Forest Soil Rehabilitation in British Columbia: A Problem Analysis

Transcription

1 L A N D M A N A G E M E N T H A N D B O O K 44 Forest Soil Rehabilitation in British Columbia: A Problem Analysis 1998 Ministry of Forests Research Program

2 Forest Soil Rehabilitation in British Columbia: A Problem Analysis C.E. Bulmer Ministry of Forests Research Program

3 Canadian Cataloguing in Publication Data Bulmer, Charles Ernest, 1954 Forest soil rehabilitation in British Columbia : a problem analysis (Land management handbook ; 44) Includes bibliographical references: p. ISBN Forest soils British Columbia. 2. Forest site quality British Columbia. 3. Soil productivity British Columbia. 4. Forest productivity British Columbia. 5. Soil conservation British Columbia. I. British Columbia. Ministry of Forests. Research Branch. II. Title. III. Series. SD390.3.C3B C Province of British Columbia Published by the Research Branch B.C. Ministry of Forests 712 Yates Street Victoria, BC V8W 9C2 Copies of this and other Ministry of Forests titles are available from Crown Publications Inc. 521 Fort Street Victoria, BC V8W 1E7 ii

4 SUMMARY Substantial amounts of money are invested each year in forest soil rehabilitation, and the amounts have increased as a result of recent investments made by Forest Renewal BC. The need for research on forest soil rehabilitation arises from our desire to ensure that these funds provide the maximum possible benefit for the lowest possible cost. Rehabilitation projects are currently being implemented, or are planned, throughout the province on a wide range of site and disturbance types. The purpose of this project was to summarize existing information and assess the need for new information to improve the results and cost-effectiveness of soil rehabilitation efforts. The conclusions presented in this paper arose from three sources: review of the scientific literature; field visits with rehabilitation practitioners and researchers working for industry and government in 15 forest districts to operational and research sites representing a wide range of conditions; and consultation with a number of provincial experts and those from other jurisdictions. Our knowledge of soil rehabilitation is limited because few research projects were initiated during the past decade, the range of site types studied so far is limited, and many of the questions we ask about these projects require information on tree growth response, which can only be collected as fast as the trees grow. Despite these limitations, our understanding of how soil processes affect forest productivity has improved substantially in the past decade, and much of this information can be used to solve problems in forest soil rehabilitation. Sections in this report on soil physical processes and soil nutrient cycling processes describe how an understanding of growthlimiting conditions can guide rehabilitation practitioners to strategies for effective and costefficient rehabilitation. Site-specific rehabilitation plans are an integral part of modern rehabilitation projects. These plans provide soil and site information that is used to develop a cost-effective rehabilitation approach. The plans also specify the equipment required and the operating conditions necessary to successfully restore soil productivity on a particular site. Planning, appropriate equipment, and trained operators are crucial to successful soil rehabilitation. People developing rehabilitation plans and implementing rehabilitation projects need information on the effects of various techniques on forest productivity, and on the potential beneficial effects of soil rehabilitation on timber supply. Specific information is needed on the productivity gains resulting from tillage, topsoil replacement, slope recontouring, revegetation, reforestation, soil amendments, and other rehabilitation techniques. Descriptions of these techniques, and the important information gaps affecting their use are provided in this report. iii

5 ACKNOWLEDGEMENTS The author wishes to thank all of the people who contributed information, and those who took the time to visit field sites. A list of these people is included in Appendix 1. Funding for this project was provided by Forest Renewal BC. Lito Arocena, Dave Polster, Paul Sanborn, Margaret Schmidt, Bob van den Driessche provided thoughtful reviews of this report. iv

6 CONTENTS Summary Acknowledgements iii iv 1 Introduction Forest Soil Rehabilitation in British Columbia History Current Situation Future Needs Productivity of Degraded and Rehabilitated Forest Soils Alleviating Growth-limiting Conditions Restoring Soil Physical Processes Processes contributing to soil structure in forest soils Methods for evaluating characteristics of the pore system Restoration of Nutrient Pools and Soil Nutrient Cycles Methods for evaluating nutrient pools and nutrient cycles Techniques for Forest Soil Rehabilitation Tillage Extensive tillage Spot treatments and site preparation equipment Information gaps and research needs Topsoil Conservation and Replacement Information gaps: research needs Slope Recontouring Information gaps: research needs Reforestation and Revegetation Techniques Coniferous and hardwood crop trees Grasses, legumes, and native shrubs for soil amelioration Information gaps: research needs Soil Amelioration: Fertilizers, Amendments, and Mulches Fertilizers Nutrient-poor residues Nutrient-rich amendments Mulches Information gaps: research needs Conclusion References Appendix 1 People Consulted when Preparing this Report v

7 TABLES 1 Results of studies on tree productivity before and after soil rehabilitation Limiting physical conditions for root growth in forest soils Comparison of organic matter and total N pools on disturbed and rehabilitated sites, along with nutrient levels in topsoil and other materials having potential as soil amendments Potential contribution of organic matter and nutrients by adding soil amendments Cost of various implements for extensive tillage of forest soils FIGURES 1 The hypothetical trend of the amount of rehabilitation activity in BC, and two possible scenarios for the future Winged subsoiler tilling a landing at the Aleza Lake Research Forest, Prince George Forest District Rock ripper being used for forest soil tillage in the Penticton Forest District Excellent growth of lodgepole pine planted in 1990 on a landing tilled with a rock ripper (Kamloops Forest District, Jamieson Creek Forestry Road, km 33) Results of a single pass with a rock ripper on a medium-textured landing in the 100 Mile House Forest District Extensive tillage with an excavator equipped with a site preparation rake (Williams Lake Forest District) Excavator with a single-tooth ripper attachment, used for loosening very rocky soils in the Penticton Forest District Close-up of the ripper attachment illustrated in Figure 7, showing the wing attachments to enhance soil tillage (Penticton Forest District) Results of partial tillage with an excavator equipped with a silvatiller attachment (Horsefly Forest District) Evaluating tillage depth (Morice Forest District, Telkwa River Forestry Road, km 15) Topsoil piles commonly found adjacent to bladed areas on level ground (Kalum Forest District) Landing rehabilitated as part of EP 777 in the Williams Lake Forest District The right-hand portion of the landing from Figure 12 (Williams Lake Forest District) Seedling damaged by cattle trampling in a rehabilitated roadside work area (Moffat Road, Williams Lake Forest District) Rehabilitated landing with good pine growth, but poor germination of birdsfoot trefoil because of delayed seeding (Kispiox Forest District) Landing rehabilitated in 1987 in the Quesnel Forest District Surface soil conditions on the landing illustrated in Figure vi

