6.10 - CUMULATIVE EFFECTS ON FOREST SOILS
(JRB note: this is a rough copy of a file used to product the large document
"Cumulative Effects of Forest Practices in Oregon" by Robert Beschta
et al., prepared for the Oregon Department of Forestry. It's my intention
to add figures as I can.)
Brief summary conclusions for each of the major effects considered in this chapter are:
(1) Long-term or cumulative effects on forest vegetation may be evident as changes in forest composition and/or structure. Floristic species richness or diversity may be increased or decreased. Overstory composition is altered directly by cutting and regeneration methods. Understory composition responds to resulting changes in site and growing conditions, as well as to direct manipulations. Forest structure may be affected by acquisition or loss of canopies layers or stratified age/size classes. A common effect on forest structure should be overall reductions in average ages and sizes of forest trees and forest stands; a related effect should be reductions in numbers and sizes of snags and fallen dead wood.
(2) Soil compaction has high potential to accumulate through time and space, which should be accentuated as more intensive forestry practices require more entries at shorter intervals.
(3) Long-term and/or chronic soil erosion may occur resulting from periodic or chronic soil exposure. The episode or "pulse" of erosion following a periodic disturbance (i.e., harvest or site preparation) is generally of short duration. This suggests that active erosion should not accumulate between rotations. On site, the next consideration involves whether the soil loss resulting from an erosion episode will be recovered by soil formation during the succeeding rotation. Viewed from a wider spatial scale, periodic disturbances on different sites distributed through time may lead to increases in the overall area-wide erosion rate. Areas of chronically exposed soil (road cut-banks and surfaces) may experience continuous or chronic erosion. As the area of soil in a chronically exposed state is expanded, spatial accumulation of erosion may result.
(4) Forestry practices such as roading, timber harvest and to some extent, application of fire, when conducted in sensitive areas, tend to lead to increases in rates of mass-movement. Evidence is not sufficient to assess effects of practices on long-term average rates of mass-movement. Forestry practices may also tend to increase the magnitude of events when they do occur. On a human (as opposed to a geologic) time-scale, the landscape-scale area affected by mass-movement will probably be limited to a small proportion of the area affected by forestry practices.
(5) There is currently little direct evidence to indicate that harvest removals in themselves lead to soil depletion over several succeeding rotations. Short-term productivity declines should be due to severe associated effects such as compaction, erosion or loss of organic layers. Eventual declines are predicted for highly intensive silvicultural regimes by some researchers. The potential for cumulative decline in soil/site productivity should be greatest when high silvicultural intensity is combined with low inherent productivity and harsh conditions; decline should be least likely when low silvicultural intensity is combined with high inherent productivity and favorable conditions.
(6) If and when significant long-term or cumulative effects on soil biota are realized, they should occur as the result of habitat alteration due to changes in vegetation and forest floor. It is not understood how the long-term or cumulative effects on vegetation or forest floor will affect the magnitude of effects on soil biota.
(7) Evidence suggests that sustained growth response is unlikely to be realized as a cumulative effect of nitrogen fertilization, but such a possibility might exist with other forms of fertilization, if they become widespread. Based on the assumption that duration of these effects is short, cumulative nutrient mineralization and/or leaching should be unlikely to occur as cumulative effects over time on single sites. However, duplication of applications across landscape areas might hypothetically lead to cumulative increases in "baseline" nutrient mobilization and/or leaching for the area as a whole.
(8) Evidence reviewed suggests that the probability of cumulative toxic effects on soil biota due to herbicide application is low, due to low inherent toxicities of the chemicals applied, the low levels at which they are present in soils, and the short duration of persistence of residues in soils relative to rotation lengths. Also based on persistence of residues in soils, it can be concluded that leaching of residues is unlikely to occur as a cumulative effect over time on individual sites. However, residue leaching might hypothetically occur as a cumulative effect over space if applications are duplicated regularly over landscape-scale areas.
(9) Changes in the composition and biomass of forest floors, associated with decomposition, nutrient mobilization, and redistribution from vegetation and forest floors to mineral soil horizons should occur commonly as a cumulative effect in managed forests, as these trends are influenced in similar or complementary ways by several forestry practices, including harvest, site preparation, fertilization and herbicide application. The long-term significance of these trends is unknown.
INTRODUCTION
Up to this point, discussion has concerned direct effects, i.e., effects of a single operation or sequence of related operations on a single site.
Ostensibly, shifting the focus to cumulative effects involves consideration of multiple operations or sequences of operations repeated in
time and duplicated in space. The possible combinations of various operations on various sites is limitless; it will only be possible to to
suggest a few hypothetical combinations in this discussion, to serve as illustrative examples.
This discussion will be based on a series of assumptions, outlined below:
cumulative effects may follow from forestry practices if: (1) practices are repeated closely enough in time that effects of previous
practices have not fully recovered by the time that subsequent practices are conducted, and/or (2) practices are duplicated closely enough
in space that the extent of landscape area affected increases, or that effects resulting from practices on different sites combine and interact
(See section 2.3).
The potential for realization of cumulative effects depends also on individual characteristics of soils, sites and landscapes; the
susceptibility of different soils and sites to direct and cumulative effects is highly variable.
Under the level of cultural intensity typically practical in forestry, soil or site productivity is an inherent characteristic which is unlikely
to be permanently or significantly improved through human cultural practices. This is aside from short- or limited-term improvement in
site quality due to fertilization. An exception may be phosphorus fertilization, which may provide long-term gains in yield (Turner &
Lambert 1988). In general, however, while soil productivity may be maintained or degraded under sustained forestry practices, but is
unlikely to be fundamentally increased. As stated by Smith (1986, p. 7), "site factors that are most subject to long-lasting harm are those
of the soil, which is the most nearly nonrenewable of the resources used in silviculture." Smith (1986, p. 8) continues: "it is entirely
within the realm of possibility to conduct forestry permanently without the degradation that is almost inevitable in agriculture... however,
realization of this potentiality is not automatic."
Concern about cumulative soils effects stems in large measure from concern about potential future adverse effects on soil or site
productivity.
More intensive forestry practices involve more direct and far-reaching interventions in natural processes, and should be likelier to lead to
cumulative changes in or effects on forest soils and vegetation than less intensive practices.
The time required for recovery from effects should be roughly proportional to the magnitude of changes or effects.
As a corollary to the above assumptions, our premise in this discussion is that the potential for forestry practices to exert cumulative
effects on forest vegetation or soils depends on:
the intensity and scheduling of practices.
the resiliency of soils combined with the tendency towards natural recovery.
Silvicultural intensity involves the effort and expense applied to modify forests in ways conducive to human purposes. It reflects the
extent to which human actions direct, manipulate, accelerate, retard, counteract or supersede natural processes. In a very extensive
forestry regime, the practices carried out might comprise selective cutting and fire protection. Yields would be limited to periodic cuttings,
natural regeneration might be relied upon, and vegetative competition regulated through maintenance of canopy cover. In a more intensive
regime (more typical of current practices in the Pacific Northwest), large proportions of the overstory or the entire overstory are often
harvested in one operation, diseases or pests might be curtailed or eliminated as as the seed- or planting-bed is cleared in site preparation,
prompt regeneration is typically secured through planting or direct seeding, competing vegetation might be curbed through chemical,
mechanical or manual means, increased growth and yield might be encouraged through thinning and fertilization, and the new stand is
commonly protected from animal damage through a variety of means.
Scheduling refers to the planning of forestry operations, including temporal and spatial considerations. Temporal considerations include
the lengths of silvicultural rotations or cutting cycles, as well as the scheduling of intermediate (stand tending) practices throughout
rotations. Spatial considerations involve arrangement of managed areas across landscapes. Dispersion or aggregation of areas under
management are two patterns common in the Pacific Northwest. These patterns relate to the arrangement of different land ownerships
across the landscape, on which the varying purposes of different landowners are reflected in different silvicultural practices. These
differences can be seen both in intensity and scheduling. The professional and legal committment of industrial and some public forestry to
meet the ideals of multiple use and sustained yield implies that lands presently dedicated to silviculture will tend to remain in the land base
for indefinite future periods (management direction on Federal lands in the Pacific Northwest is unclear as of this writing). To meet the
continuing demand for forest products, the intensity of silvicultural practices may be expected to increase on some proportion of the
timber land base.
In assessment of cumulative hydrologic or atmospheric effects, a key factor is the emanation of materials or pollutants from a source
(point or non-point), moving downstream or into surrounding areas. In a forestry context, the forest soil or site acts as a source for such
materials, such as eroded soil reaching water sources as sediment, dissolved nutrients or chemicals being leached into groundwater, or
smoke from slash burning moving into the atmosphere. From the hydrological or atmospheric perspective, emphasis is placed on
sediment or smoke entering hydrological or atmospheric systems, rather than on the soil itself as the site of origin. In assessment of soils
effects, by contrast, emphasis is placed on the significance of losses, of soil or nutrients, for example, to the site of origin. Soil transported
from the site in erosion or dissolved nutrients leached to ground water are assumed to be losses from the site, and as such, are not
discussed further in this chapter in relation to the site or its productivity. This is not to deemphasize the significance of the potential for
on-site soils effects to lead to subsequent off-site effects, but to delineate the scope and emphasis of sections 6.11 - 6.19. The idea of
on-site effects moving off-site, combining and interacting with other effects and natural processes is integral to the broad concept of
cumulative effects as discussed in section 2.3.
Materials entering the soil system can also be important in effects on soils, especially in terms of recovery. Examples include atmospheric
deposition of nutrients or chemicals, influx of nutrients in precipitation, deposition of sediments, application of fertilizers, or (very
exceptionally) deposition of fresh unweathered parent materials in events such as volcanic eruptions.
From a soils perspective, other classes of effects may be localized on-site, potentially significant to site condition but not necessarily
having potential for transport of materials or pollutants off-site. One example is soil compaction, which in terms of productivity, is
localized as an on-site effect. However, from a hydrological perspective, compaction may also alter drainage and/or concentrate runoff,
potentially accelerating erosion. Another example might be alterations in on-site nutrient element dynamics, such as the eluviation of
elements such as Calcium or potassium from organic to mineral horizons over a number of years following harvest or fire, but not
resulting in net loss from the site.
