When European naturalists first visited the New World Tropics they saw vast forests that seemed untouched by humans. While indigenous people often lived in these forests, their populations were small. This led to a perception of tropical forests as primeval, “virgin” forests. In the last few decades, this perception has changed - large areas now covered by mature forests have a history of cultivation. In many cases, “primeval” forests are less than 500 years old.
La Selva biological station in Costa Rica is one of the premier research stations for Neotropical biology. Prior to archaeological study of the site, much of it was assumed to be free of human influence. However, the discovery of pre-Columbian artefacts led to the discovery that the site had been occupied at least 3000 years ago. Charcoal was more abundant in alluvial terraces (flatter areas with deeper, more fertile soil) and less abundant in the less fertile upland soils. A chronology, established by Sol (2000)1, divided the La Selva into four archaeological phases: La Cabaña 1000 – 1550 CE; La Selva 500 – 1000 CE; El Bosque 300 BCE– 500 CE; La Montaña 1500 –300 BCE.
To better understand the history of the site, Lisa Kennedy of Virginia Tech and Sally Horn of the University of Tennessee, Knoxville, undertook a study of sediment cores extracted from the Cantarra swamp2, a 0.5 ha wetland dominated by perennial herbs. They used pollen, charcoal and macrofossils to reconstruct the environmental history of the site. Wetlands are frequently used to reconstruct vegetation histories. As sediments accumulate in bodies of water, plant pollen, fern spores and charcoal fragments are trapped. Pollen coats are extremely tough, and decay takes place very slowly in waterlogged soils. If the vegetation surrounding the site changes, different types of pollen will be deposited into the site. Someone with the patience to sort through these cores can observe thousands of years of history in a few metres of sediment.
The most obvious evidence of human activity is the presence of corn (Zea mays subsp. mays) pollen. Corn is a cultivated species which does very poorly without human intervention. Thus, the presence of corn pollen in the wetland sediments is direct evidence of agriculture. Corn pollen shows up from 880 CE to somewhere between the mid-1600s and mid-1800s. Pollen of other species like Amaranths, Asteraceae (the sunflower family), and other grasses and herbs also peak during and before the “corn zone”, often at the same time that charcoal density peaks. This may also reflect cultivation, although it could represent weedy species establishing after fires. Corn pollen was found in sediments about 1300 years older at another lake about 2 kn distance from this one. The authors suggested that disturbance in this time period at Cantarra swamp may have represented the cultivation of root crops (which don’t leave the kind of pollen signature that corn does.
As a forest ecologist, I find some of the “other evidence of disturbance” to be the most interesting. There are several peaks of Cecropia pollen, and to a lesser extent Trema pollen. These are fast-growing species that are usually associated with large gaps in the forest - specifically the type that human agricultural activities may have suggested. Other peaks of pollen belonging to forest species suggests that periods of forest recovery were interspersed with the cultivated times.
This is very interesting stuff. We are too inclined to interpret forests as “primeval”. In many cases, what our eyes see as ancient is only a few centuries old. It is important to understand that if we want to construct realistic models of forest dynamics.
- Sol, C., R. F. 2000. Asentamientos prehispánicos en la Reserva Biológica La Selva, Sarapiquí, Costa Rica: Sistemas de explotación de recursos naturales en un bosque tropical lluvioso. Licenciatura thesis, School of Anthropology and Sociology, University of Costa Rica.
- Lisa M. Kennedy, Sally P. Horn. A Late Holocene Pollen and Charcoal Record from La Selva Biological Station, Costa Rica . Biotropica 40(1):11-19.