8 1 INTRODUCTION This report summarizes issues and problems in forest soil rehabilitation in British Columbia. It presents an up-to-date review of the scientific literature and the activities of rehabilitation specialists and practitioners working for the forest industry and government. It is aimed at people who carry out rehabilitation projects, and those faced with developing and evaluating cost-effective new techniques for soil rehabilitation. The focus of this report is on techniques for restoring soil productivity, with the implied objective of re-establishing a productive forest ecosystem on a site that has suffered degradation. The causes of forest soil degradation and avoidance techniques are not addressed in detail, as these are dealt with elsewhere (e.g., Lousier 1990; Lewis 1991; and various Forest Practices Code guidebooks). Also, techniques for stabilizing slopes, preventing erosion, and for manipulating and restoring drainage patterns (i.e., techniques for road deactivation) are not specifically discussed, as these practices are also described elsewhere (e.g., Carr 1980; Chatwin et al. 1994; Moore 1994). Various terms, such as restoration, reclamation, and rehabilitation, have been used to describe a range of mitigation activities to counter the effects of environmental degradation. In this report, the term soil rehabilitation refers to activities that aim to improve soil productivity to a state where a productive forest can develop on sites that have suffered some form of soil degradation. This usage reflects terminology adopted in the Forest Practices Code. The objectives of this project were: 1. to review the literature on techniques for forest soil rehabilitation; 2. to describe the current and projected extent of forest soil rehabilitation activities in the province based on consultation with rehabilitation specialists working for industry, government, and other agencies; and 3. to identify gaps in the information base, and the need for future research. 2 FOREST SOIL REHABILITATION IN BRITISH COLUMBIA 2.1 History According to Carr (1983), the Cariboo Regional Policy was the first to require landing rehabilitation in British Columbia. These policy guidelines resulted from landing rehabilitation trials in the 1970s initiated as Project E.P. 777 (Vyse and Mitchell 1977). The landing rehabilitation measures specified in the Cariboo Regional Policy were: 1. ripping or scarification to a depth of 30 cm, if deemed necessary; 2. burning of slash piles; and 3. respreading of topsoil and ash piles on the landing. Mitchell (1982, cited by Carr 1983) evaluated the policy s effect and noted several shortcomings in its implementation, including excessive site disturbance associated with landings, lack of topsoil conservation, ineffective scarification and soil loosening, and poor performance of planted trees. Mitchell made several recommendations and a new landing policy was drafted for the Cariboo Region (Cariboo Forest Region 1983). In subsequent years, several forest regions and districts have developed guidelines and policies regarding the rehabilitation of landings and skid roads (e.g., Prince George Forest Region 1986), but many of the concerns expressed in Mitchell s (1982) report remain valid. In 1986, a workshop held at the University of British Columbia (Still and Lousier [editors] 1988) highlighted the growing awareness among land managers for increased attention to soil conservation programs. Utzig and Walmsley (1988) outlined the costs to the provincial economy associated with forest soil degradation. By 1990, a more-or-less complete framework of standards was in effect throughout the province s forest regions and districts, which restricted the amount of soil disturbance allowed 1

9 during timber harvesting and site preparation, and specified measures for soil rehabilitation. The degree to which these early policies were enforced was variable, as was the quality of the rehabilitation work. In addition, landing rehabilitation objectives were not always clearly stated, and efforts to produce commercial forests were often secondary to those aimed at erosion control or cattle grazing site maintenance. Many landings had their soil loosened and were grass-seeded, but a much smaller number were planted to trees. Even where trees were planted, little follow-up work was conducted to evaluate survival and growth. Several forest companies and forest districts in the interior experimented with road and landing rehabilitation during the 1980s using brush blades, rock rippers, and winged subsoilers. By the late 1980s and early 1990s, substantial landing rehabilitation programs were under way (e.g., Pope and Talbot, Midway; Weyerhauser, Kamloops; West Fraser, Quesnel) in several areas. For example, Crestbrook Forest Industries of Cranbrook began rehabilitating skid roads in Several landing rehabilitation projects that were carried out in the early 1990s (e.g., Pope and Talbot, Midway; Kispiox Forest District; Kalum Forest District) used the winged subsoiler for tilling, followed by seeding with mixtures of grass and legumes, and planting with lodgepole pine. When preparing this report, I became aware of several other forest companies in the interior and in coastal regions that were rehabilitating disturbed lands. It is likely that further contacts with people working for the Forest Service and forest industry will reveal additional examples of rehabilitation projects that were implemented during the past decade. Landing rehabilitation during the 1980s and early 1990s was carried out in a difficult climate, with rehabilitation practitioners facing widespread skepticism about the benefits of restoring soil productivity relative to the costs. Most foresters and researchers did not consider that the existing techniques were capable of restoring full site productivity. Currently, however, many foresters and soil scientists are more optimistic about the potential productivity of rehabilitated sites. Knowledge and techniques have improved, and a renewed commitment to restore the forest exists. Consequently, modern rehabilitation projects are using site-specific techniques that may have been considered impractical only 10 years ago. The regional and district policies and procedures, along with follow-up documents (including Timber Harvesting Branch 1993 and Vancouver Circular Letter 1994), provided the background that led to the soil conservation requirements in the Forest Practices Code. The Forest Practices Code Act, regulations, and guidebooks provide comprehensive, province-wide standards for soil degradation and rehabilitation. These instruments supersede most of the regional and district policies, except where more detail is required. 2.2 Current Situation The amount of soil disturbance allowed on harvested areas is now controlled by the Forest Practices Code, and rehabilitation is required for all temporary access structures. Even before the Code was fully implemented, the amount of land being degraded each year as a result of forestry operations was declining. Key provisions of the Code that affect soil rehabilitation include the following: 1. The amount of permanent access (roads and landings) is limited to that necessary for the harvest, usually less than 7% of the area. 2. All temporary access structures (roads, landings, and trails not required for the long-term management of the area) are to be rehabilitated to restore soil productivity. 3. The amount of dispersed soil disturbance in the net area to be reforested is limited to what is necessary for the harvest, usually less than 5% for coastal sites, and less than 10% for interior sites, although on many sites the allowable level is lower. Therefore, rehabilitation is required for planned (temporary) access that is not part of the permanent access system and for situations where soil is disturbed excessively during forestry operations. The amount of rehabilitation that occurs as part of ongoing operations depends on the logging system and the site characteristics. Forest Renewal BC provides funding to rehabilitate degraded areas or those areas at risk because of past forestry practices. Many of these sites suffer reduced productivity as a result of equipment traffic or road construction. However, while soil rehabilitation activities have increased, large areas remain that require treatment to restore productivity. 2