Most studies explicitly addressing cumulative effects address hydrological effects directly and soils effects indirectly as sources of
sediment, chemical residues or dissolved nutrients. Much of what is presented here is necessarily conjectural, based on what is
known about direct effects combined with projections of trends in forestry practices. Discussion of cumulative effects will follow a
pattern similar to discussion of direct effects. cumulative effects on forest vegetation will be discussed first, in section 6.11, followed by
discussion of cumulative compaction, soil erosion, mass-movement, nutrient mobilization and loss, effects on soil biota, fertilizers and
pesticides (6.12 through 6.18)
6.11 FOREST VEGETATION - CUMULATIVE EFFECTS
Forestry or silvicultural practice is synonymous with the removal, destruction and/or replacement of some component of the forest vegetation to obtain desired yields or to steer development of the vegetation towards desired goals or conditions. It follows that the effects of forestry on vegetation should tend to be long-term in nature, due to the nature of forestry as a long-term enterprise. Forestry inevitably involves long time periods due to the long life spans of forest trees, especially in the Pacific Northwest, and the long growing periods and rotation lengths of forest stands, spanning time periods of several decades to over a century. The conversion of "wild" forest to "domesticated" or managed forest can be regarded as effectively permanent since the forest is unlikely to be allowed to return to an unmanaged condition, assuming that the lands involved will not be permanently removed from the silvicultural land base.
Smith (1986) acknowledges the potential for cumulative effects on forest vegetation, in particular, those exerted by various cutting regimes:
"It is important to understand the long-term, cumulative effect of cutting operations in building, or degrading, a forest."
This generalization can be extended to other practices, such as site preparation, reforestation and release, since the form taken by harvest
itself in large measure dictates the necessity for associated practices and the means by which they are carried out.
The potential for cumulative effects on vegetation through time should depend on the extent to which vegetative composition is allowed to
occur between repeated cutting cycles or rotations, which would be related to the intensity and scheduling of silvicultural regimes.
Pathways followed by succession would depend on the extent and severity of disturbance due to forestry practices; periods over which
succession would act would depend on rotation or cycle length. In this context, disturbance refers broadly to the overall disturbance of
soil and vegetation involved in a silvicultural regime. With respect to vegetation, it refers to the extent to which vegetation is curtailed,
killed or removed in order to clear growing space for desired species.
Hypothetically, the successional status of vegetation should tend to fluctuate between those conditions prevailing immediately prior to the
end of one rotation, and those prevailing immediately following the initiation of the succeeding rotation (Figure 6.39). However, the
figure is oversimplistic in at least one important respect. The figure implies that all vegetative layers or stories on a site would be at
similar stages of successional development at any given point in a rotation. In fact, species associated with several seral stages may
coexist. If harvest is conducted with extensive soil disturbance, for example, and followed by site preparation and /or release operations,
the site may be to a large extent cleared for influx of colonizing species, as well as for the desired regeneration. To some extent,
succession of both desired and competing vegetation would have been reinitiated. However, if harvest is conducted without extensive soil
disturbance, and site preparation is omitted, the ground vegetation present at the time of harvest may maintain its presence throughout the
establishment and development of the new stand. Succession and development of desired vegetation would have been reinitiated, whereas
that of the ground vegetation would be affected only to a minor degree.
The recovery patterns of forest overstories should be relatively easy to predict over repeated rotations, since they are directly manipulated
and determined by silvicultural systems and rotation lengths. This predictability should be higher for intensive regimes than for extensive
ones. The recovery patterns of understory or ground vegetation follow successional processes, as influenced by disturbance and/or direct
manipulation, and should be less easy to predict.
Effects of forestry practices on forest vegetation will be considered in terms of changes in composition and structure.
Figure 6.39 Hypothetical trends in vegetative succession under various silvicultural systems.
6.11.1 Changes in Forest Composition
Overstory forest composition may be altered directly, as in artificial regeneration (planting or seeding) or site conversion, or indirectly, as
when changes in growing conditions following cuttings favor the regeneration of some species over others. Ground vegetation also
changes in response to altered conditions following silvicultural practices. Effects on ground vegetation and its recovery are discussed in
more detail in section 6.2.
Depending on the intensity and types of forest harvest cuttings, and associated practices, overstory and understory species richness may
increase or decrease. That is to say, succession may be accelerated or retarded, or early, middle and/or late seral species may be allowed to
coexist. In regeneration, emphasis is given to favored commercial species, such as Douglas-fir in western Oregon or ponderosa pine in
eastern Oregon. A factor that should not be overlooked is fire suppression, which has led to changes in forest composition and structure
(Geppert et al. 1984). Geppert et al. (1984) also point out the potential for cumulative effects on the genetic composition of forests
through seed source selection and genetic improvement programs. Locally, this practice may increase genetic diversity.
These generalizations are illustrated by the following examples. Kirby et al. (1991) surveyed stands in Bialowieza Forest, Poland, which
had undergone varying levels of management over the last century. Over- and understory species richness increased in a stand harvested
in small clearcuttings, relative to an undisturbed stand. This was attributed to the entry of early seral trees and vegetation, allowed by the
relatively open canopies. However, stands regenerated by planting following extensive clearcutting and associated grazing, showed more
uniform composition and structure than undisturbed old-growth.
Fralish et al. (1991) observed differences between presettlement, second-growth and old-growth forest stands in the Shawnee Hills of
southern Illinois. The composition and structure of presettlement forest was reconstructed from General Land Office survey records
dating from the years 1806-07. Second-growth forests were defined as those in some stage of recovery from cutting, burning, and/or
grazing, and which also show the influence of fire protection over the past 40-60 years. Old-growth forests were defined as forests allowed
to proceed to natural vegetative climaxes due to the absence of harvest and the interruption of natural fire regimes. On north-facing slopes
and terraces, the composition of second-growth forest was more similar to that of presettlement forest than to that of old-growth (Figure
6.40). The composition of presettlement forest was regulated largely by the natural fire regime; these conditions were continued to some
extent by the cutting, burning and/or grazing which took place in what is now second-growth forest. Early successional oaks played
predominant roles in both these forest types. In old-growth stands, where succession was allowed to proceed by the absence of
disturbance, later successional sugar maple and red oak dominate forest canopies. Second-growth stands have received fire protection for
over 50 years; late seral sugar maple is increasing in prevalence in these stands. The principles illustrated in this study are worthy of
consideration for Oregon forests.
Figure 6.40 Species richness for three forest types in southern Illinois (adapted from Fralish et al. 1991).
After studying successional trends in the pine-hardwood forests of east Texas, Glitzenstein et al. (1986) suggested that the overall effect of
natural and human disturbance over wide areas is the maintenance of species diversity. Disturbance reinitiates succession often enough
that early- or mid-seral species can maintain a foothold among late-seral species. Locally, the direction followed by succession after
disturbance depended on the condition of the vegetation as well as the intensity of disturbance. Disturbance in early seral stands tended to
accelerate succession; disturbance in near-climax stands tended to reinitiate succession.
Simulations conducted with the use of the model FIRESUM supported observations that harvest combined with fire suppression is tending
to increase the dominance of Douglas-fir over ponderosa pine and western larch in the ponderosa pine/inland Douglas-fir forests of central
and eastern Oregon (Keane et al. 1990). A simulation involving complete fire exclusion predicted establishment of Douglas-fir in the
understory, leading to overstory dominance over a period of 140 years. Simulations involving fires at 50-year intervals predicted
understory dominance by Douglas-fir in about 150 years (Figure 6.41). Overstory dominance would be shared by ponderosa pine and fir
in about 300 years, with western larch retiring to a minor role. It was suggested that these trends would be reinforced by harvest of large
overstory ponderosa pine. This view is shared by Geppert et al. (1984) who conclude that partial cutting practices, such as shelterwood
harvest, should exert a cumulative effect on forest composition, in that the regeneration of tolerant species, such as the true firs or western
hemlock would tend to be favored over that of intolerants, such as the larches or pines.
Concern has been raised that the common practice of clearcutting combined with regeneration by planting tends toward the establishment
of pure stands or "monocultures" of commercially favored species, e.g. Douglas-fir, over wide landscape areas. When regeneration is
artificial, as in planting or seeding, the species mix of succeeding stands is determined more by the selection of species for regeneration
than by the ecological effects of the silvicultural system. Nevertheless, Geppert et al. (1984) do not believe that the combination of
clearcutting and planting will lead to permanent or cumulative changes in forest composition, arguing that efforts to establish pure stands
are seldom completely successful. Because planted stock is often not from limited local seed sources, diversity may be increased.
Another trend in forest composition that might be forecast is the eventual decline of hardwoods and shrubs as components of forest stands,
due to efforts to eliminate them through site preparation or release operations. Such effects may be observable locally, but it is unlikely
that efforts to eliminate noncommercial species such as hardwoods would be completely or universally successful, especially with current
streamside management restrictions. Many hardwood or shrub species which compete with conifer regeneration are capable of vigorous
regeneration by seed or sprouting. In many cases, especially in western Oregon, it is difficult to restrain the establishment and/or growth
of hardwoods, shrubs and herbaceous vegetation long enough to secure the establishment of the desired regeneration.
Figure 6.41 Simulated trends in forest composition in the ponderosa pine/inland Douglas-fir forests of central and eastern Oregon. Above: simulated 50-year fire interval; below: simulated fire exclusion or suppression. (from Keane et al. 1990).
6.11.2 Changes in Forest Structure
Structure refers to the presence, sizes and spatial relationships of trees and ground vegetation comprising forest stands. Relevant characteristics include: the size distribution of forest trees, and the presence and prevalence of high shrubs, low shrubs, herbs, grasses and mosses. Overstory structure involves considerations such as the number of stories or stratified age or size classes. Several trends in forest structure may accompany sustained forestry practice, which include: (1) addition or loss of one or more canopy layers or stories, (2) overall reductions in tree age and size. Stands comprised of two or more stories or stratified age/size classes may be consolidated to single storied stands, as when old-growth Douglas-fir is replaced by
second-growth Douglas-fir through evenaged silviculture, or additional canopy layers may be added, as when fire suppression allows the
formation of a tolerant understory in forests of central and eastern Oregon. On the landscape scale, there has been a shift toward younger
age classes as old growth is harvested and plantations established.
In the Bialowieza Forest in Poland, stands that had undergone varying levels of harvesting had smaller overall tree sizes than associated
undisturbed stands (Kirby et al. 1991). Managed stands showed fewer snags and amounts of fallen dead wood than undisturbed stands.
What snags and dead wood were present were smaller in size than in the undisturbed stands. Kirby et al. (1991) concluded that "a large
reduction in the dead wood resource is likely to be typical of all managed forests."
Fralish et al. (1991) found differences in stand structure between presettlement, second-growth and undisturbed stands in southern Illinois.
Presettlement forests were considered to have the lowest densities and basal areas, but the highest mean diameters. Second-growth stands
showed the highest stand density and the lowest mean diameter. Undisturbed stands had the highest basal area and mean diameters
higher than those in second-growth stands (Figures 6.42 and 6.43). Simulation of fire regimes in ponderosa pine/Douglas-fir forests
suggested that artificially lengthened fire intervals would lead to increased stem densities in smaller size classes (Keane et al. 1990).
