Runoff and soil erosion plot-scale studies under natural rainfall: A meta-analysis of the Brazilian experience

- Runoff and soil loss records from 71 plot-scale studies in Brazil are compiled.
- The rates vary along the 52 observation sites due to the plots' conditions.
- Sustainable management practices in agriculture can reduce erosion rates.
- Trends, history, and perspectives of plot-scale studies in Brazil are discussed.
- 49% of the experimental studies have 2 years or less of monitored data.

Research to measure soil erosion rates in the United States from natural rainfall runoff plots began in the early 1900′s. In Brazil, the first experimental study at the plot-scale was conducted in the 1940′s; however, the monitoring process and the creation of new experimental field plots have not continued through the years in either country, and are relatively rare in other parts of the world. To better understand runoff and soil erosion rates in Brazil, we review the plot-scale studies that have been conducted across the country. We also evaluated trends, challenges, and perspectives of plot-scale studies in Brazil. Runoff and soil loss records under natural rainfall were compiled from peer-reviewed journals, books, M.Sc. theses, and Ph.D. dissertations, and we organized a database containing the following information: geographic coordinates, region, rainfall, runoff, soil erosion, length of record (plot-years), land cover, tillage system, slope length, and slope gradient. We found mean values of observed annual soil loss ranging from 0.1 t ha y− 1 (grassland and pasture in the southern region) to as great as 136.0 t ha y− 1 (tilled fallow plots, without plant cover, in the northeastern region). Our findings indicated that the southern and southeastern regions of Brazil have the greatest number of runoff and soil loss monitoring sites and length of records (plot-years), corresponding to 67% and 88% of the totals, respectively. In addition, the number of plot-years of data collected has decreased 86% in the last 15 years. However, the number of published documents has increased 10 times in the same period. Around 50% of the experimental studies have 2 years or less of monitored data. In order to reduce the variability of the observed data, develop models, and support decisions, it is important to increase the monitoring period of the experimental sites. Moreover, efforts should continue to allow field observations in all regions of Brazil where data is scarce, in particular paying special attention to the central-western region, where crop production is the main land-use, and along the largest Brazilian agricultural expansion frontier (northern region).