10 2.3 Future Needs Future needs in soil rehabilitation will depend on numerous factors, including the amount of land suffering degraded productivity, legislation, funding, and public perception. The short-term trend has been towards increasing amounts of rehabilitation as Forest Renewal BC projects reach full implementation, and areas requiring rehabilitation are identified and treated. A considerable inventory of land in the province is unproductive because of soil degradation; the quantity of the unproductive land rehabilitated will depend on the willingness of the public at large to fund this activity in view of other demands on public funds. The availability of funds will inevitably reflect the perceived benefits of rehabilitation, including increased amounts of fibre, forage, or other products, improved water quality, and increased withdrawal of atmospheric carbon, relative to the costs. In the long term, when all suitable areas degraded by past forestry practices are rehabilitated, soil rehabilitation efforts will involve mainly routine rehabilitation of temporary access, as specified in the Code. This situation is illustrated in Figure 1, which shows the amount of land being rehabilitated now compared with the amounts in the past, and the amounts expected in the future. The increasing, then decreasing, trend of expected rehabilitation activity is similar to the trend of road deactivation projected on U.S. Forest Service lands to the year 2045 (U.S. Department of Agriculture 1996). The amount of rehabilitation of temporary access that occurs in the future will depend on several factors including access requirements, site conditions, logging system, success of the rehabilitation efforts, and relative costs of rehabilitation compared to the costs of alternatives that prevent soil degradation. A comprehensive evaluation of the relative costs and benefits of various logging systems that incorporate rehabilitation versus those that prevent soil disturbance has not been carried out. Preventing soil disturbance will be a preferred option in most cases, but situations likely exist where creating soil disturbance followed by routine rehabilitation of the disturbed area will provide cost advantages or other benefits such as access to seasonal wood supplies. For example, cable logging systems produce much less soil disturbance than ground-based systems, but Amount of rehabilitation occurring in ha/yr FIGURE 1 Time > The hypothetical trend of the amount of rehabilitation activity in British Columbia, and two possible scenarios for the future. The arrow points to the situation in 1997, where the amount of rehabilitation work is increasing in response to the Forest Practices Code and the availability of funds from Forest Renewal BC. The declining portion in the future represents a time when backlog work is completed and rehabilitation activity is driven mostly by the requirements of the Code. The diverging lines, in the future, illustrate that the level of future activity will ultimately depend on rehabilitation success and costs relative to alternative strategies for maintaining productivity. the cable systems are generally more expensive. Some forest companies have achieved cost advantages by employing ground-based systems on steep slopes (skidding on constructed trails) followed by rehabilitation of the trails. The initial results from this system are promising, but it is too early to reliably assess the effects on long-term productivity. One of the most important considerations determining the amount of rehabilitation in British Columbia over the long term is the cost. Obviously, we would like to reduce the costs while ensuring the effectiveness of the work. Also, the benefits and costs of alternative approaches for restoring lost productivity on a landscape scale should be examined. For example, it may be cost-effective to simply revegetate degraded areas for erosion control and aesthetics, while restoring lost productivity through more intensive management of other, undisturbed areas. Research aimed at evaluating the costs and benefits of soil rehabilitation and other alternatives will provide information to assist decision-makers faced with such choices. 3

11 3 PRODUCTIVITY OF DEGRADED AND REHABILITATED FOREST SOILS Plant productivity on degraded forest soils may be limited by such factors as: compaction (Greacen and Sands 1980; Froehlich and McNabb 1984); erosion (Miles et al. 1984; Carr 1987a); nutrient displacement (Ballard and Hawkes 1989); unsuitable moisture, thermal, and aeration regimes (Standish et al. 1988; Sutton 1991; Day and Bassuk 1994); and dysfunctional nutrient cycles (Dick et al. 1988). These factors also interact with each other in various ways to produce conditions unsuitable for tree growth. Utzig and Walmsley (1988) acknowledged that measurements of forest productivity on degraded and rehabilitated soils integrate all of these effects. They used results from 18 studies, most conducted in the 1970s and 1980s, to estimate productivity losses caused by harvesting-related disturbance. In their calculations of lost productivity associated with soil degradation in British Columbia, they used a 50% reduction in volume (m 3 /ha) for all types of disturbance, even though the results from individual studies varied widely. Several previous studies (summarized by Andrus [1982], cited by Froehlich and McNabb [1984] and Andrus and Froehlich [1983]) found that tillage of forest soil can improve productivity. Most of these results show the success of tillage, with productivity gains in both survival ( 9 to 38%) and height growth (8 73%). Some more recent studies are summarized in Table 1. As expected for diverse ecosystems with different disturbance types and rehabilitation treatments, the response to machine traffic and rehabilitation treatments varies widely in these studies. To illustrate how different initial soil conditions and effect of treatments on growth-limiting factors influence the success of rehabilitation, two of the studies are described in more detail below. Miller et al. (1996) showed that for skid trails at coastal sites in Washington State initial increases in bulk density did not affect eight-year survival of planted Douglas-fir and Sitka spruce. Tillage improved survival and early growth of western hemlock, but tree heights and volumes of hemlock, Douglas-fir, and Sitka spruce did not differ among treatments after eight years. The soils at these sites contained very high amounts of organic matter in the surface 50 cm, and were generally not adversely affected by soil disturbance associated with skidder traffic. The authors concluded that initially favourable soil conditions, and subsequently favourable climatic conditions can compensate for the potentially negative effects of soil disturbance. These results are likely not applicable to heavily disturbed roads and landings with low organic matter levels in surface soil. However, they may be relevant to studies of isolated ruts and dispersed disturbances, which occur frequently in wetter portions of the landscape where surface soils contain large amounts of organic matter. Carr (1987a) reported that seeding a grass and legume mix and adding 450 kg/ha of fertilizer to a site with eroded subsoil on Vancouver Island (CWHdm) increased foliar nitrogen levels in Douglas-fir and resulted in a 300% increase in height growth compared to the untreated control. The gravelly sandy soil at this site was substantially eroded by road construction, which left a compacted nutrient-poor subsoil exposed at the surface. The rehabilitation treatment obviously alleviated a severe nitrogen limitation on this site, but the longer-term effects of soil compaction and loss of organic matter on trees growing at this site are not known. Although short-term results are often the only information available to researchers attempting to evaluate the effect of management activities on forest productivity, problems emerge when short-term results are used to predict long-term changes in productivity. According to Morris and Miller (1994), acceptable evidence for long-term changes in site productivity must meet three conditions: 1. growth differences must be attributable to differences in site conditions, rather than to differences in resource allocation among target and non-target species or to differences in plant potential; 2. growth results must be available for a sufficient period of time so that the influence of ephemeral conditions is diminished, and the capacity of the site to support trees is emphasized; and 3. adequate experimental control must exist. Morris and Miller (1994) considered measures of site index based on 10 (fast-growing stands) or more than 20 (slow-growing stands) years of data as 4