Harvest of larger overstory trees would also accelerate this trend.
6.11.3 Summary - Effects on Vegetation
Forestry is synonymous with the manipulation, removal, destruction and replacement of forest vegetation. Since forestry is a long-term
enterprise, the effects of forestry practices on vegetation should prove long-term; they may also tend to accumulate through the repetition
of practices through time and their duplication and expansion across the landscape. Many lands presently dedicated to the practices of
intensive silviculture will probably remain so dedicated for indefinite future periods. Simultaneously, the intensity of silvicultural practice
may be expected to increase on some proportion of the land base.
The potential for long-term or cumulative effects on forest vegetation depends on : (1) the intensity and scheduling of practices, (2) rates
of vegetative recovery, and (3) the distribution of silvicultural regimes among various ownerships across the landscape.
Figure 6.42 Stand densities for three forest types in southern Illinois (adapted from Fralish et al. 1991).
Figure 6.43 Stand basal area (above) and mean diameter (below) for three forest types in southern Illinois (adapted from Fralish et al. 1991).
The conversion from "natural" to "managed" forest is permanent for practical purposes. Long-term or cumulative effects on vegetation
may be observable as changes in forest composition and structure. Overstory and understory species richness may increase or decrease;
succession may be reversed or reinitiated by disturbance. Forests may be maintained in earlier or later successional stages, or mosaics of
varied successional stages be allowed to coexist. Obviously, Commercial or other desired species are highly favored in forestry practice;
the decline of hardwoods, shrubs or other ground vegetation as components of forest stands may be realized locally but probably not
widely or universally across landscapes.
Changes in forest structure may include trends toward: (1) addition or loss of canopy layers, stories or stratified age/size classes, and (2)
overall reductions in tree age and/or size, accompanied by increases in stand density. The extent of the former trend (1) should vary
widely, depending on silvicultural regimes and forest type. The latter (2) may prove almost universal in managed forests. Reductions in
tree size should lead to reductions in the sizes and numbers of snags and fallen dead wood. Fire suppression tends to complement the
effects of other practices in the alteration of forest composition and structure.
6.12 SOIL COMPACTION - CUMULATIVE EFFECTS
In this report, compaction is understood conceptually as an immediate effect resulting from the movement of heavy equipment on the soil during the conduction of forestry practices (see Figure 6.44). Forestry practices such as timber harvest, site preparation and road building involve the use of heavy equipment or machinery, and consequently, can involve the occurrence of compaction. This applies to sites on gentle terrain on which the use of heavy equipment is feasible. Relevant characteristics of heavy equipment include the machine type and running gear, and the weight of machine and load; the extent and severity of compaction that may result depend on the number of passes on the soil and the extent of area affected (see section 6.3.1).
Aspects of harvest relevant to the extent and severity of compaction include primarily the harvest system, particularly the harvest system,
and secondarily, the silvicultural system, which may influence the extent of access for machinery required throughout a harvested area.
Site preparation may involve a choice of mechanical methods or prescribed burning. Mechanical methods may involve extensive coverage
of an area by machinery, but may also involve tillage, which may alleviate associated compaction. Burning may involve wide coverage of
a site by machinery, if slash is windrowed or piled, although this is not necessarily the case. Obviously, roads, skid trails and landings are
intentionally compacted to ensure their stability and retard erosion from their surfaces. These areas are typically not expected to remain
productive, but road-related compaction may be hydrologically significant.
The extent and severity of compaction on individual sites is influenced by characteristics such as: (1) topography, which influences
compactive effort, (2) soil properties, such as texture and structure, which influence initial bulk density and soil strength,(3) soil moisture,
which influences soil strength at the time of loading, and (4) organic matter content which may cushion compactive forces (see section 6.3.2).
Figure 6.44 Conceptual cause/effect relationships between forestry practices and site factors in the occurrence of soil compaction.
6.12.1 Accumulation of Compaction through Time
Compaction may accumulate in time when repeated entries involving heavy equipment are made on a site throughout a rotation and
between rotations. Compaction effects might prove cumulative if areas compacted during one operation were re-compacted during
subsequent operations An example might be the use and reuse of "formal" or designated skid trails during successive operations, such as
the commercial thinning of a young stand, following the final harvest of the preceding mature stand. Another might be periodic entries at
decade-long intervals into stands under partial or selective harvest regimes. This scenario is seen as desirable, since the area compacted by
succeeding entries is limited to the existing trail network.
6.12.2 Accumulation of Compaction through Space
Compaction could accumulate in space if areas were compacted during successive entries that were not compacted during the initial entry,
and insufficient time had elapsed for recovery from the initial compaction. An example might be a case in which different skid trail
systems were used in two or more successive entries (Figure 6.45).
Accumulation of compaction would be especially likely if successive entries were carried out in which routes of vehicles and heavy
equipment were unplanned, and therefore could not be duplicated. An example might be tractor skidding with "informal" skid trails.
Others examples might include grapple skidding or mechanical harvesting, in which it is necessary for the equipment to cover a wide area,
in order to gain access to trees and/or logs. Another example might be the piling or windrowing of slash, in which a cat or other vehicle
might cover a large proportion of the area while gathering slash (Figure 6.46). Assuming that old skid trails and landings are reused,
compaction on those areas would be perpetuated in time, although additional areas would not be compacted.
At a wider scale, such as the sub-basin or basin, compaction would accumulate as networks of roads, landings and skid trails were
extended throughout the area, and as harvest expands through the area. In gentle terrain, compacted skid trails would accompany that of
roads and landings. In steeper terrain, assuming cable systems were prevalent, significant compaction would be limited to roads and landings.
When the permanent road network is completed in an area, compaction might not be expected to accumulate further due to road expansion. It could, however, continue to expand, due to expansion of harvest and associated practices. When the road network is complete, it might be appropriate to consider it as a long-term or permanent component of overall compaction in the
basin, rather than as a cumulative component. Compaction due to other practices could continue to accumulate both in time and space.
A characteristic of soil compaction contributing to its potential to accumulate is its duration, which can generally be considered in terms of
decades (See subsection 6.3.4). This increases
Figure 6.45 Hypothetical cumulative compaction over two successive harvest entries.
Figure 6.46 Hypothetical movement patterns of heavy equipment while creating windrows.
the likelihood that compaction from previous entries will not have recovered fully by the time subsequent entries are made.
6.12.3 Summary - Soil Compaction
The significance of soil compaction is related to its potential for effects on soil productivity and its potential to accumulate over timber
growing areas. Accumulation of compaction could occur through a combination of increased intensity of practices and the long time
periods required for recovery. Increasingly mechanized and intensive forestry practices may require repeated entries for harvest cuttings,
site preparation and stand tending.
Greacen & Sands (1980) describe the problem in this way:
"Slow but continued irreversible compaction of sandy soils under our forests may present a problem of considerable significance in the
long term, even if short-term consequences appear insignificant".
This statement was applied specifically to sandy soils in Australia, but it should apply broadly to other soil types in other regions, such as
the Pacific Northwest. Greacen & Sands (1980) also point out that soil damage such as compaction may not always be striking or
obvious. For this reason, the significance of compaction effects may not be recognized until the problem is far advanced.
H.A. Froehlich, who conducted extensive research on soil compaction and its effects in the Pacific Northwest, offers a similar caveat
(Froehlich 1979):
"In my opinion the growth reduction caused by compaction is great enough to indicate that existing skid trails should be used to the
extent possible in future harvests. Repeated entries into a stand markedly increase the area covered by skid trails, and that impact is cumulative."
In addition, the presence of compaction, an on-site effect, can contribute to the occurrence of offsite effects, the most obvious of which is
erosion and sedimentation. Infiltration of precipitation into compacted soil is impeded, increasing the efficiency and concentration of
runoff, which increases its depth, velocity, and erosivity. This suggests that semipermanent or permanent compaction may contribute to
chronic or cumulative surface erosion, which is discussed in the following section.
6.13 SURFACE EROSION - CUMULATIVE EFFECTS
In this report, surface erosion is understood conceptually as an intermediate effect, which may result from the immediate effects of soil
compaction, disturbance, and/or exposure, which can be associated with various forestry practices to various extents (Figure 6.47). Soil
exposure leaves soil susceptible to erosive forces such as rainfall or runoff. Compaction reduces soil porosity directly, by reducing
macroporosity, and exposure may induce surface "crusting"; both of these may serve to increase runoff, divert it, and concentrate it, which
may increase its erosive force.
Aspects of harvest practices relevant to the extent and severity of erosion include the harvest system and the silvicultural system.
Ground-based harvest systems are generally associated with higher levels of soil disturbance and exposure than are cable systems, which
are associated with higher levels than are aerial systems. It is difficult to distinguish definite associations between different silvicultural
systems and rates of disturbance, but the extent of disturbance may be loosely associated with harvest intensity. Prescribed fire, depending
on its intensity, may consume forest floors, resulting in exposure of mineral soil; intense fire may also induce temporary hydrophobicity in
surface soil layers. Mechanical methods involving churning or scarification are intended to disturb and expose soil, as a means of
achieving silvicultural goals, whereas techniques involving chipping or crushing of debris may increase ground cover. Road cut-banks,
fill-slopes and surfaces may remain chronically exposed to erosive forces. Compacted surfaces may divert and concentrate runoff.
The susceptibility of individual sites to surface erosion is influenced by site and soil characteristics such as: (1) soil erodibility, the
inherent susceptibility of a soil to erosion, as determined by soil parent material, soil texture, structure, and organic matter content, (2) the
presence or absence of erosive forces such as runoff, (3) the presence or absence of soil cover, and (4) topography, which influences the
volume and velocity of erosive runoff. In addition to compaction, the single most important soil characteristic influencing rates of erosion
is the presence of ground cover, generally in the form of a litter layer (the forest floor). It will be assumed that erosion rates are negligible
when adequate ground cover is present. Speculation about the potential accumulation of erosion through time and/or space will be based
on assumed patterns of soil exposure, which may be short-term, periodic or chronic.
6.13.1 Accumulation of Erosion through Time
The first case we will examine is that of erosion resulting from repeated operations on the same site. An example would be some form of
harvest followed by site preparation. We assume that enough soil disturbance and exposure result from these operations that some amount
of surface erosion follows. This may be described as a "pulse" of erosion that generally declines rapidly (within 2-5 years) as revegetation
restores soil cover. This is illustrated in Figure 6.48. The severity and duration of erosion will depend on the severity of disturbance and
the the extent of soil exposure.
