Spatial Model for Predicting Communities of Predatory Mites (Mesostigmata) of Leaf Litter

Abstract

The importance of the predatory mites Mesostigmata of the litter lies in its function of population regulation to maintain the balance of soil organisms. Leaf litter is one of the main habitats of these organisms, however, the effect of its properties on the spatial distribution of mites is still unknown. The objective of this research was to know the effect of physical and chemical properties of litter on the abundance of mites. Litter from coniferous, deciduous, and mixed forests was sampled, these forests were located 90 km northwest of Peace River, Alberta, Canada. The generalized additive model (GAM) with the negative binomial distribution modeled the overdispersion of the mite frequency, with an explained deviance of 71.8%. Second degree non-linear relationships were significant between the total abundance of Mesostigmata mites and the variables of Elevation, depth, temperature, humidity and pH. The relationship between abundance and geographic coordinates was fifth order, indicating that the greater abundance of mites was the result of geographic variability. The optimal conditions for the production of Mesostigmata mites are: Elevation below 700 meters above sea level, depth greater than 12 cm, temperature between 11 to 12 degrees centigrade and acidic pH. The least favorable percentage of humidity is between 60 to 90%. The number of mites in the coniferous forest (CD) was statistically equal to that in the deciduous forest (DD), but different from the mixed forest (MX) with p<0.05, the latter having a higher abundance of mites. It is concluded that GAM models are useful to estimate the abundance of mites and predict them in adjacent areas that have not been sampled.