Soil characterization and reconnaissance survey of the Ranomafana National Park area, southeastern Madagascar.

Abstract

The Madagascar rainforests display some of the richest biodiversity found anywhere on earth. This unique biodiversity is imperiled by deforestation resulting mainly from the accumulated impacts of small-scale slash-and-burn agriculture (tavy), practiced by subsistence farmers. The Ranomafana National Park (RNP), created to conserve an especially species-rich rainforest region, essentially depends upon developing sustainable agriculture and forestry alternatives to tavy. The main objectives of this study were to document soil systems of the region, provide detailed soil characterization data, and to provide soil interpretations for sustainable agriculture and forestry development.

The regional geology consists mainly of high-grade felsic metamorphic rocks (gneiss, migmatite) and granites. Saprolite on higher, stable landscape positions commonly reached 10-20+ meters thick, while saprolite on lower geomorphic surfaces was generally shallow to rock due to stream downcutting and geologic erosion. Alluvium was generally shallow (<2-3 meters) to gleyed saprolite in lower-order stream valleys, but deeper in major river valleys.

Long-term climate data indicated a udic soil moisture regime throughout the region. Soil temperature measurements indicated an isothermic soil temperature regime, and were highly correlated with elevation (r2 = 0.97) and estimates of mean annual air temperature. The soil temperature data suggested that mean annual air temperature equals mean annual soil temperature, with no correction factor required.

Upland soils over thick saprolite profiles were generally classified as Oxic Dystrudepts. Some soils contained horizons that met all criteria for an oxic horizon, except that Soil Taxonomy considers sand-sized kaolinite pseudomorphs as weatherable minerals. Upland soils found in lower landscape positions, or places with thin saprolite profiles, tended to have higher cation exchange capacities and relatively high biotite and/or feldspar contents, and classified mainly as Typic Dystrudepts. Argillic horizons were rarely observed and were weakly expressed, and no mappable areas of soils with argillic horizons were found. Umbric epipedons occurred as small patches on the upland landscape, but represented only inclusions within potential soil map units.

Alluvial soils classified into various subgroups of Humaquepts, Dystrudepts, Fluvaquents, Udifluvents, and Udipsamments. Ochric, umbric, and histic epipedons were all described for these lowland soils.

Total elemental analyses of representative rock and saprolite samples revealed extremely low levels of bases and phosphorus. Soil characterization data generally indicated pH values of 4-5, high to very high aluminum saturation levels, extremely low exchangeable base levels, and extremely low Olsen phosphorus values, regardless of landscape position. One area contained a geologic anomaly with extremely high phosphorus contents, but was limited in extent. With minor exceptions, the RNP area soils reflected not only very low available nutrients, but also very low total nutrient reserves.

The combined geologic, total elemental, and soil characterization data revealed that soil nutrient levels in the RNP area are insufficient for sustained crop production without nutrient inputs. Those inputs must be obtained either from the rainforest biomass through burning, through nutrient harvesting of live biomass for compost, or through chemical fertilizers. Rainforest burning is incompatible with the RNP conservation objectives. Nutrient harvesting can concentrate nutrients in agricultural fields, but at the expense of other portions of the ecosystem. Inputs in the form of chemical fertilizers therefore represent the only long-term solution for sustainable agriculture if present population densities are maintained. To suggest that "alternative" agronomic practices (e.g., SRI, or system of rice intensification) can overcome these fundamental biological and chemical constraints only provides a false hope that will be ultimately proven unsustainable in the long-term. That will eventually spell calamity for both the Malagasy rainforest and the Malagasy farmers.