Another important consideration is whether soil detached and transported on a site is also transported from the site, representing a loss
from the site as well as redistribution on the site. On-site deposition or off-site transportation depend on several factors. Anything causing
erosive waters to infiltrate or slow down can cause eroded sediment to be deposited. Examples of potential deposition sites include
reductions in slope gradient, which retard flow or ravel,
Figure 6.47 Conceptual cause/effect relationships between forestry practices, erosive forces and site factors in the occurrence of surface erosion.
Figure 6.48 Hypothetical surface erosion and cumulative soil loss following levels of high and low soil disturbance. (adapted from
Warrington 1978).
increases in surface roughness, which retard and disperse flow, and the presence of ground cover, which retards flow and allows it to
infiltrate. In forests, soil detached and transported by
rilling and gullying should stand a greater chance of being transported off-site than soil transported by sheet or splash erosion. In hilly
forested terrain, off-site transportation typically means delivery into a stream channel or other water source. In the Washington State
Watershed Analysis Manual, sediment is assumed delivered to streams if a vegetative or "break in grade" buffer is not present between
potential sites of detachment and deposition (TFW 1992). This may or may not be a valid assumption for all sites.
Episodic erosion following operations over several rotations, along with hypothetical cumulative soil loss is represented in Figure 6.49.
At the beginning of each rotation, the duration of the erosion episode is short, as the site revegetates, but the amount of soil lost from the
site may accumulate from rotation to rotation. An important consideration at this point is the rate at which soil formation at a particular
site might lead to recovery of humus and topsoil lost to erosion. This is represented conceptually in Figure 6.50. This figure represents a
situation in which recovery from erosion losses occurs before the succeeding rotation; we would assume that loss of productivity due to
soil erosion would not be a cumulative effect on this site. This would represent a "best case scenario", and may never occur in managed
forests due to slow rates of soil formation.
In contrast, Figure 6.51 represents a situation in which rates of soil formation are not adequate to compensate for erosion losses from
preceding rotations. In such a case, soil loss would be cumulative over these rotations. We would also assume that productivity
reductions due to soil erosion would also be a cumulative effect. In this scenario, soil loss does accumulate, but at a slower rate than if we
assumed no recovery between rotations, as in Figure 6.49.
The approach described above sounds plausible in theory, but in practice, it would prove very difficult for managers to predict rates of soil
formation and associated recovery from previous erosion, or to assess when a site has recovered from erosion due to past practices.
Chronic Soil Exposure. So far the discussion has described how cumulative erosion may result from periodic disturbance and associated erosion episodes. It is assumed that erosion rates decrease rapidly as ground cover is restored. Another possibilty is that of chronic or long-term soil exposure, i.e., once the soil is disturbed and exposed, it shows a tendency to remain exposed and susceptible to erosion for long periods or permanently, if remedial measures are not taken. Examples of areas that may remain chronically exposed include road surfaces, cut-banks and fill slopes, rills and gullies, and landslide scars. (Figure 6.52). In addition, harsh sites undergoing severe disturbance may remain exposed for longer periods, due to delayed revegetation. An example might be a site on a steep southern exposure, which undergoes clearcutting followed by an intense burn.
Figure 6.49 Hypothetical trends in surface erosion rates and cumulative soil loss on a forest site over several rotations, assuming no
recovery between rotations.
Figure 6.50 Hypothetical trends in soil erosion and cumulative soil loss assuming recovery of soil loss before onset of successive rotations.
Figure 6.51 Hypothetical trends in soil erosion and cumulative soil loss assuming incomplete recovery between rotations.
Figure 6.52 Soil surfaces indefinitely exposed to erosive forces.
Such areas may remain exposed to erosive forces indefinitely. Theoretically this should lead to some level of chronic erosion above the predisturbance baseline. Erosion rates would not be consistent throughout the year but should fluctuate seasonally. In western Oregon, rates should be highest in fall and winter, when the bulk of precipitation falls, making erosive runoff available. Episodes of spring snowmelt runoff may also contribute to high erosion
rates. In addition, dry ravel may occur extensively on exposed areas during the dry season. Soil detached and/or transported by dry ravel
may then be available for transport by runoff during the following wet season.
Soil erosion may prove cumulative through time if periodic disturbances occur at intervals too short for soil formation to bring about
recovery, or if soil surfaces are exposed to erosive forces for long periods or indefinitely.
6.13.2 Accumulation of Erosion through Space
To theorize about ways in which erosion may accumulate through space, the concepts applied to cumulative erosion through time
described above are projected over spatial scales large enough to include large numbers of sites, e.g., a basin or sub-basin.
Over a basin-scale area undergoing continuous harvest, some proportion of the harvested area may be exposed (and assumed to be
eroding), and some proportion will be in some stage of revegetation and recovery. The relative proportions would depend on the length of
time forestry had been practiced in the basin, and the rates of harvest over that period. Again theoretically, we would assume that some
portion of the basin is exposed and assumed to be eroding in any given year or period. (Figure 6.53). This suggests that the basin-wide
baseline level of erosion would be increased by some increment. Whether the increase would be significant enough at the basin scale to
rate as a problem might be difficult to assess.
Erosion could also accumulate through space as the area of soil in an indefinitely exposed state is extended across the landscape. This
might occur in any number of ways, such as through the extension of networks of roads and landings; the area occupied by road surfaces,
cut-banks and fill-slopes would be increased. The proportion of an area, such as a basin, occupied by the road network varies highly but
might range from 1-30%, depending on design and silvicultural system (Megahan (1988) in Miller et al. 1992). At the low end of this
range, the soil area that is indefinitely exposed may be negligible; at the high end of the range, the soil area indefinitely exposed may be
large enough to be significant as a potential source of eroded sediment.
Soil erosion may accumulate through space as areas of soil compacted and exposed, periodically and/or indefinitely, are extended across
the landscape.
6.13.3 Summary - Surface Erosion
In the preceding section, hypothetical means through which surface erosion might accumulate through time and/or space have been
presented. The assumption underlying the discussion is that
Figure 6.53 Hypothetical patterns of soil disturbance and erosion susceptibility over four harvest entries in a small watershed.
significant erosion will occur only on areas of exposed soil. On this basis, it is assumed that cumulative or long-term soil erosion will
follow patterns of cumulative or long-term soil exposure, which may be periodic or chronic. Periodic episodes of soil erosion may follow
harvest and site preparation. Chronic erosion may be associated with areas such as roads, skid trails or landings. Geppert et al. (1984)
conclude that cumulative surface erosion should result from the construction and existence of road networks, but that forest harvest and
site preparation should not result in cumulative erosion, except when poorly applied on poor or harsh sites.
6.14 MASS-MOVEMENT - CUMULATIVE EFFECTS
In this discussion, mass-movement is considered an intermediate effect which may result from various forest practices such as soil
disturbance, excavation, filling and compaction in sensitive mountainous terrain. Associated intermediate effects may include: (1)
alteration or interruption of slope drainage patterns, (2) disruption of slope balance or support, (3) reductions in rooting cohesion due to
decay of root networks, and (4) increased soil moisture due to reduced evapotranspiration on the site. These effects may alter the balance
between shear stress and shear strength in the soil mantle (Figure 6.54).
A relevant aspect of timber harvest is the silvicultural system (cutting intensity) which influences post-harvest soil moisture levels, as well
as the extent of mortality in root networks. Depending on its severity and duration, fire can result in mortality of overstory and understory
vegetation, exerting similar effects on soil moisture and root networks. Aspects of forest roads relevant to mass-movement potential
include road location relative to unstable terrain, construction techniques, and the establishment and maintenance of drainage systems.
Natural factors influencing the susceptibility of individual sites to mass-movement include: (1) landform (slope gradient and shape),
which influences the dispersion or concentration of drainage waters, (2) geology, (parent material, orientation of formations and bedding
layers, faulting, fracturing, and jointing) which influence susceptibility to weathering, and the presence or absence of potential failure
planes, (3) soil properties (texture, weathering, and clay mineral development), which influence soil cohesion, internal friction, depth and
drainage, (4) hydrological regimes and events, which influence soil moisture content (short- and long-term) and the extent and depth of
weathering of soil and bedrock, and (5) vegetation, which influences soil moisture seasonally, and supplements soil cohesion through
development of root networks.
Evidence suggests that forestry practices carried out across a landscape and over long periods can increase rates of mass-movement or landsliding in all forms, such as debris avalanches, debris torrents, slump/earthflow, and soil creep. This in itself might be viewed as a basis for concluding that cumulative mass erosion effects have occurred. In the following section, we
suggest ways in which mass-movement and its associated effects might accumulate over time and through space.
It is often not possible to specifically predict the occurrence of a mass-movement event following forestry practices. An alternative
approach is to predict, qualitatively or quantitatively, the
Figure 6.54 Conceptual cause/effect relationships between forestry practices, precipitation and site factors in the occurrence of mass-movement.
probability of mass movement under undisturbed conditions or following forestry practices. In comparison to surface erosion behavior,
mass movements are generally more random and erratic which greatly complicates attempts to predict or assess cumulative effects
associated with mass movements.
Before considering the influence of forestry practices on the probability and rate of occurrence of mass-movement, it is helpful to consider
the rates at which various mass-movement processes occur under undisturbed forest conditions. Recurrence intervals for geological
processes are depicted graphically in Figure 6.55. Soil creep and earthflow operate on relatively short time frames (years to decades); they
are often viewed as persistent or continuous processes, although rates of movement vary widely in response to various factors, including
precipitation trends and forestry practices. Debris slides and avalanches, by contrast, tend to operate on longer time frames (centuries to
millennia); they are considered as episodic processes, which not only occur infrequently but which also exert significant, long-term effects
when they occur. The relative importance that continuous or episodic erosion processes play with regard to overall sediment transport and
landscape formation is likely site- or basin-specific.
6.14.1 Accumulation of Mass-Movement through Time
Following various forestry practices, particularly road building, harvest and/or burning, a site may undergo a period a elevated mass
movement hazard, which is gradually reduced as overstory and understory vegetation is restored on the site. From this it should follow
that slope stability or the probability of failure should fluctuate between various maxima and minima over a series of cycles or rotations.
Sidle (1992) devised a model which predicts trends in the probability of failure on a site undergoing various harvest regimes, such as
clearcutting with or without thinning, partial cutting or shelterwood cutting. The model applies to translational slides in shallow soils
overlying bedrock. Simulated probability trends (Figure 6.56a) suggest that the probability of failure on a given site could vary by a factor
of 5 between clearcutting and partial cutting in which 75% of the overstory is removed. The depiction in Figure 6.56a also assumes that
rotation length is sufficient for full recovery of slope stability between rotations. The simulations conducted by Sidle also suggest that a
cumulative increase in the probability of failure could occur if rotations are progressively shortened, or if rotations are not long enough to
allow full stability recovery before the next harvest cycle (Figure 6.56b).
