A Guide to Soil Quality Monitoring for Long Term Ecosystem Sustainability on Northern Region National Forests

SOLO HOME > HABITAT TYPES > FOREST HABITAT TYPES OF MONTANA > CHARACTERIZATION AND DISTRIBUTION OF HABITAT TYPES

Forest Habitat Types of Montana
PRODUCTIVITY/MANAGEMENT AND SOIL EXCERPTS

[Excerpted from: Pfister, Robert D., Bernard L. Kovalchik, Stephen F. Arno, and Richard C. Presby. 1977. Forest habitat types of Montana. Gen. Tech. Rep. INT-34. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest & Range Experiment Station. 174 p.]

CHARACTERIZATION AND DISTRIBUTION OF HABITAT TYPES

Soils

Characteristics of the upper 10 cm of soil are summarized in appendix D-1 and as a paragraph in each habitat type description. Soil samples were first examined in the laboratory by an experienced soil scientist (Ronald McConnell, USDA Forest Service, retired) to determine structure, character of horizons, and textural class. Air-dry samples were then weighed, sieved (2 mm) to separate the gravel, and reweighed to determine percent gravel content. The soil separate was analyzed for wet color, dry color and pH, using the water-paste method with a 12-hour delay before reading with a glass electrode pH meter. The soil paste was then used to confirm the textural class designation. The gravel and larger coarse fragments were examined by geologists (Sigrid Asher-Moore and Cynthia Heliker, University of Montana, Missoula) to determine major parent materials.

Soil sampling and analyses were designed to obtain a simple characterization of surface soils for each habitat type, rather than detailed soil-vegetation relationships. Even our limited data make it evident that some habitat types are strongly controlled by edaphic or topo-edaphic factors and have a narrow range of soil characteristics; other habitat types occur on a broad range of soils.

One of the strongest influences of soil on vegetation in Montana is the presence of calcareous parent materials. The PIPO/SYAL h.t., BERE phase as well as the PICEA/PHMA, PICEA/SEST, ABLA/CLPS, and most of the Pinus flexilis h.t.s have a strong affinity for calcareous substrates. Dark-colored surface horizons were generally found in habitat types having either a grass-dominated undergrowth or calcareous substrates, although they were occasionally found in other habitat types as well.

Exposure of mineral soil and rock was greatest on the warm, dry habitat types of the Pinus flexilis, Pinus ponderosa, and Pseudotsuga series. Litter accumulation was lowest in these habitat types.

Several habitat types are associated with water tables close to the surface during part of the year (e.g., the PICEA/EQAR, THPL/OPHO, ABLA/OPHO, and ABLA/CACA h.t.s). Our samples indicate that these habitat types have less gravel, finer textures, lower pH, and deeper litter accumulation than adjacent upland sites; however, more complete soil descriptions would be necessary to adequately document these relationships.

A few research studies have documented soil characteristics by habitat type in the Northern Rockies. McMinn (1952) and Daubenmire (1968a) showed that soil moisture depletion rates differ substantially among habitat types, and this helped explain the differences in vegetation. Work is currently being conducted by the Soil Conservation Service (Harold Hunter, Bozeman) and the Bitterroot National Forest (B. John Losensky, Hamilton) to measure relationships between habitat types and soil temperatures. It is often theorized that vegetation or habitat types can be predicted from soil characteristics. But R. and J. Daubenmire (1968) have emphasized that the correlation between habitat types and soil types (classified on the basis of standard soil profile characteristics) is too weak to allow prediction of habitat types from soil types, or vice versa. We subscribe to this viewpoint as a general rule for several reasons. First, the development of a soil profile reflects a long-term integration of soil forming factors, whereas vegetation development is much more sensitive to current climatic conditions. Second, soil classification systems are not designed to primarily reflect influences on vegetational development; therefore, predictive capabilities should not necessarily be expected. Third, vegetational development depends on many factors, of which soil characteristics is only one. According to the principle of factor interaction, species are able to grow on a wide range of substrates when other factors provide compensatory effects.

In summary, land managers should be cautious about attempting to “shortcut” inventories of either vegetative potentials or soils through the process of “assumed correlations.” Some useful correlations undoubtedly exist; but they must be developed objectively, tested adequately, and extrapolated with caution.

Vegetation: Timber Productivity

Timber productivity is one of the key management implications for which data were collected during this study. Site trees were selected to determine the potential height growth of relatively free-growing trees. One site tree of each species was selected for each stand wherever possible. Site trees showing marked diameter-growth suppression for a period of 10 or more years were rejected during analysis of the increment cores. Old-growth and stagnated trees were not used for productivity estimation. Even though only a single site tree per species per stand was used, the data are reasonably consistent. Comparisons appear to be valid, and the sample size (794 stands) permits comparison of productivity among habitat types as well as within each habitat type.

Determination of site index from height-age data requires specific procedures for each tree species. The number of years to reach breast height (4.5 feet) must be measured or estimated for species having height-total age site curves. If a site curve is not available, a curve from another species must be selected as a substitute. Criteria used to determine total age, as well as sources of site index curves and yield capability data for this analysis, are summarized in table 7.

We used Pinus ponderosa curves for determining Pseudotsuga site index rather than Brickell's (1968) Pseudotsuga curves, because the curve shapes for Pinus ponderosa are more realistic for our data (giving closer estimates for different aged site trees in the same stand). Furthermore, since Pinus ponderosa yield tables are currently used to estimate Pseudotsuga yields in the Northern Rocky Mountains, it is more logical to use Pinus ponderosa site index for estimating Pseudotsuga yields.

