Forest floors near Port Hardy, British Columbia, Canada

The forest floors of three biogeocoenoses, from northern Vancouver Island, were stratified into LF and H horizons, sampled by use of a stratified random sampling procedure, and analyzed for a number of chemical properties. The biogeocoenoses represent Xeric, Mesic, and Hygric sites. The objectives of the study included: characterization of the forest floors in terms of chemical properties, estimation of property variation and sample requirements, selection of properties best suited for distinguishing the forest floors and organic horizons of each site, and examination of nutrient relationships within the forest floors. To assess variability of chemical parameters, 28 properties were used. The values for total nitrogen, iron, manganese, and exchangeable aluminum were found to increase from Xeric to Hygric sites. Increased leaching losses of potassium and calcium and accumulations of iron, aluminum, and manganese occur as the forest floor materials decompose. The least variable horizon in terms of overall sample requirements was the Mesic LF, while the most variable was the Hygric H. The LF horizons on all three sites tended to be less variable than the corresponding H horizons. The chemical properties found to have the least variability were considered to have the best potential for classifying forest floors. These were pH(H₂0), water content of an oven-dried sample, pH (0.01 M CaCl₂), loss on ignition, total carbon, pH (1 N NaCl), and cation exchange capacity measured at pH 7. Least value for prediction of variability and thus for classification were properties such as total manganese, aluminum, iron, calcium, sodium, thickness, exchangeable calcium and magnesium displaced by 1 N NaCl and 1 N NH₄OAc. The second phase of the study selected properties that were best suited for separating the organic horizons and forest floors of the three sites. The properties examined included those measured in the variability study as well as 12 derived or calculated parameters. Two-way analysis of variance in combination with the Student-Newman-Keuls range test was utilized to determine which parameters would best distinguish the forest floors of sites, the LF and H horizons overall, and the horizons of individual sites. The best properties found to separate Xeric, Mesic, and Hygric forest floors were total potassium, exchangeable sodium measured at pH 7, and the ratio of loss on ignition to total carbon. The best parameters for separating LF and H horizons were total potassium, total zinc, exchangeable calcium displaced by 1 N NH₄OAc, exchangeable potassium displaced by 1 N NH₄OAc and 1 N NaCl, pH measured in water and 0.01 M CaCl₂, loss on ignition, base saturation at pH 7, the ratio of total calcium to total magnesium, and the ratio of total calcium to total potassium. The horizons of the individual sites could not all be separated by any individual parameter. The univariate analysis indicated that the order of increasing difficulty of characterization was the LF and H horizons overall, the forest floors of sites, and the horizons of individual sites. A multivariate analysis was performed to find the combination of variables which best distinguishes the horizons of individual sites. Stepwise discriminant analyses using nine, five, and two variables correctly classified 94%, 81%, and 71%, respectively, of the cases examined. The best approach for using multivariate characterization would be to use a minimum number of variables and to include the parameters total nitrogen, potassium, phosphorus, the ratio of loss on ignition to total carbon, and cation exchange capacity measured at pH 4. The final phase of the study examined nutrient relationships in the forest floor horizons as well as the impact of decaying wood and fine roots on forest floor properties. A correlation matrix was produced for the LF and H horizons. The correlation matrices indicated that certain groups of properties were highly correlated and that only one form of the nutrients calcium, magnesium, and potassium need to be measured. The iron, aluminum, and manganese values were highly correlated, which shows that these elements are involved in similar processes, such as biocycling and podzol formation. Several other highly significant correlations indicate that the nutrients calcium, magnesium, and potassium are predominantly in exchangeable forms and that calcium is the dominant cation in these forest floor systems. .Time available for decomposition is the main factor distinguishing LF from H horizons. A final relationship to be examined was the pH-dependent cation exchange capacity. The values for this property increased significantly downslope and significantly greater values were found in the H horizons of each site - a result of increased functional groups associated with the formation of humus. The total nutrient concentrations of decaying wood were measured and compared to the concentrations found in the corresponding LF and H horizons. The decaying wood was found to be a nutrient-deficient material that is distinct from both LF and H horizons. Bulk density measurements demonstrated that the forest floor and decaying wood materials are not significantly different, although the latter material is more variable. .Therefore, decaying wood represents a substantial input of nutrient-deficient biomass to the forest floor. The nutrient concentrations of fine (