12 TABLE 1 Results of studies on tree productivity before and after soil rehabilitation Productivity of Productivity of rehabilitated sites rehabilitated sites Disturbance Rehabilitation Species/ (% of undisturbed) (% of disturbed) type treatment age a Survival Ht Vol Survival Ht Vol Source Compact/blade Seed/fertilize Fd/9 n.d. b n.d. n.d. n.d. 325 n.d. Carr 1987a:11 Skid trails Tillage Hw Ss Fd/ Miller et al c Winter landings Auger planting Pl/5 n.d. 47 n.d n.d. Arnott et al d Winter landings Auger planting Pl/5 n.d. 44 n.d n.d. Arnott et al e Deep ruts In-fill/fertilize Pe/5 n.d. 107 n.d. n.d. 158 n.d. Tiarks 1990 Medium ruts In-fill/fertilize Pe/5 n.d. 107 n.d. n.d. 125 n.d. Tiarks 1990 Shallow ruts In-fill/fertilize Pe/5 n.d. 121 n.d. n.d. 110 n.d. Tiarks 1990 Compact trail Bed/fertilize Pt/4 n.d. 108 n.d. n.d. 227 n.d. McKee et al. 1985:17 f Site preparation Shallow till Pp/5 n.d. n.d. n.d n.d. McNabb and Hobbs (rock ripper) 1989:244 a Tree species: Fd = Douglas-fir; Hw = western hemlock; Ss = Sitka spruce; Pl = lodgepole pine; Pe = slash pine; Pt = loblolly pine; Pp = ponderosa pine b n.d. = no data c For comparisons of rehabilitated sites with undisturbed, data from one (Central Park) of three locations were used. For comparisons with disturbed sites, data from all three locations were used. Survival data (values for class 2 3 disturbance from Table 5); height and volume data (values for class 2 3 disturbance from Table 7). d For container seedlings (auger planted = rehabilitated versus dibble planted = control) e For bareroot seedlings (auger planted = rehabilitated versus mattock planted = control) f Data from Table 2. reasonable criteria for meeting the first two conditions. For experimental control, they cited requirements presented in Mead et al. (1991), which include: at least four replications in randomized or randomized complete-block designs; suitable pre-treatment site and crop information for individual experimental plots; measurement plots of at least 400 m 2 ; at least 12 trees left at the end of the experiment; and buffer areas at least 10 m wide surrounding each measurement plot. While some investigations of the changes in longterm productivity that result from soil disturbance meet, or soon will meet, some of the criteria described above, a wide gulf exists between the evidence needed to verify effects of soil rehabilitation on long-term productivity and the information that is available now, or is likely to become available in the near future. Few existing studies (one exception is described by Sanborn et al. 1996) meet the criteria for experimental control outlined above, and only one study, currently dormant (Vyse and Mitchell 1977), comes close to meeting the age criteria. For these reasons, we are left with the use of short-term studies to evaluate the effectiveness of rehabilitation, and therefore need to develop other techniques, most likely based on evaluating soil properties, to predict the long-term productivity of rehabilitated sites. Establishing long-term experimental sites where productivity is assessed for rehabilitated areas is consequently a high priority. Retrospective studies of operational projects initiated in the late 1980s and early 1990s may provide information until the results of long-term experiments are available. Estimates of site productivity are needed to determine the effectiveness of various rehabilitation treatments, and to evaluate the contributions of these lands towards timber supply (e.g., Pedersen 1996). These calculations would likely involve estimating the site index based on tree height and age for rehabilitated areas. Two methods of determining site index are commonly used for young stands (B.C. Ministry of 5

13 Forests 1995). The biogeoclimatic method predicts site index based on the subzone and site series, and is suitable for trees with less than three years growth above breast height. The biogeoclimatic method is based on site factors to predict site index (e.g., Wang et al. 1994). The growth-intercept method predicts site index from height and age measurements taken on selected sample trees. Growth-intercept tables are available for various coastal and interior species, and the method is suitable for trees with at least three years of growth after achieving breast height (1.3 m). With a few exceptions, the trees growing on most rehabilitated sites in British Columbia are too young for reliable site index estimation using the growthintercept method. Evaluating the relative success of rehabilitation treatments should be based on the survival and early growth of trees on rehabilitated sites, and the soil conditions, compared to nearby undisturbed areas. Innovative approaches could be developed to obtain early estimates of site index for rehabilitated sites, possibly using some combination of the existing soil- and site-based approach for young stands (biogeoclimatic method), and some evaluation of tree performance relative to expected performance on nearby undisturbed sites. Such estimates of site index would be useful to evaluate the potential contribution of rehabilitated sites to future timber supply. 4 ALLEVIATING GROWTH-LIMITING CONDITIONS Soil degradation occurs when machine traffic, or other types of disturbance, alter the soil physical, chemical, and biological properties in such a way that site productivity is reduced (Smith 1988). A workshop held in Edmonton, Alberta ( Ziemkiewicz et al. [editors] 1980) sought to develop guidelines to characterize desirable characteristics of reclaimed forest soils after surface mining, and to determine the depth requirements for reclaimed soils. Workshop participants concluded that the chemical and physical characteristics of reclaimed soils could likely be accurately described by referring to published reports (some examples are presented by Ballard 1980), but that soil depth criteria needed to be developed on a site-specific basis. Isolating the effect of individual factors or soil processes that cause reduced productivity may prove difficult because soil disturbance often results in changes to more than one soil process. Therefore, when developing rehabilitation prescriptions to alleviate growth-limiting conditions, the effect of numerous interdependent processes on productivity must be considered. This section describes two major groups of processes that become dysfunctional in degraded soils. These processes must be restored to achieve successful rehabilitation. Soil physical processes, such as infiltration and transport of soil water and air, are controlled by soil porosity and the pore size distribution. The soil pore system is adversely affected by compaction (an increase in bulk density) and puddling (a loss of macropores). Soil physical properties also affect the soil s thermal regime and the resistance of soil to plant root growth. On many degraded sites, nutrient pools are depleted by the removal of forest floor and surface soil. Nutrient cycling processes are also impaired by the removal of nutrients and microbial inoculum. As well, unsuitable conditions for nutrient cycling organisms may exist, particularly the altered soil moisture, aeration, and thermal regimes that result from soil disturbance. The restoration of nutrient cycles involves replacing lost nutrients and organic matter, and providing an environment suitable for the organisms responsible for nutrient cycling. 4.1 Restoring Soil Physical Processes On many sites, alleviating compaction and restoring the pore system so that water and air can move freely to and from plant roots is the most important requirement for restoring soil productivity. When a soil is compacted, a number of conditions are created that inhibit root growth (Bathke et al. 1992). Compaction may increase soil strength to the point where roots cannot penetrate the soil. Also, compaction and puddling cause soil aggregates to collapse, destroying the macropores essential for water and air transport. Excess water drains from macropores rapidly after a rainfall, leaving them filled with air. Oxygen transport to growing root tips, and the movement of physiologically active gases such as 6