The simulations described above project the probability of failure over several silvicultural rotations. We have considered trends in slope stability associated with silvicultural practices but
have so far assumed that failure has not yet occurred. That is to say, we have considered trends in the probability of failure leading up to
the occurrence of the first mass movement event attributable to the effects of land use.
If we assume that the slope failure actually occurs, the next question concerns the probability of recurrent failure on that site , i.e., on the
scar or in the flow itself, or in the immediate vicinity.
Figure 6.55 Characteristic frequencies of landscape-forming geological events (from Kelsey 1982b).
Figure 6.56 (a) Hypothetical trends in the probability of slope failure on a forest site over five 50-year rotations. (adapted from Sidle 1992).
(b) Hypothetical progressive increase in the probability of slope failure on a forest site onver five 50-year harvest rotations. (adapted from Sidle 1992).
In some cases, once a site has failed, the potential for recurrent failure might be negligible, until recovery processes have provided a soil
mantle of sufficient depth to "recharge" the failure potential of the site. Such a situation might apply in low-order drainages, swales or
"hollows", which recover from avalanching by surface erosion, root throw and soil creep as a preliminary stage to the eventual recurrence
of failure (Swanson & Fredriksen 1982). In the Van Duzen River basin of northern California, Kelsey (1982a) found that recurrent debris
avalanches following severe winter storms were more likely to affect slopes in some stage of revegetation and reweathering than more
recently failed slopes in the process of recovery. Historically, major incidents of avalanching in headwater drainages were estimated to
occur every 300 to 2000 years (Kelsey 1982a).
The assumption that mass-movement potential on failure sites is reduced below that of the undisturbed forest bears implications for trends in future landscape mass-movement potential. Swanson & Fredriksen (1982) outline several hypothetical situations, two of which will be considered (Figure 6.57):
(1) If the time required for the site to recharge is shorter than the prevailing harvest rotation, then mass-movement potential would be similar to the undisturbed "baseline" condition at the onset of each rotation which implies that the overall "baseline" potential would not increase (Figure 6.57a).
(2) If recharge has not occurred by the beginning of the succeeding rotation, sites are not "ready to fail" when they are reharvested, which
implies that the baseline failure potential may decrease (Figure 6.57b).
However, in other cases, a site on which failure has occurred might be subject to repeated failures, with or without further forestry
disturbances. Examples of this situation might be represented by earthflows, whose headwalls may recede through repeated slumping, or
roadcuts. In addition, slump-earthflow movement may leave over-steepened slopes, which are subject to slides or avalanches (Swanson &
Fredriksen 1982).
6.14.2 Accumulation of Mass-Movement through Space
As forestry practices such as road-building and harvest are extended throughout forested areas, the number of sensitive or unstable areas
disturbed are likely to increase, depending on the rates at which practices are carried out. Commonly observed increases in mass
movement rates associated with harvest, road-building and/or fire demonstrate this tendency, although roading is believed to be the single
most significant cause of accelerated mass-movement erosion.
Two characteristics of mass-movements contribute to the likelihood that areas which normally experience mass failures may tend to
increase in frequency as forestry practices occur across a landscape. These are: (1) the relatively long duration of elevated
mass-movement hazard following disturbance, (ca. 20 years), and (2) the long time-periods required for recovery from mass movement
once it has occurred (10-100+ years). On unstable sites, the period of elevated
Figure 6.57 Hypothetical trends in landscape landslide potential, relative to recovery rates and rotation lengths. (adapted from Swanson &
Fredriksen 1982).
landslide hazard is long enough to generally ensure that a storm of magnitude sufficient to trigger failure will occur. The chances of a
5-year storm occurring over 10-20 years is 89-99%, and the chances of a 100-year storm occurring over the same period are 10-20%
(Geppert et al. 1984). Megahan & King (1985) caution that delays in landslide activity should not be allowed to give a sense of false
security, because of the "virtual certainty" that the consequences of poorly planned and/or conducted forestry practices will become
"painfully clear in the long run".
An important issue at this point concerns the proportions of landscape-scale areas expected to be affected by mass-movement erosion over
time-scales meaningful to forestry planning. As a background, consideration of the natural residence times of soil mantles is helpful in
addressing this issue. Soil mantle residence periods are related to the recurrence intervals of geologic events (Figure 6.55). Residence
periods are the time periods over which the entire soil mantle of a region might be removed and/or replaced by geological and climatic
processes. For example, Kelsey (1982a) estimated the entire soil mantle of drainage basins in Northern California to be affected by
avalanching every 17,000 to 50,000 years. This estimate is based on the assumptions that 2-3% of the area would be denuded every
500-1000 years, and that different areas would be affected by each periodic episode. Swanson et al. (1982) estimated that the residence
time of a soil profile about 3 feet deep at approximately 10,000 years, based on an assumed average soil loss rate of 100 t/km2/yr.
Variation in natural residence times is to be expected, considering that the role played by mass movement in landscape formation varies
between landscapes, depending on variation in geology and climate. The influence of these factors on mass-movement is discussed in
section 6.4.4. An estimate of residence time implies rates of mass-movement, which suggests the proportion of the landscape area that
might show the effects of movement at any given time. In the geologically active, unstable landscape of northern California, Kelsey
(1982a) suggested that all headwater drainage areas would eventually be affected by debris avalanching. In more stable landscapes,
mass-movement might be limited to sensitive areas which occupy a relatively small proportion of the landscape, such as headwater
channels, headwall areas, bedrock swales or concave slopes. Assuming that forestry practices accelerate rates of mass-movement would
suggest that residence times might be reduced, but the magnitude of potential reductions cannot be accurately predicted at present.
Theoretically, it might seem that when forestry practices have occurred across much of a drainage, the potential effects of additional harvesting and roading might become of less concern. A point might be reached, so to speak, when "the places liable to fail have failed". Such a situation might arise, for instance, when the road network is complete and all the area in the timber base has been harvested at least once. However, considering the duration of post-disturbance hazard, it would be decades following the cessation of widespread forestry practices in a given watershed before such a conclusion could be drawn with assurance. If we assume that such a point were ever reached, concern might then center around recurring failure and
erosion on existing failure sites and scars. In addition, widening and reconstruction in a completed road network might renew hazard or instability.
Arguments have been made that partially support the preceding view; Swanson & Fredriksen (1982) summarize comments from various
researchers that forestry practices (in particular,
Figure 6.58 Natural and clearcut-associated avalanche rates over a 40-year period in southern British Columbia (from Howes 1987).
harvest) do not increase the overall rate of mass-movement, but rather alter the timing of movement activity. This conclusion is based on
the premise that the period of accelerated mass-movement activity following an episode of management activity is followed by a period
during which the rate of mass-movement is lower than that for undisturbed areas.
Applying the estimated rates of mass-movement cited above to the time scales over which forestry practices are planned and conducted,
would suggest that the spatial extent of mass-movement activity and effects will be limited to a relatively small proportion of a
landscape-scale area, even accounting for the influence of human activity on mass-movement occurrence. Long-term concern may deal
with persistent or recurrent failure within this area. Of course, the spatial extent of mass-movement is not the sole criterion for judging
their importance.
6.14.3 Summary - Mass-Movement
The previous discussion of the potential for forestry practices to lead to cumulative mass-movement effects has primarily involved several
perspectives concerning the recurrence of mass-movement over time and as forest practices occur across a watershed or landscape. A
study in southern British Columbia provides evidence that clearcut harvesting has increased the rate of occurrence of debris avalanching
over the period 1950-81 (Howes 1987) (Figure 6.58). The clearcut-associated avalanche rate was about 1.7 and 4.1 times the natural rate
during the 1953-1962 and 1963-1981 periods, respectively. The clearcut-associated rate during 1963-1981 was about 1.8 times that for
1953-1962. The natural avalanche rate remained relatively constant over the entire period considered (Figure 6.58).
While the previous discussion has primarily addressed the influence of forestry practices on the rate or frequency of mass-movement, it is
also possible that land use may influence the magnitude of failure events. Kelsey (1982b) suggests that, due to human influence, which
would in part include forestry practices, rainfall events of similar magnitude may result in mass-movement events of greater magnitude
than previously.
Evidence is available which suggests that human activities such as forestry practices can increase rates of mass-movement activity. This
evidence provides a basis for rough qualitative predictions of trends in mass-movement, although it is not yet possible to predict the
occurrence of failure events with precision, or to assess the long-term effects of forestry practices on long-term mass-movement rates.
As the probability of elevated mass-movement activity increases, the probability also increases that sediment and debris will be delivered
off-site, which in steep forested terrain typically means stream channels. This increased potential for sediment delivery is in itself
sufficient to consider mass-movement as a potential cumulative effect.
6.15 NUTRIENT REDISTRIBUTION AND LOSS - CUMULATIVE EFFECTS
In this discussion, nutrient redistribution and loss is understood conceptually as several intermediate effects which may follow from the
immediate effects of removal and mortality of vegetation, and associated creation of varying amounts of slash and debris. Associated
intermediate effects which tend to follow include: (1) altered site microclimate and soil moisture, and (2) accelerated decomposition of
debris and forest floor; these may lead to a "flush" or "pulse" of nutrient elements being released (nutrients may also be volatilized in
intense burns). What may follow is short-or long-term transfer of nutrient elements from vegetation and forest floor into the mineral soil.
Subsequent shifts or transfers of nutrient elements among soil horizons, and alterations of soil chemical processes may also show long
duration. Nutrient element flushes or transfers into the mineral soil may be accompanied by (generally short) periods of accelerated
leaching (Figure 6.59).
Relevant aspects of forest harvest include: harvest intensity (the proportion of the forest canopy removed), utilization standards (the
proportion of biomass harvested, i.e., stem-only harvest, whole tree harvest, etc.), and rotation length or cutting frequency. The extent and
severity of fire effects are related to fire intensity, duration, and amount of vegetation, forest floor or debris consumed. The extent of
nutrient mobilization related to fertilization should depend on application rates, timing of applications, the capacity of various soils to
store added nutrients, and ther extent to which fertilization may stimulate soil mineralization processes. The extent of nutrient
mobilization associated with herbicide application should be related primarily to the effectiveness and extent of vegetative kill following applications.
Site characteristics which influence the extent and course of nutrient redistribution and loss include: (1) soil properties such as parent
material composition, which influences the capacity of soils to receive and retain additional nutrients and moisture, (2) hydrological
regime, which may influence ammounts of drainage waters abailable for percolation through soils, and (3) vegetative recovery following
disturbance, which increases moisture and nutrient uptake.