We used Alexander's (1967) Picea engelmannii curves for Picea rather than Brickell's (1966) because: (1) Alexander's are based on breast-height age (data available) rather than total age (estimate required); (2) the curve shapes are more realistic for our data (giving closer estimates for different aged site trees in the same stand); and (3) yield data related to the curves are available (Alexander and others 1975). We also used Alexander's (1967) Picea engelmannii curves for several other species that lack site-index curves; because they do not require breast-height age estimates. Thus a possible source of estimation error is eliminated.

The site-index data (base age 50 years) have been summarized by species within habitat types (appendix E-1). Because of regional differences in habitat-type occurrence as well as apparent regional differences in productivity for some habitat types, all timber productivity data were summarized separately for west-side and east-side forests. The mean site index was calculated whenever three or more values were available. With five or more values, a 95-percent confidence interval for estimation of the true population mean was calculated. (The confidence interval narrows with both decreased variability and increased sample size.) The same procedure was used for summarizing basal areas of sample stands.

The maximum heights observed in old-growth stands (>200 years) are presented in appendix E-2. These data can be used for simple comparisons and for identifying sites where height is severely limited.

Although site productivity can be compared by using site index alone, a more useful assessment can be made by using the estimated net yield capability of the site (cubic-foot production). Until managed-stand yield tables are completed, the best approach is to use natural-stand yield tables for assessing yield capability. As stated by Brickell (1970), “Yield capability, as used by Forest Survey, is defined as mean annual increment of growing stock attainable in fully stocked natural stands at the age of culmination of mean annual increment.” (In other words, yield capability = maximum mean annual increment attainable in fully stocked natural stands.)

The curves used to estimate yield capability from site index are presented in figure 48.

Yield capability values are based on cubic feet of all trees (>0.5 inch d.b.h.). The Larix occidentalis curve was derived from Schmidt and others (1976). (Brickell's 1970 curve for this species was only for trees greater than 5.0 inches in diameter.) The Larix curve was also used for Pinus contorta because Brickell's (1970) curves (trees >5.0 inches) are almost identical for the two species, and because natural stand yield data have not been published for Pinus contorta.

The Picea curve was derived from original data used in developing managed-stand yield tables (Alexander and others 1975). We calculated mean annual increment for all trees for 21 of Alexander's fully stocked natural stands near the age of culmination of mean annual increment (ages from 97 to 165 years). A linear regression of yield capability on Alexander's (1967) site index was conducted, converted to site index at base-age 50, and plotted in figure 48. [Yield Capability = −26.0 + (1.84 × 50-year site index.) R² = 0.66]. The other curves were developed by Brickell (1970) from natural-stand yield tables.

The spread in these curves indicates that natural-stand yield capability for a given site index is considerably higher in Abies grandis- and Pinus monticola-dominated stands than for other species. This illustrates the importance of using species specific curves for estimating productivity.

Our best current estimates of yield capability (in cubic feet/acre/year) for each habitat type are shown in appendix E-3 (west-side) and E-4 (east-side). Procedures used to develop these estimates were:

1. Yield capability was estimated for each site tree from appropriate species curves according to the criteria in table 7. These values were plotted by species within habitat types and phases for a visual display of distribution.
2. Mean yield capability for all site trees in each habitat type was calculated and cutoff points were established to approximate 90 percent of the range of our data.
3. For habitat types where stockability appears to limit productivity, a stockability factor was developed. Basal area data for plots in these types were compared with Meyer's (1938) basal area data for fully stocked “normal” stands, following the approach of MacLean and Bolsinger (1973). From these calculations and additional observations, an average mean stockability factor was determined for several habitat types. This factor was multiplied by yield capability for a given site index to determine the adjusted yield capability. A factor of ±0.10 was used to expand the estimated range of productivity.

These current best estimates (appendixes E-3 and E-4) portray both relative productivity of habitat types and the range of productivity within a habitat type. From these, it is possible to assign a ranking or qualitative rating of potential timber productivity of natural stands for use in planning.

As Daubenmire (1976) emphasized, natural vegetation serves as a convenient indicator of productivity over large areas of land. However, productivity within habitat types (appendix E) often shows substantial variability. The following points help explain this variability, and give suggestions for reducing it.

1. Site-index curves were used to obtain productivity data from yield tables. Different height-growth patterns undoubtedly occur on different sites, but data to account for this variation are not available.
2. Yield tables and site curves have not been developed for all species, making extrapolation necessary.
3. Yields of mixed species stands can be estimated by several individual species' yield tables. We found that a range of 30 to 40 cubic feet/acre/year in yield capability was common in individual stands, depending upon the species used for estimation.
4. Some variability in productivity within a habitat type is logical in a natural classification system. The habitat type classification is based on abilities of species to reproduce and mature under competition, not on their rates of growth. The correlation between this and productivity is imperfect. (For instance, in some stands tree roots draw on underground water tables and achieve excellent growth rates, while surface drought limits development of tree seedlings and undergrowth.)
5. Where a more accurate estimate of productivity is needed for local areas, we recommend taking additional site-index samples.
6. It has been suggested that productivity estimates for habitat types could be improved by incorporating classifications of soils, topography, or climate. We have demonstrated a major regional difference in productivity by separating west-side and east-side data (appendix E). Differences in productivity within a habitat type due to topography or soils are also apparent in some local areas. However, because of the limitations of existing site index curves and yield tables, further refinement of productivity data for large areas should be based on more precise methods of measuring productivity.
7. Natural-stand yield capability by habitat type could be estimated more precisely by direct measurements of volume growth, rather than by using site index to enter a yield table based on averages. This would require analysis of existing timber inventory plots representing maximum growth potential or new field measurements.
8. Recent stand growth models (Stage 1973, 1975) utilize growth coefficients based on habitat types. These add a new dimension to yield prediction, provide the basis for developing managed-stand yield tables, and should improve our knowledge of productivity within and between habitat types.

<< Back to top

On to Appendixes >>