14 CO 2 and ethylene away from roots to the soil surface, occurs much more rapidly through air-filled than through water-filled pores. Low temperatures may also be an important factor inhibiting root growth on compacted and puddled soils, which retain water and therefore warm more slowly in the spring. Some examples of growth-limiting values for soil strength (penetration resistance), bulk density, and aeration porosity (the proportion of the soil volume occupied by macropores) are presented in Table 2. Growth-limiting values provide targets for the soil conditions needed to restore productivity. However, some difficulties arise when using critical values to predict root or shoot growth (Greacen and Sands 1980). For example, soil compaction may lead to compact root systems, which occupy less soil volume, but shoot growth may be unaffected if air, water, and nutrient supplies are adequate. Also, a high degree of spatial variation exists in the strength of natural soils that are compacted by logging machinery, especially where traffic is unevenly distributed over the area. Because roots preferentially penetrate pockets of lower strength, they may also penetrate compacted soil layers if sufficient zones of weakness are present. Compact soils resist penetration by plant roots. This may be related to the presence of small pores, which roots cannot enter, or to rigidity of the pore system, which prevents growing roots from deforming the soil. Penetrometer resistance measures soil strength, or resistance to deformation, and is considered to approximate the resistance encountered by plants roots. Thompson et al. (1987) showed that penetrometer resistance and bulk density adequately predicted the performance of hybrid corn on loamy rehabilitated mine soils. Although bulk density was slightly better for predicting the effective rooting depth, penetrometer resistance values were obtained more quickly and therefore allowed more replication, which is an advantage in highly variable soils. Determining a critical value for penetration resistance is complicated because instrument set-up (e.g., cone size and shape), operating conditions (e.g., rate of insertion), and site factors (e.g., soil moisture content, state of soil aggregation) influence the values obtained (Atwell 1993). In many situations, TABLE 2 Limiting physical conditions for root growth in forest soils (as cited by various sources) Soil type Plant species Soil strength (kpa) Bulk density (kg/m 3 ) Aeration porosity Source Various Various 2500 a Greacen et al Pine 2500 Zyuz 1968 Sandy Radiata pine 3000 Sands et al % Grable 1971 Sandy Sands and Bowen 1978 Various Various 1640 b Jones 1983 Sandy Radiata pine Greacen and Sands 1980 c % d Brady 1996 Various Not specified Campbell et al Loess/tilled Ap horizon 3600 Taylor et al Loess/untilled Ap horizon Taylor et al Silt loam Yellow poplar 1400 e Simmons and Pope 1987 Clay loam White oak 1500 Tworkoski et al a Mean value from several studies, range kpa. b Mean value from 10 studies using 20 soils where root growth equaled 20% of optimum. c Extrapolated from a graph of previously unpublished data presented in Greacen and Sands (1980:178). d Individual studies are not cited in the text; this value represents a range described as being commonly accepted by soil physicists. e Average values for trees grown with and without mycorrhizae. The trees were grown at bulk density 1250, 1400, and 1550 kg/m 3. Root length was reduced for the middle treatment; root weight was reduced for the highest level. 7

15 penetrometer resistance values above 2500 kpa result in restricted root growth (Table 2). Determining a critical value for bulk density is not straightforward. Any value will depend on soil (e.g., texture and organic matter content) and site (e.g., climate and soil moisture regime) characteristics, and on the criteria used to evaluate when growth is affected. For example, Jones (1983) considered the critical value to be a bulk density level where root growth was reduced to 20% of optimum. Froehlich and McNabb (1984) summarized studies of the effect of bulk density on growth for six tree species on a range of soil types. They found that a linear relationship existed between the percentage increase in soil bulk density and the percentage decrease in seedling height growth. According to their results, for example, a 50% increase in bulk density led to a 35% decrease in height growth. Subsequently, Miller et al. (1996) showed that even this relationship did not provide a complete picture of the effect of bulk density on productivity. For soils near the Washington coast with surface horizons rich in organic matter and bulk densities well below critical values, increases in bulk density of 40% had little effect on conifer productivity. Daddow and Warrington (1983) summarized several studies and concluded that growth-limiting bulk density values for sandy loams and loamy sands were near 1750 kg/m 3, while clay, silty clay loam, silty clay, and silt soils had growth-limiting bulk density values near 1400 kg/m 3. Lousier (1990) considered bulk density values of kg/m 3 to be growthlimiting for most ecosystems. The variable results described above help to explain why a single critical value for bulk density is unrealistic for all situations on all sites. The relative importance of soil strength and aeration as growth-limiting factors depends on soil water content and texture, as well as other site factors such as climate, plant water demand, and site drainage characteristics (Froehlich and McNabb 1984). Root growth in compacted, coarse-textured soils may be directly related to soil strength, while aeration is often a more serious problem in compacted, fine-textured soils. Soil strength is also expected to limit growth in dry soils, while aeration is more likely to limit growth under wet conditions (da Silva et al. 1994). Seven years after landings were rehabilitated, Kranabetter and Denham (1995) found that aeration porosity and bulk density were both significantly correlated with height and height increment of lodgepole pine, but the relationships were stronger for aeration porosity. Because of highly variable soil conditions and tree growth response on these landings, they used a microplot method to evaluate soil properties in proximity to groups of six trees. For aeration porosity, the values (~ 10%) cited in Table 2 are lower than those extrapolated from Kranabetter and Denham s (1995) data, which showed that values near 20% might be associated with reduced height growth in lodgepole pine. Soil temperature is an important factor affecting root growth in forests (Sutton 1991). Where compacted soils result in poor drainage and higher moisture content, lower soil temperatures often occur because of the high heat capacity of water relative to soil minerals and organic matter Processes contributing to soil structure in forest soils In undisturbed soils, macropores result when individual sand and coarse silt particles are packed together. In medium- and fine-textured soils, macropores only result when silt and clay particles are arranged into larger structural units called aggregates. These aggregates, and their associated macropores, are stabilized by organic matter, plant roots, fungal hyphae, clay binding, and aluminum and iron hydroxides. Root expansion, soil fauna, and freezethaw and wet-dry cycles are also important to create and stabilize soil aggregates. Organic matter is a key ingredient affecting aggregate stability in surface soils. In clay-rich subsoils with low organic matter content, fine films of clay that line the pores are thought to play an important role in stabilizing structure. In some forest soils, iron and aluminum hydroxides may also be important. Processes contributing to aggregate stability are reasonably well understood for agricultural soils (Allison 1968; Tisdall and Oades 1982), but relevant information about forest soils is limited. Borchers and Perry (1992) reported that timber harvesting reduced forest soil aggregation, and that aggregates in a medium-textured soil were an important reservoir of physically protected soil organic matter. Bartoli et al. (1993) found that organic matter content was positively correlated with aggregate size and water stability for a mull A horizon and a Podzolic B horizon. Oxalate-extractable aluminum 8

