of a trend which has been continuing for much longer, triggered by a past disturbance, then pollen records should show pronounced changes over time (particularly in the last 100-1000 years). If the forests were in fact in equilibrium before monitoring began, then the pollen assemblages should show no significant change during this interval. Although pollen analysis has previously been undertaken at a number of sites in Amazonia, there are two problems with the existing records. Firstly, there are few records with sufficient temporal resolution in the last thousand years to detect fluctuations in forest cover (in most cases, there are only one or two samples in the last thousand years). Secondly, sites with palaeoecological potential situated close to established monitoring plots are extremely rare. The recent discovery by Lähteenoja et al. [20] of rapidly accumulating peats in an area of forest monitoring in northern Peru offers a unique opportunity to resolve these problems. This project seeks to establish (1) the current composition and ecology of the vegetation at these sites today; (2) the potential of these peat sequences to provide good pollen records; (3) the history of the forest communities; (4) the relationship between the modern pollen rain and the surrounding vegetation to assist in interpreting the pollen. This study will form the foundation for future work in this region where there may be considerable potential for integrated neo- and palaeoecology to address questions of long-term ecology.