Nevertheless the economical importance of the forests and jungles in the state of Durango, there is no evidence about basic research carried out in detail to which statistical analysis have been applied for ecological analysis that allow us to analyze and to relate characteristics of the vegetation and soil impacted due to the forestry utilization, therefore, the objetives of this trial were to know interrelation existent among variables of: vegetation, soil, hydrology and phisiography, and to detect which variables of the soil are affected due to the forestry utilization in the forests of Durango State, México.
Forty two variables were evaluated in 80 sites distributed in 20 properties: nineteen vegetation variables, sixteen soil variables, three physiographic variables and four hydrological variables. Sampling units were located in areas that had been used for at least 2 years before.
Canonical correlation was used to obtain models to determine the magnitud of the relationship among the vegetation variables (dependent variables), and the soil, physiographic, and hydrological variables (independent variables) measured in three different condition of forestry.
Based on production, the 80 sites were divided in three condition classes, twenty one sites with production above 100 m³/ha were classified as class 1 (no impact), twenty seven sites with production between 50 and 100 m³/ha, were classified as class 2 (medium impact), and thirty two sites with a production of 50 m³/ha or less were classified as class 3 (severe impact). Models obtained on each one of those classes showed a high and positive linear correlation (.94, .95 and .96 respectively) with very low significance levels (.0002, .005 and .0000 respectively) indicating reliable results. The dependent variables, which were consistent in the three models were: pine tree production, oak production and fire potential, meanwhile the independent variables were: phosphorus, nitrogen, pH, slope, altitude and clay.
The average of the variables that were consistent in the models of canonical correlation for the three condition classes were compared by using the "t" Student test to identify statistically significant differences between specific variables to determine wich variables were impacted due to the forestry utilization.
In the sites with medium and severe impact, the soil characteristics and pine plantule density show promise a good natural reforestation in the future, but when the values of the variables that showed consistency in the canonical correlation models were compared it was observed that in the sites with severe impact the soil is less acid (pH =  5.9) and phosphorous (2.3 ppm) and nitrogen (0.23%) content is lower.