Consideration of potential for long-term site nutrient depletion is an exception to the rule that cumulative effects on soils have received
negligible attention. On the contrary, the prospect of eventual site depletion under continuous forest culture has received considerable
attention and discussion (Boyle 1976; Dyck & Mees 1990; Evans 1990; Johnson 1983; Kimmins 1977; Powers etk al. 1990; Sollins et al. 1983).
The concept and practice of sustained yield management has been based on the tacit assumption that soil and site potential were constant
or could be assumed constant for practical purposes, and that sustained yield could be realized through the sole expedient of prompt and
consistent regeneration. Concern about reports of alleged declines in forest plantations in second or third rotations have brought this
assumption into question.
Powers et al. (1990) and Evans (1990) have reviewed reported cases of productivity decline in successive rotations. Powers et al. (1990)
concluded that some direct evidence of productivity
Figure 6.59 Conceptual cause/effect relationships between forestry practices, site factors and nutrient redistribution and loss.
decline due to forestry practices exists, but that it is meager, scattered and of limited general applicability. This does not mean that no
potential for decline exists, but that focussed observations spanning time periods of sufficient length to draw firm conclusions have not
been conducted. Evans (1990) came to similar conclusions: "there is little direct evidence that intensive plantation forestry practice will
in itself lead to decline in productivity with successive rotations."
Both reviews concluded that productivity declines, at least in the short term, would be attributable not so much to harvest removals per se,
but rather to associated effects such as compaction, erosion, and loss of organic layers. On this basis, we can conclude that for most
silvicultural regimes on most sites, productivity decline is unlikely to materialize in the short term, i.e., over the first several rotations. If
and when site depletion occurs, it should follow from repeated culture or cropping occurring over longer time periods of indefinite
duration. For this reason, site depletion may be seen as a cumulative effect by definition.
Nutrient dynamics is one of the most complex and difficult classes of effects to attempt to assess or predict. Nutrient cycling and forest
nutrition involve interactions of physical, chemical and biological processes interwoven in webs of baffling complexity. Processes are
poorly understood, and data are difficult to project over time or transport to other locations.
6.15.1 Interactions between Forestry Practices and Site Quality
The potential for nutrient depletion or site degradation is determined by interaction between inherent site productivity or quality,
silvicultural regimes and associated forestry practices. Our assumption is that reduced productivity under silvicultural practice is most
likely to occur when high silvicultural intensity is combined with low inherent productivity and harsh site conditions. Reduced
productivity is least likely to occur when low silvicultural intensity is combined with high inherent productivity and favorable site conditions.
6.15.1.1 Silvicultural Practices
Characteristics of silvicultural practices requiring consideration in assessment of productivity trends include: (1) silvicultural system, (2)
rotation or cycle length, and (3) other silviculture-related nutrient losses or redistributions. See subsection 6.6.1 for a more detailed
discussion of these factors.
A crucial aspect of a silvicultural regime is the "silvicultural system". In a narrow sense, the silvicultural system refers to combinations of
cutting intensity and utilization standards. Cutting intensity means the proportion of standing trees harvested, i.e., clearcutting vs.
shelterwood vs. selection cutting. Utilization standards refer to the portion of a tree considered merchantable and removed from the site in
harvest, i.e., conventional or "stem-only" harvest as opposed to "whole-tree" or "biomass" harvest.
Another critical aspect of a silvicultural regime is the rotation or cycle length. Rotation length determines the intervals at which the site is
entered and disturbed and nutrients are removed, redistributed or lost. Rotation length is especially significant from the point of view of
cumulative effects since it determines the time periods allowed for recovery between harvests. In addition, rotation length determines the
age and developmental stage at which trees are harvested; nutrient demand and the proportion of the overall nutrient pool in the standing
crop vary with age. It has been suggested that rotation age may be of greater significance than harvest intensity (Aber et al. 1978 in
Powers et al. 1990).
Other silviculture-related nutrient redistributions or losses requiring consideration include: decomposition, leaching, erosion and volatilization. For example, site preparation, which forms an integral part of silvicultural regimes of moderate to high intensity, can result in additional nutrient losses, depending on the technique employed. Mechanical site preparation and fire both show potential to disturb or expose soil, and remove, displace or destroy the forest floor remaining following harvest.
Nutrient losses other than harvest removal depend on interactions between forestry practices and site characteristics. An obvious example is soil erosion, which can be significant when soil is exposed on steep slopes, but probably negligible when soil is exposed on level ground (rilling and gullying may provide exceptions to this rule). For a given soil texture, slope gradient and level of soil exposure, erosion might be higher in areas or seasons in which the local precipitation regime tends toward short storms of high intensity rather than those in which low intensity storms of longer duration are prevalent.
6.15.1.2 Site Characteristics, Nutrient Distribution, and Nutrient Cycling
Characteristics which influence the conditions and inherent productivity of a site include: climate, geology, soils, topography (slope,
aspect and landform), and vegetation. Characteristics influenced by those just listed which are also important to consider are nutrient
distribution and nutrient cycling.
Nutrient Distribution. Nutrient distribution refers to how the total nutient pool in the system under consideration is distributed among
various system components such as the vegetation (overstory and understory), forest floor and mineral soil. Figure 6.60 (a,b) shows the
wide variation in nutrient status and distribution possible between sites. Figure 6.60a shows nutrient distributions in two hemlock/spruce
stands in the Oregon Coast Range, 26 and 120 years old, respectively. Figure 6.60b shows the nutrient distribution of a 35 year-old
Douglas-fir stand growing on a river terrace formed by glacial outwash. The total ecosystem nitrogen pool of the hemlock/spruce stands
contains approximately 10 times as much nitrogen as that of the Douglas-fir stand, i.e., 30,600-32,400 lb N/acre vs. 2970 lb N/acre.
Another aspect of distribution is the relative sizes of total and available nutrient pools. Available or exchangeable nutrient pools often
comprise only a fraction of total nutrient pools. Figure
Figure 6.60a Distribution of nitrogen and phosphorus for hemlock/spruce stands at Cascade Head Experimental Forest, Oregon (adapted
from Grier 1976).
Figure 6.60b Distribution of nitrogen and phosphorus for a second-growth Douglas-fir forest growing on coarse, well-drained glacial
outwash soils (adapted from Cole et al. 1968).
6.60a illustrates dramatic differences in the sizes of total and available N pools for the Coast
Range hemlock/spruce stands described above. Available N, composed of NH4+ and NO3-, was estimated at 175 lb/acre, which comprises
approximately 0.6% of the ecosystem N. Total soil N, in which the available soil pool is contained, comprises 98% of ecosystem nitrogen
(Grier 1976).
Figure 6.61 also illustrates these relationships for a spruce/fir stand in British Columbia. The available Ca pool is largest relative to the
total pool (to 1 m), followed by K, P, and N (Kimmins and Hawkes 1977 in Kimmins 1977). Available nutrient pools may be larger than
these examples indicate, if phenomena such as "lithoponics" (i.e., nutrient uptake from "unavailable" forms in rocks) are taken into
account (Kimmins 1977).
Impacts of silvicultural removals and associated losses on available nutrient pools is of greater short-term (first several rotations)
importance than are magnitudes of silvicultural removals and losses in relation to total pools. Using Figure 6.61 as an example,
whole-tree harvest in the British Columbia spruce/fir stand was estimated to result in removal of about half of the available Ca (to 1m),
about one third of the available K , about one quarter of the available P, and about five times the available N. Removals in conventional
harvest would be expected to be about half as large.
Attempts have also been made at construction of nutrient distributions for forest stands of various ages at various locations. Constructs
such as these provide crude "snapshots" of nutrient distribution at a particular stage of stand development.At the least, such distributions
provide rudimentary ideas of proportions of nutrients in trees, vegetation, forest floor and soil, which may be susceptible to change by
forestry practices.
Nutrient Cycling. Rates of cycling between system components determine rates at which nutrients can become available to recovering
vegetation and succeeding tree crops or stands.
A common approach to study of nutrient cycling is the nutrient "budget", which describes nutrient flows into, out of and between nutrient pools. Nutrient flows typically taken into account include: additions in precipitation or atmospheric fallout, litterfall, plant uptake and leaching; underground processes such as root turnover are less commonly considered. The budget is formed of measured or estimated additions and losses of nutrients from the system of concern, which serve to describe net gains or losses of specific nutrients from soil or site at the stand age considered. Figure 6.62 shows a nutrient budget for the hemlock/spruce stands described in Figure 6.60a. The rapidly growing 26 year-old stand is making large demands on the available soil N pool, and moderate demands on the pool of exchangeable P. The
120 year-old stand, by contrast, is accumulating small amounts of N and is making very small demands on the pool of exchangeable soil
P. Figure 6.63 shows similar budgets for for N, P, Ca and K for the Douglas-fir stand described in Figure 6.60b.
The scope of the system considered to house the total nutrient pool is an important consideration in the conception of a budgetting
approach. The total pool may be considered to reside in an
Figure 6.61 Total pools of various nutrients to several soil depths for a spruce/fir stand in British Columbia in relation to available pools
and nutrient removals in whole-tree harvest. (from Kimmins 1977).
Figure 6.62 Above-ground annual nitrogen and phosphorus budgets for hemlock/stands at Cascade Head Experimental Forest, Oregon.
(adapted from Grier 1976).
Figure 6.63 Nutrient budgets for four nutrients in a 35 year-old Douglas-fir forest stand (from Cole et al. 1968).
aggregate of vegetation, forest floor and soil; in this approach, nutrient transfers such as plant uptake, litterfall and litter decomposition are
considered as internal reallocation, rather than as additions or losses. If the mineral soil is regarded as the "central" element of the system,
then uptake is regarded as a loss, and litter decomposition is regarded as a return or addition. In both approaches, harvest removal,
leaching (beyond the rooting zone), and erosion are regarded as losses, and precipitation influx is regarded as an addition. The approach
in Figure 6.62 is mixed. The central element of the nutrient cycling system is considered to be forest floor and mineral soil as an
aggregate. Litterfall is regarded as a return or addition, and uptake is regarded as a removal.
The magnitude and frequency of silvicultural nutrient removals and losses, relative to the sizes of system nutrient pools, combined with rates of nutrient cycling and addition, determine the potential for short- and/or long-term nutrient depletion.
If magnitudes and rates of nutrient removal are high relative to available pools, and nutrient cycling is slow, nutrient deficiencies may manifest themselves fairly quickly.
If magnitudes and rates of nutrient removal are low relative to available to available pools, and/or cycling is rapid, nutrient deficiencies
may not become apparent for long periods.
6.15.2 Prediction of Trends in Productivity
With the considerations above in mind, various attempts have been made at predicting or assessing potential for eventual site/soil
depletion. These range from simple input/output budgets (Bormann & Gordon 1989; Boyle 1973; Adams 1980) to complex computer
models (Kimmins 1988; Kimmins et al. 1988).