16 was negatively correlated with aggregate size. Proximity to the roots of 300-year-old fir trees had no effect on aggregate size or stability. In Japan, forest soil aggregates were also stabilized by organic matter, but soil development resulted in aluminum hydroxy cation formation, which contributed to aggregate stability (Itami and Kyuma 1995). These authors showed that reclaimed subsoils with low levels of organic matter had lower aggregate stability and exhibited structural collapse following forest clearing and land levelling. Soils with clays dominated by low-charge minerals, such as aluminum vermiculite and kaolinite, were less sensitive to dispersion (and structural collapse) than those with high-charge clay minerals, such as smectite and illite. Aggregate stability is probably a factor that controls the phenomenon of resettling, or structural collapse (puddling), which is implicated as causing poor soil physical properties after tillage was used to rehabilitate fine-textured soils in British Columbia (Carr 1988b). While some resettling and consolidation is expected after tillage, resettling to the point where the rehabilitation project fails is of concern. The conditions under which resettling is likely to cause failure of the rehabilitation project have not been documented in any detail. Subsoils with high clay content are most likely to experience resettling, but mineral soils with more than 15% clay are also subject to plastic behaviour (Brady 1996), and may also resettle to unfavourable physical conditions. Little information exists about the effects of resettling for the many soils in the province with intermediate clay contents (15 30%). Also, the clay fractions of soils in northeastern British Columbia are dominated by high-charge clays, while soils west of the Rocky Mountains generally have clays with lower charge. Therefore, resettling might pose more of a problem for soils in the northeast. The influence of organic matter on structural collapse has not received specific attention, but is of interest, especially because of the possible role of organic amendments in preserving the benefits of tillage in fine-textured soils. From another perspective, Itami and Kyuma s results (1995) suggested that Podzolic B horizons with large quantities of hydroxy iron and aluminum might provide a more stable structure and be less likely to resettle after tillage. Plant roots and fungal hyphae are important for stabilizing aggregates, but plants are not equal in their ability to promote aggregation. For example, true prairie grasses, which are strongly associated with mycorrhizae and perennial species of the Compositae are associated with soils having larger water-stable aggregates than non-prairie grasses, such as Agropyron repens and Bromus inermis (Miller and Jastrow 1990). Plants with long-lived root systems, and those more dependent on mycorrhizal associations will probably promote soil aggregation to achieve optimal growth. Detailed studies of these attributes for native forest grasses, herbs, and shrubs would be of interest to evaluate their potential role in stabilizing soil structure Methods for evaluating characteristics of the pore system One of the most common measures of soil compaction is bulk density. Core samplers provide a rapid method to collect bulk-density samples (Culley 1993), although excavation methods are more suitable in rocky soils (Blake and Hartge 1986). Nuclear methods can provide large amounts of data quickly, but the reliability is also limited by rocks, and licensing is costly. Predictions of soil productivity based on bulk density could be improved if individual bulk-density samples were also analyzed for organic matter and moisture content. This could provide a means of isolating the effects of these important factors on soil bulk density. While it is possible to calculate total porosity from bulk density and a knowledge of the density of individual soil particles, the aeration porosity is a much more sensitive indicator of soil physical condition. Aeration porosity is a measure of the large pores that drain rapidly, and is usually evaluated by determining the percentage of soil volume occupied by air-filled pores at low water tension (commonly MPa). While determining bulk density and aeration porosity is reasonably straightforward in principle, considerable care is required to produce consistent results, especially for aeration porosity. Also, since both measurements rely on small samples (< 0.5 L), considerable replication is required to produce reliable estimates. Soil strength is measured in several ways, but the most commonly used measurement relevant to plant root growth is penetration resistance to a cone 9

17 penetrometer (e.g., Vepraskas and Miner 1986). Under good conditions (a major limitation in the use of penetrometers occurs where soils have high coarse fragment content), individual measurements of soil strength taken with a cone penetrometer are more quickly obtained than bulk density or aeration porosity values. Penetration resistance was closely correlated with bulk density for loamy soils in the British Columbia interior (Smith and Wass 1994a, b), and was fairly reliable for characterizing various disturbance types associated with skid road construction and stumping. Penetration resistance has potential for more widespread use as an index of soil physical condition in forest soils in British Columbia. However, for the method to have predictive ability, the relationships between penetration resistance, soil moisture content, bulk density, aeration porosity, and seedling growth need working out for various forest site types. Other methods of evaluating soil physical condition on disturbed and rehabilitated sites may also have application in studies for evaluating soil rehabilitation. Air permeability, infiltration rate, and hydraulic conductivity can be determined and provide information about the characteristics of the pore system. Direct measurement of soil air composition would determine when a soil aeration problem exists. Image analysis of intact samples could also prove particularly helpful when evaluating the factors that affect the stability of soil aggregates formed by various rehabilitation treatments. 4.2 Restoration of Nutrient Pools and Soil Nutrient Cycles Many roads and landings have low nutrient content. This condition results primarily when topsoil is displaced during construction and levelling. Nutrient losses associated with removing surface soil and forest floor are well documented (Smith and Wass 1985; Carr 1987b, 1988a). Several rehabilitation techniques have the potential to restore nutrient pools on degraded sites, including topsoil replacement (Carr 1987a; Kranabetter and Osberg 1995), fertilization (Carr 1987a), establishing nitrogen-fixing plants (Carr 1987a; Power 1994), and amending with nutrient-rich byproducts (McNab and Berry 1985; Bauhus and Meiwes 1994; Rose 1994). In general, measures of nitrogen status are more likely to predict productivity on forested sites than measures for other nutrients (Powers 1980; Klinka et al. 1994). Table 3 provides examples of organic matter and nutrient contents that represent typical values for disturbed, undisturbed, and rehabilitated ecosystems. The potential to enhance site nutrient pools with organic amendments is illustrated in Table 4. Forest floor and topsoil removal and other types of disturbance, and the effects of subsequent soil rehabilitation techniques on nutrient-cycling organisms and processes, are more subtle than the effects on nutrient pools. The effects on forest productivity are more difficult to evaluate. When soils are compacted, some studies show that populations of nutrient-cycling organisms and biological activity are reduced, but can be restored by soil rehabilitation treatments. For example, Dick et al. (1988) observed reduced microbial biomass carbon and lower enzyme assays on compacted skid roads compared to undisturbed clay loam soils in Oregon. Rehabilitation treatments involving subsoiling alone, or subsoiling and discing, seemed to restore the biological processes. All of these observations were made four years after site treatment. Somewhat different results were obtained by Smeltzer et al. (1986) for a loamy sand soil under mixed northern hardwood forest in Vermont. In this coarse-textured soil, total fungal and bacterial populations were reduced on compacted plots compared to control plots, and a surface mulch applied after compaction had little effect on populations levels. However, in contrast to the results of Dick et al. (1988), the effect of compaction was relatively short-lived four years after treatment no differences were evident in populations between the treatments. The loamy sand soil was apparently less sensitive to compaction than the clay loam soil. The effect of compaction on nutrient-cycling processes is partly related to changes in soil moisture and aeration. Linn and Doran (1984) showed that microbial activity in agricultural soils, as indicated by CO 2 and N 2 O production, was closely related to the percentage of water-filled pores. Maximum microbial activity occurred at 60% water-filled pore space (this water content represents field capacity for a hypothetical soil with 20% aeration porosity and bulk density of 1400 kg/m 3 ), and declined rapidly above 70% water-filled porosity (this water content represents field capacity for the same hypothetical soil with 12% aeration porosity). These results suggest that soils 10

