Tree influence on soil and pasture: contribution of proximal sensing to pasture productivity and quality estimation in montado ecosystems

摘要

Montado is a silvo-pastoral ecosystem of the Mediterranean region, a mixed system of trees and pasture, subject to animal grazing. Farmers need information on pasture production and quality in order to assess the direct effect of tree presence on the productivity of their pastoral system, and to devise management that balances farm production and profitability with sustainable soil management. The main objectives of this work were (1) to evaluate tree influence on soil and pasture parameters and (2) to evaluate the use of proximal sensing techniques that have potential for monitoring aspects related to spatial and temporal variability of pasture productivity and quality in montado ecosystems. Both objectives can support the decision-making process of the farmer. The study field is located in Mitra farm, in Southern Portugal. During October 2015, 24 geo-referenced composite soil samples (12 under tree canopy and 12 outside tree canopy) were collected from the 0.0-0.3m soil layer. The soil samples were analysed for texture (sand, silt, and clay content), moisture content, pH, organic matter, total nitrogen (N), phosphorus (P), potassium (K), magnesium (Mg), and manganese (Mn). The evolution of the pasture was recorded in the 24 sampling points at five monitoring dates: at the end of autumn (December 2015), at the end of winter (March 2016), and then monthly during spring 2016 (April, May, and June). The following pasture parameters were measured: normalized difference vegetation index (NDVI), capacitance, temperature, green and dry matter, ash, crude protein (CP), and neutral detergent fibre. Soil under tree canopy had significantly higher levels of organic matter, N, P, K, and Mg, and better pasture quality while the pasture productivity was higher outside tree canopy. The correlation between pasture direct measurements and sensor parameters was more consistent between capacitance and pasture productivity and between NDVI and CP. The use of fast and efficient tools associated with geo-referenced systems can greatly simplify the pasture monitoring process, which is the basis for estimating feed availability in the field. The knowledge of biomass quality and quantity is fundamental to support decision-making regarding animal stocking rates and rotation among grazing parcels.