Dyck & Mees (1990) describe various approaches taken to assess and predict trends in productivity. Common approaches taken include
site classification, empirical research and computer models.
In site classification systems, sites are organized by criteria such as soils, vegetation or site index. Classes are interpreted in terms of
productivity and susceptibility to effects, which provides an indication of which silvicultural practices might be suitable on particular sites.
An example of an integrated site classification system, the "edaphic grid", (Kimmins 1988), is illustrated in Figure 6.64. A primary
classification by fertility and moisture conditions serves as the basis for associated classifications in terms of productivity, vegetative
associations, and "suitable" silvicultural practices. This system is widely used in British Columbia.
Under the heading of "empirical research", Dyck & Mees (1990) included nutrient budgets, chronosequence studies and retrospective studies. In the nutrient budget approach, silvicultural and non-silvicultural nutrient removals or losses are compared with nutrient additions in order to assess nutrient status and infer trends. In chronosequence studies, a group of stands of similar composition and history, but forming a series of age classes, is used to infer the effects of
forestry practices over long periods (see Frazer et al. 1985 and Snyder & Harter 1985). In retrospective studies, an attempt is made to infer
the effects of past causes from conditions
Figure 6.64 A site classification scheme "the edaphic grid" designed to serve as a basis for silvicultural prescription in British Columbia
(from Kimmins 1988).
observed in the present. Retrospective approaches have been applied successfully, but are hampered by lack of knowledge or control of
multiple causal factors. Field trials involve the establishment of formal field experiments designed to monitor and assess the effects of
silvicultural practices on site condition through time. Such experiments have the advantage of some degree of experimental control, but
long time periods must elapse before results are available.
Sollins et al. (1983) point out that the forestry profession and society can ill afford to wait for the results of long-term studies. For this
reason, a number of computer models of varying complexity have been developed to predict silvicultural effects and productivity trends.
Prominent examples are the LINKAGES and FORCYTE 11 models. For a description of the background and capabilities of FORCYTE
11 (now FORCAST), see Appendix 6B. Computer models have extensive data requirements for calibration, use and validation, which
may prove time consuming to meet, and involve a high degree of labor intensity. Several attempts at prediction of soil productivity trends
under sustained forestry practice are described below. These include three nutrient budget studies and one computer simulation.
6.15.2.1 Examples of Productivity Prediction
An early attempt at nutrient budget evaluation was made by Boyle et al. (1973) for a 40 year-old mixed aspen/hardwood stand in
Wisconsin. It was estimated that soil reserves of N, P, and K, combined with additions from weathering and precipitation could support
an "infinite number" of rotations, even under a whole-tree harvest regime. However, potential deficiencies of Ca were predicted following
nine 30-year whole-tree harvest rotations. It was suggested that alterations in silvicultural practices could be made to circumvent this limitation.
Federer et al. (1989) used data from several nutrient cycling studies in the eastern U.S. to project trends in forest nutrient status 120 years
into the future, assuming the application of whole-tree clearcutting regimes. The nutrient elements N, P, Mg, Ca and K were considered.
It was estimated that three 40-year whole-tree harvest rotations, combined with accelerated leaching attributed to acid precipitation (not a
major factor in the PNW) could deplete organic and soil Ca pools by about 50% over 120 years. Depletion of Mg and K was predicted,
but to a lesser extent. These authors predicted eventual Ca depletion in the east unless countermeasures, such as fertilization, are taken.
The two studies described below place their emphasis on nitrogen rather than calcium. Bormann & Gordon (1989) used an input-output
budget to evaluate the nitrogen sufficiency of several forest types under whole-tree and conventional (stem-only) harvest regimes, such as
red alder, lupin/pine, Douglas-fir and loblolly pine. Plantations and natural stands were compared. They concluded that forest types
containing a permanent or temporary N-fixing component (e.g., red alder, snow brush ceanothus) should generally be able to maintain N
sufficiency even under intensive silviculture. Forest types lacking N fixation would undergo gradual N depletion under intensive
silvicultural regimes (i.e., high utilization, short rotations). Two caveats were added: (1) that the presence of N-fixers does not guarantee
the occurrence of fixation, and (2) even though N depletion may not be a concern in N fixing forests, concurrent depletion of other
nutrients may be, i.e., N fixing forests are thought to have high P demands.
Sachs & Sollins (1986) used the computer model FORCYTE 10 to develop projections of nitrogen cycling and availability over a simulated 540 year period in coastal western hemlock stands. Four silvicultural regimes were simulated, ranging from stem-only harvest at 90-year rotations, to whole-tree harvest at 30-year rotations. The least intensive regime (90-year rotations, stem-only harvest) was projected to maintain stable soil organic matter and N levels. Regimes involving higher utilization, thinning and shorter rotations (30-45 years) were projected to lead to eventual declines in levels of soil organic matter and available N. (Figure 6.65). Interestingly, predicted changes in future yields were much less dramatic, remaining almost constant over the simulation period. Two regimes: (1) stem-only clearcut harvest with two thinnings on 45-year rotations, and (2) whole-tree clearcut harvest with two thinnings on 45-year rotations, showed gradual predicted yield increases over the first two centuries, followed by gradual declines over the remainder of the simulation period. The predicted results of stem-only clearcut harvest at 45-year rotations, involving two thinnings, were more similar to whole-tree harvest at 30-year rotations than to stem-only harvest at 90-year rotations. Yields were predicted to decline over the 540-year simulation period but declines were gradual and small. A number of assumptions about nutrient cycling processes were made in the development of the simulation, such as: rates of fine-root turnover, rates of decomposition of soil organic matter, rates of leaching (NO3-), variations in C:N ratio of soil organic matter, and changes in nitrogen concentration of organic matter during decomposition.
6.15.2.2 Evaluation of Predictive Methods
As mentioned at the beginning of this section, little direct evidence is currently available demonstrating productivity decline as a result of
silvicultural practice. However, available projections, simulations or models provide conflicting predictions.
The questions that remain are: how seriously should we take such predictions, and how should forestry practice be altered to accomodate them?
Predictive methods such as nutrient budgets and computer models are open to criticism. The nutrient budgetting approach has been
criticized as being too static and/or simplistic. Incorporating "snapshots" of fertility status at a particular point in stand development, they
do not account for, or only crudely account for trends in nutrient demand or dynamics through time. Nutrient budgets tend to overlook
underground processes such as root turnover, which may be at least as significant as analogous processes such as litterfall. The budgetting
approach tends to ignore the roles of soil biota in nutrient cycling, which are also highly significant.
Computer models are very time consuming to construct, calibrate and apply, and may have very high sampling requirements. (Adams 1980). Dyck & Mees (1990) caution that the ability to
predict productivity trends with even complex models is hampered by lack of detailed understanding of soil processes, and their
relationships to productivity. They add that the
Figure 6.65 Projected effects of four silvicultural regimes on the forest floor, soil organic matter, and yields over a simulated 540-year period (from Sachs & Sollins 1986).
System A: stem-only clearcut harvest on 90-year rotations, no thinning.
System B: stem-only clearcut harvest on 45-year rotations, precommercial thinning at age 15, commercial thinning at age 30.
System C: Same as B but with whole-tree harvest.
System D: whole-tree harvest on 30-year rotations.
validity of model predictions cannot be adequately verified without results from long-term field studies, which are currently not available.
It may be argued that any attempt at prediction of productivity trends is tenuous. Johnson (1983) points out that consideration of nutrient
distribution and budgets provides a coarse image of the significance of harvest (and other practices) to productivity, but cautions that this
picture is lacking in important respects. He argues that our inability to predict changes in soil nutrient availability precludes meaningful
predictions of changes in soil productivity.
The dearth of direct evidence and the unknown reliability of predictions does not mean that productivity decline cannot occur, or that
concern for productivity decline is groundless. For example, a broad consensus exists that whole-tree harvest and or short rotations hold
high potential for eventual site/soil depletion. Intensive silviculture is a fairly young enterprise, and Raison & Crane (1986) point out that
cumulative effects of forest cropping have had little time to express themselves, much less undergo assessment or measurement.
6.15.2.3 Research Needs
Authors of predictive studies are typically among the first to point out the inadequacy of their methods, and call for the establishment of
long-term field studies to validate their predictions. Pointing out inadequacies and gaps suggests research needs. We will end this section
with a listing of such requirements, compiled in part from Dyck & Mees (1990), Johnson (1983) and Raison (1986):
1. Effects of forestry practices on soil nutrient availability. Relevant practices include: harvest, site preparation, pesticide application, and
fertilizer application.Examples of relevant considerations include: effects on forest floor and soil acid/base balance, and the effects of
canopy openings on decomposition of litter. An underlying assumption is that reductions in pools of available nutrients mean reductions in
productivity.
2. Root dynamics and their relationships to nutrient cycling and productivity.
3. Relationships between soil properties and productivity. Potential productivity declines are often inferred from changes in soil
properties regarded as indicators of productivity, such as macroporosity or organic matter content. Yet the relationships between such
properties and soil productivity, while acknowledged as broadly valid, remain vague and tenuous.
4. Behavior of soil processes and their relationships to productivity. Examples of such soil processes include: litter decomposition,
mineral weathering, atmospheric additions, nutrient immobilization and mineralization, fixation, nitrification, and denitrification.
Researchers such as Dyck & Mees (1990) and Sollins et al. (1983) recommend the conduction of process studies to improve
understanding of soil processes themselves, and to help decide which processes can be regarded as "key" processes, on which attention
should be focussed. Information gained in this way can be used to refine and improve research- or management-oriented models, whether
conceptual, empirical, or computer-based.
5. Roles of soil flora and fauna in nutrient cycling, nutrient availability, and productivity.
6.15.3 Temporal and Spatial Considerations
Temporal considerations of nutrient redistribution and loss involve considerations of silvicultural intensity and rotation length, in
combination with site characteristics and conditions.
A concept helpful in integrating these considerations is that of the "ecological rotation", discussed by Kimmins (1977). In terms of soil
fertility, an "ecological rotation" can be understood as a rotation of sufficient length to allow site fertility to return to a state similar to that
prevailing before rotation (See Figure 6.66). In theory, rotations shorter than the ecological rotation (for a given silvicultural intensity)
should eventually lead to site/soil depletion. Practical implementation of ecological rotations may be hampered by several difficulties:
(1) Determination of ecological rotation may prove a difficult exercise in practice. Setting ecological rotations may require exercise of
knowledge of productivity that we do not possess. However, it may be possible to make broad determination of ecological rotation
through model predictions. For example, results of computer simulations made by Sachs & Sollins (1986) suggest that ecological rotation
in some coastal hemlock/spruce stands may be no longer than 90-100 years (although they may be shorter).