18 with aeration porosity values below 12% may experience impaired nutrient cycling for extended periods after rainfall. Removing nutrients by disturbing forest floors and topsoil results in changes to many soil physical and biological processes. Some of these changes are possibly similar to those arising from site preparation, where survival and early growth of seedlings appears to depend as much on temperature and moisture changes as on nutrient status. Shortterm results might prove misleading when used to predict long-term performance, but the short-term changes observed by Munson et al. (1993) and Ohtonen et al. (1992) provide insight into the potential effects of disturbed and rehabilitated soil on soil nutrient cycles. Munson et al. (1993) showed that the effects of controlling vegetation were greater than those of TABLE 3 Comparison of organic matter (om) and total N (tot N) pools on disturbed and rehabilitated sites, along with nutrient levels in topsoil and other materials having potential as soil amendments. The organic matter and nutrient pools presented here strongly depend on the soil depth chosen in the original study. Since the rooting depths vary, these results should be interpreted as only general trends within a particular row, and are not suitable for comparisons between rows. Undisturbed kg/ha Rehabilitated kg/ha Degraded (kg/ha) Description om tot N om tot N om tot N Reference Burned windrows a Ballard and Hawkes 1989 Mine soils b Heilman 1990 Fort St. James Carr 1988a Vanderhoof c Carr 1987a Koksilah c Carr 1987a a Calculated from the authors data for forested sites (15 cm depth) at Carp Lake. Degraded sites are scalped areas between burned windrows; rehabilitated sites are under the burned windrows. This example relates to a common observation that the best tree growth near landings is often where debris piles are burned. In the past, rehabilitation often involved simply spreading this material across the road or landing surface. Concentrations were converted to kg/ha using bulk density values provided in the original text. b Values for mine soils to a depth of 76 cm. The degraded sites are recontoured spoils without topsoil; rehabilitated sites are recontoured spoils with topsoil. c Two-year results for the Vanderhoof site after rehabilitation with 300 kg/ha of fertilizer and seeding with a legume mix (achieved 65% cover). Two-year results for Koksilah after rehabilitation with 450 kg/ha of fertilizer and grass-legume seed mix (achieved 70% cover). TABLE 4 Potential contribution of organic matter and nutrients by adding soil amendments (the size of the effect depends on the application rate) Amendments Organic matter (kg/ha) Total N (kg/ha) Source Shredded bark (5 cm = 75 dry t/ha) a Saini and Hughes 1973 Sewage sludge (5 cm = 55 dry t/ha) b Simpson 1985 Wood waste/sewage sludge compost (5 cm = 75 dry t/ha) b Simpson 1985 a Shredded tree bark used to ameliorate potato fields in New Brunswick; an arbitrary depth of 5 cm is presented as an example; the conversion to weight assumes 150 kg/m 3 bulk density (data on bulk density were not provided). In the study, the bark was applied at 30 t/ha. b Sewage sludges and compost used in a trial of nursery potting mixes; sewage sludge is from the Greater Vancouver Regional District; assumes bulk density of 100 kg/m 3 ; compost is average value for two composts derived from sewage sludge (Kelowna and the Greater Vancouver Regional District) and wood waste; assumed bulk density for compost was 150 kg/m 3. 11

