| This study tested the effectiveness of Forestry Best Management Practices (BMPs) to reduce erosion and sedimentation rates in the pine forest of Central Honduras. These BMPs were evaluated during two consecutive years (1999 to 2000) through field experiments.; Timber harvesting experiments were conducted in two watersheds during the rainy seasons of 1999 and 2000. Treatments included timber harvesting using skyline cable (BMPs), tractors (traditional harvesting), animal logging (oxen, BMPs), and undisturbed forest (control, no-harvesting). Simulated rainfall data were used to assess the individual effects of BMPs. Results showed that BMPs such as cable logging, animal logging, slash treatment, and water bars are highly effective in reducing up to 90% of soil particles being eroded from harvested areas.; Soil losses from roads were evaluated for two years including road surfaces and cut and fill slopes. Results showed that sediment production, on both treated and untreated road surfaces were reduced from more than 500 tonnes/km/year to 225 tonnes/km/year with the implementation of BMPs.; Revegetation measures were implemented as BMPs to control erosion on cut and fill slopes of roads. Hyparhenia rufa (Jaragua) and Vetiver zizanioides (Vetiver) were used as revegetation grasses in the first and second years, respectively. Hyparhenia rufa proved to be very effective, reducing up to 90% of erosion and sedimentation rates, while there were no significant differences between the treated and the control plots using Vetiver zizanioides.; The Universal Soil Loss Equation (USLE) and the WEPP Model (Water Erosion Prediction Project) were compared with actual erosion under harvested and no harvested forest to determine their accuracy and precision. The correlation coefficient (r) between WEPP model predictions and actual erosion was 0.80, while the correlation between USLE predictions and actual erosion was 0.39 for the non-harvested forest site. Correlation coefficients for the harvested forest were: 0.29 for WEPP model and 0.45 for USLE. The results indicate that these models may not be very useful in the tropical area such as Honduras. Further model calibration should be made before researchers start recommending these models in tropical environmental conditions. |