(2) Decisions of rotation length are made not only on the basis of considerations of nutrient cycling recovery, but also on the basis of a
variety of social, political and economic criteria, which compete for top priority.
Spatial considerations of cumulative nutrient redistribution and loss involve the distribution of silvicultural regimes of varying intensity across the landscape. Until recently, probably the most common silvicultural regime on state and industrial lands in Oregon has involved clearcut harvest on rotations of moderate length. However, this trend may be changing. According to the Oregon State Forestry 1992 Annual Reports, partial cut acres exceeded clearcut acres in both western and eastern Oregon during fiscal year 1992 (ODOF 1992). Regeneration is commonly by planting, preceded by site preparation by means of prescribed fire or mechanical methods. Stand tending may include conifer release by mechanical or chemical methods,
precommercial thinning, fertilization and commercial thinning. In Oregon during fiscal year 1992, about 557,400 state and private acres
were harvested, (91,400 acres clearcut and 466,000 partial cut), about 122,000 acres reforested, about 54,000 acres precommercially
thinned, and about 83,000 acres fertilized (ODOF 1992).
Figure 6.66 Hypothetical trends in site productivity under various silvicultural regimes and site conditions (adapted from Kimmins 1977).
6.15.4 Summary - Nutrient Redistribution and Loss
Except in extreme cases, there is little direct evidence currently available to indicate that declines in soil productivity are resulting from
the conduction of forestry practices on most sites. The practice of forestry, especially intensive forestry, is a young enterprise; potential
cumulative effects on soil productivity have had little time to become evident, much less undergo assessment. Focussed observations over
areas wide enough or time periods long enough to draw firm conclusions have not been conducted.
For most silvicultural regimes on most sites, soil productivity decline should not be expected as a short-term effect, as a result of harvest
and plantation establishment per se. Short-term declines are likelier to result from associated effects such as compaction, loss of organic
layers, or erosion. Soil productivity decline might be expected as a long-term effect, occurring gradually or incrementally following
periods of continuous cropping or management. An exception might be very intensive regimes such as whole-tree harvest on short
rotations, which are commonly expected to lead to productivity declines over relatively short periods.
The probability of productivity decline resulting from forestry practices is highest when silvicultural regimes of high intensity are
performed on sites with low productivity or harsh conditions; the probability is least when regimes of low intensity are performed on sites
with high productivity or favorable conditions. Important characteristics of silvicultural regimes include: the silvicultural system (harvest
intensity and utilization standards), rotation or cycle length, and related nutrient losses (decomposition, leaching or erosion). Important
site characteristics include: nutrient distribution and nutrient cycling, as influenced by climate, geology, topography, soils, and vegetation.
With the lack of direct evidence, numerous attempts at prediction of trends in productivity have been made, with approaches such as site
classification, nutrient budgetting, and computer simulation. Such studies predict eventual declines in one or more critical nutrients over
time periods of about two centuries or more, resulting from intensive silviculture.
Temporal aspects of potential cumulative effects on soil productivity involve considerations of silvicultural intensity and rotation/cycle
length, under given site and soil conditions. Spatial aspects involve the distribution of silvicultural practices of varying intensity across
the landscape.
6.16 SOIL BIOTA - CUMULATIVE EFFECTS
We believe that effects of forestry practices on soil biota may be realized through several pathways, which are discussed in more detail in
section 6.7.3 but are listed below in summary form: (1) Direct mortality occurring during the conduction of forestry practices (especially
burning and pesticide application), (2) Alteration of site microclimate (especially harvest and site preparation), (3) Removal of resources
used by soil biota, for example, photosynthate sugars
Figure 6.67 Conceptual cause/effect relationships between forestry practices, site factors and effects on soil biota.
delivered by root systems to rhizosphere organisms. (4) Alteration of the composition and condition of litter and debris layers, including
the litter layer, humus layer, and coarse and fine dead wood, and (5) Alteration in the composition of forest vegetation and directions and
rates of ensuing succession (Figure 6.67).
Conceptually, number (1) can be viewed as an immediate effect, and numbers (2)-(5) as intermediate effects, with numbers (4) and (5) influencing numbers (2) and (3). Our assumption is that long-term or cumulative effects on soil biota would occur as the result of habitat alteration, occurring through changes in vegetation and forest floor, i.e., numbers (4) and (5) above. Following this assumption, we may assume that cumulative effects on soil biota can be expected to the extent that cumulative effects on vegetation and forest floor are expected. This
assumption should be qualified by adding that it is not understood how the magnitude of long-term effects on biota would be related to the
magnitude of effects on vegetation.
As mentioned above, different forestry practices can influence the composition of vegetative communities and the direction and rates of
succession. Conceivably, harvest without burning could lead favor one vegetational community and successional pathway, whereas
burning without harvest could favor another, and harvest and burning could favor yet another. For example, in the case of harvest without
burning, much of the vegetation present on the site at the time of harvest may retain its foothold following harvest and during the
establishment of the succeeding stand, coexisting to some extent with pioneer or early seral vegetation, depending on the extent of soil
disturbance. In the case of burning without harvest, understory growing space should be cleared by the burn, but the composition and
density of post-burn vegetation would tend to be regulated by the density of the overstory canopy. Harvest combined with burning may
tend to lead to dominance of pioneer or early seral species, depending on the severity and extent of the burn. Site conditions resulting from
the conduction of practices such as harvest and/or site preparation can increase contributions made by symbiotic N-fixing plant species
such as red alder and snowbrush ceanothus (Jurgensen et al. 1979). The same would apply to combinations of harvest with other site
preparation techniques.
Computer simulations made using the model FORCYTE 10 predicted that silvicultural regimes of moderate to high intensity could reduce
the biomass of forest floors under coastal hemlock/spruce stands by about 50% over a period of 100 to 120 years, after which a continued
gradual decline was projected (Sachs & Sollins 1986). Silvicultural regimes of "moderate" intensity included stems-only clearcutting
preceded by precommercial thinning at age 15 and commercial thinning at age 30, carried out at 45-year rotations. The high intensity
silvicultural regime simulated comprised whole-tree harvest at 30-year rotations.
Many forms of soil biota are associated with specific vegetative communities or successional stages. This implies that the conduction of
silvicultural regimes should to lead to changes in the composition of associated soil biotic communities (Molina & Amaranthus 1987).
Changes in overstory and understory vegetative composition may lead to changes in rhizosphere conditions, which may induce changes in
floral populations (Jurgensen et al. 1979). For example, species and groups belonging to late successional biotic communities might
disappear on a site if short rotations or cutting cycles are sustained over a period of time.
6.16.1 Recovery
The potential for cumulative effects on soil biota also depend on the ability of various biotic species or groups to recover from
disturbances and the rates at which they are able to recover, relative to the intervals at which forestry practices are repeated. As a group of
soil biota of known significance to forest health and productivity, the factors influencing the recovery of mycorrhizal fungi serve as a good
example.
The recovery and/or reestablishment of mycorrhizae on a site disturbed by harvest and/or slash-burning depend on a number of factors.
These include: host conditions, i.e., the presence, absence, or recovery of hosts or potential hosts, the composition and status of other soil
microflora, the presence and persistence of fungal hyphae and propagules, and the diversity of mycorrhizal species (Perry et al. 1987).
Rapid revegetation and /or reforestation enhances the recovery of mycorrhizal fungi on the site. Fungal hyphae and spores may persist
from several months to several years in the soil in the absence of hosts, although the inoculation potential of a site tends to decrease with
time without the reestablishment of suitable hosts (Perry et al. 1987).
The presence of a diverse fungal community tends to buffer a site against disturbance (Perry et al. 1987). The presence of a variety of
species increases the likelihood that the vegetation and conditions following harvest and/or burning will prove favorable to at least some
of the species present.
As with vegetation, when considering cumulative effects on soil biota, it is important to remember that different biotic groups and species
do not respond to forestry practices and effects in a consistent way. Some groups or species may be favored, others suppressed, or no
significant effect may be apparent. The overall ecological effects with respect to processes such as soil development and nutrient cycling
are unknown, and will prove difficult to assess over wide scales and long periods under field conditions.
6.17 FERTILIZATION - CUMULATIVE EFFECTS
In section 6.8, several effects and potential effects of forest fertilization were discussed, namely, the growth response of forest trees and
vegetation, effects on nutrient pools and cycling, effects on vegetative succession, and effects on soil biota. The potential for the
vegetative growth response and effects on nutrient cycling to occur as cumulative effects will be discussed. Our present information base
is insufficient to meaningfully discuss potential cumulative effects on vegetation or soil biota.
6.17.1 Growth Response
In general, the growth response of forests to nitrogen (N) fertilization should not be realized as a cumulative effect, as the response is of
insufficient duration to be sustained into successive rotations (Miller 1988). However, since the potential for a sustained response cannot
be entirely ruled out as impossible, conditions under which it might occur will be considered. Potential for a sustained or cumulative
growth response would depend on the number of applications of fertilizer over a rotation and their spacing. A growth response might be
sustained or cumulative if applications were repeated at intervals of about 15 years or less. A growth response might also be sustained
indirectly if the large proportion of added nutrients which are immobilized are eventually mineralized and made available to vegetation.
Sustained responses might also be more likely with other nutrients which may commonly be added in fertilizers in the future, such as
phosphorus (P).
6.17.2 Nutrient Mobilization
Fertilization comprises direct addition of nutrients to soils, and may also stimulate mineralization in the soil. Effects on mineralization are
generally hought to be of short duration, which suggests that the probability of cumulative effects are low. Added nutrients immobilized
in soils are assumed to remain unavailable indefinitely, or to be only very slowly available.
The above assumption might be rejected, i.e., an assumption that fertilization effects on soil nutrient pools are proving cumulative might
be adopted if the following conditions could be demonstrated:
(1) If nutrients added in one application or those subsequently mineralized were retained in mineral soils at the time of a succeeding application.
(2) If mineralization processes stimulated by one application remained active at the time of a succeeding application.
Nutrient additions or mineralization due to fertilization may not be of magnitude or duration sufficient to prove cumulative in themselves, but may prove cumulative in combination with other practices that also lead to accelerated decomposition and
nutrient mobilization and redistribution. Cumulative increases in overall leaching levels might in some cases be expected to follow
cumulative nutrient mobilization and redistribution.
Duplication of fertilizer applications across a landscape-scale area might lead to cumulative changes in overall nutrient cycling and/or leaching for the area as a whole, even assuming that these effects were not cumulative over time on individual sites. Accelerated leaching following fertilization is not typically of long duration on any given site; changes in nutrient dynamics may persist considerably longer. If fertilization is being carried out continuously across the landscape, accelerated leaching may presumably be occurring on some site or sites across the landscape on a contin