Abstract

An evaluation of hauling operation was realized in five farms located in Sahara, Bataán (83°20’32”O, 10°05’09”N) Limon Province, Costa Rica. Through a series of time and movement observations data were collected on two harvesting conditions: thinning and final cut, two methods of hauling with buffaloes: sulky and chains, as well as the different buffaloes used: buffalo 1 and buffalo. The following hypotheses were verified by means of tests of normality, variance analysis, Tukey’s test (normal data) and Bonferroni (data that did not show normal behavior): There is no statistical difference between: a) silvicultural treatment applied, B) method of trawling used and c) buffalo used for trawling and Ha: there is statistical difference between the groups of data mentioned above.

Variables evaluated were hauling distance in meters, number of hauled logs, total length of the working day, Volume in cubic meters, efficiency percentage, productivity (m³/hr). The analysis was performed for every production cycle and working day.

In order to estimate productivity, two regression models were created according to Hughell methodology (1990). Best-fit equations were selected according to the coefficient of variation and the adjusted R2 determination coefficient and were validated with a 25 % of previously selected data.

Two models were obtained to estimate the wood productivity in melina logs carried by buffaloes in two harvesting conditions: thinning and final cut

a) Productivity (m3/h) = 1,633 + 0,154*V - 0,002*D - 0,109*Du, with R2 of 93,2 %.

b) Productivity (m3/h) = 5,042 + 0,199*V - 0,001*D - 0,275*Du - 0,041*E, with R2 of 86,7 %.

Key words: Aprovechamiento forestal; modelos de regresión; Gmelina arborea