19 removing forest floor or fertilizing on four-year height growth of eastern white pine and white spruce. As a result of vegetation control, foliar nitrogen levels increased, even though the site s nitrogen capital was depleted by 23% (577 kg/ha) compared to control plots. Removal of forest floor layers from the sandy loam Podzolic soil resulted in an initially large nitrogen loss, and reduced nitrate levels in surface mineral soil after four years, but microbial biomass carbon and nitrogen, foliar nitrogen levels, and tree growth were not significantly affected (Ohtonen et al. 1992). The long-term effects of nitrogen loss were not described. Forest soil nutrient cycles are strongly influenced by the presence of plant roots and the associated bacteria and fungi that colonize the rhizosphere. Mycorrhizae are a major factor affecting nutrient uptake by forest trees (Read 1991). Mycorrhizal colonization increases the nutrient-absorbing surface, and in some cases mycorrhizae may gain access to organically bound nutrients that are otherwise unavailable to plant roots. Perry et al. (1987) described some factors influencing the formation of mycorrhizae on seedlings, and the subsequent effects on tree performance. Following disturbance, the number of mycorrhizae formed might be controlled by the balance of propagule input to mortality, the rate of recovery of host plants, and the diversity of fungal species present. On severely disturbed sites, especially where topsoil is lost, reduced levels of inocula are probably present. If unfavourable soil conditions slow the establishment of hosts on these sites, the rate of mycorrhizal development would be limited, possibly establishing a cycle of reduced inocula and poor seedling performance, which could make reforestation difficult. Seedling performance is likely affected both by the numbers of mycorrhizae and by their diversity. Kranabetter et al. (1996) evaluated the survival, early growth, and mycorrhizal colonization of birch seedlings planted in rehabilitated roads in the central interior of British Columbia. After two growing seasons, seedling survival and height growth was not significantly different for seedlings inoculated at time of planting with fresh soil from a nearby undisturbed area compared to seedling inoculated with sterilized soil. Soil conditions remained poor on the rehabilitated site two years after a treatment that involved respreading a 30-cm layer of sidecast topsoil and subsoil onto an untreated roadbed. Frost heaving was the primary cause of seedling mortality. Even though mycorrhizal diversity was low for both treatments after two growing seasons, it was significantly greater for seedlings inoculated with fresh soil compared to those inoculated with sterilized soil, which indicates that some biological function was restored Methods for evaluating nutrient pools and nutrient cycles Techniques for evaluating site nutrient pools are well established. Total nutrient concentrations of plant material and forest floors can be determined by routine procedures that involve dissolving the sample in acid, followed by analysis of the solution for nutrient elements. To determine nutrient pools, these concentrations are combined with measures of plant biomass or forest floor mass, which are calculated on a unit area basis. Field-based techniques for measuring mass of forest floor, plant material, and mineral soils are well established, although good measurements require care and patience. For mineral soils, nutrient content can be determined by combustion for elements associated primarily with soil organic matter, such as carbon, nitrogen, and sulphur. For metals such as calcium, magnesium, and potassium, and for phosphorus and iron, all-purpose extractants are suitable (Ballard and Carter 1984). Combinations of these techniques were used in many studies to evaluate changes in site nutrient pools. A more difficult problem emerges when measures of plant-available nutrients, and interpretations of their effects on tree nutrition and growth are required. In general, soil analysis has proven of limited use for predicting tree growth on undisturbed soils, and foliar analysis is much more reliable for evaluating fertilizer needs. One exception is the use of soil mineralizable nitrogen, determined by aerobic or anaerobic incubation, which has been used to evaluate site productivity with considerable success in British Columbia and elsewhere (Powers 1980; Klinka et al. 1994). The difficulties associated with using measures of extractable nutrients to explain the effect of soil disturbance on nutrient cycles may relate to temporal changes in rhizosphere nutrient cycles. Gosz and 12

20 Fisher (1984) discuss a body of evidence that indicates that plant roots exert considerable control over microbial activity and nutrient availability within the rhizosphere. This occurs mostly through root effects on the availability and quality of organic substrates, through the inhibiting effect of moisture and nutrient uptake, and through the production of allelopathic substances. These authors suggest that tightly coupled forest nutrient cycles develop in response to the episodic growth of fine roots and mycorrhizae, which results in several peaks in root biomass during a year, followed by periods with increased mortality. Tree roots absorb more water and nutrients during periods of root growth, and release organic substrates during periods of mortality. Microbial activity is inhibited during periods of root growth, and enhanced during periods of root mortality. Therefore, periods of root demand may coincide with microbial release of nutrients, and periods of root supply of substrates may coincide with periods of microbial demand. Other measures of nutrient cycling and microbial activity were used by Parmelee et al. (1993), who observed that microbial growth rates, biomass carbon and nitrogen, and faunal populations all increased as the density of pitch pine roots increased, and that most of the activity was in rhizosphere soil. Faunal abundance approached zero for non-rhizosphere soil. Measurements of mycorrhizal colonization of seedling roots and other biological attributes may help to identify and characterize the stock types used for rehabilitating sites. Recently, soil scientists from many disciplines including forestry have tried to define soil quality, or the soil attributes that provide benefits to society (Warkentin 1995). Attempts are under way to define the minimum data sets required to determine whether a particular agricultural or forest soil is in a healthy or a degraded condition (Doran and Parkin 1994). Some of the measures proposed include bulk density, organic matter content, mineralizable nitrogen, and microbial biomass carbon and nitrogen. These attempts to quantify soil attributes associated with quality may guide efforts aimed at quantifying the effects of soil rehabilitation on nutrient cycles and soil productivity. 5 TECHNIQUES FOR FOREST SOIL REHABILITATION In general, soil rehabilitation aims to restore soil physical, chemical, and biological conditions to a state where the site s productivity will mirror that of undisturbed soils on similar site types in the area. This section describes current or potential techniques to restore productivity to soils in various ecosystems throughout British Columbia. 5.1 Tillage Tillage involves the mechanical manipulation of soil to improve soil conditions for plant growth (Hillel 1980). When rehabilitating forest soil, the primary goal of tillage is to alleviate compaction, thereby reducing the resistance of soil to root penetration, and providing improved drainage and aeration to enhance seedling establishment and tree growth. The secondary goal may involve creating a seedbed for cover crops, or mixing and incorporating organic amendments. Many implements are available for tilling forest soils. The most appropriate implement for a particular area depends on: site conditions (e.g., slope, soil texture, moisture content, organic matter content, presence of unfavourable subsoil); tillage objectives (e.g., depth of tilling based on a knowledge of natural rooting depth in undisturbed soils, amount of mixing); and cost and availability of equipment. The success of a tillage operation depends on ecologically appropriate objectives, suitable equipment, and a conscientious and knowledgeable equipment operator. Therefore, no single prescription or implement is ideal for all sites. While the tillage techniques used to rehabilitate forest soils may be similar to some used in agriculture, forestry tillage is unique in several respects. Because forest crops have long rotations, tilling forest soil may occur only once every years on a 13

21 particular site. In most situations, tillage must be completed before the trees are planted. Prescriptions must rapidly create suitable growing conditions for tree seedlings, while also restoring growing space that the roots of mature trees will exploit 10 or more years later. Although multiple tillage operations or combinations of different implements and machines are possible, many degraded forest soils are located in remote areas. Substantial costs are involved in transporting equipment and materials to such sites. Furthermore, degraded forest soils may occur on steep slopes, they vary more than agricultural soils, and they may contain large rocks or buried wood, which hamper tillage operations. Soil texture and moisture content influence tillage results. Effective tillage requires that the soil is strong enough to transfer the energy of the implement through a substantial portion of the soil profile, creating numerous fracture planes. Sandy soils derive their strength largely from internal friction between grains; strength is not highly dependent on water content. In contrast, cohesion forces play an important role in binding clay soil particles together. As soil water content is increased in clay soils, moisture films weaken inter-particle bonds and reduce internal friction, thereby reducing soil strength (Hillel 1984). Soil strength, therefore, is influenced by moisture content (Greacen and Sands 1980), as well as other soil properties such as clay content and mineralogy (Barzegar et al. 1995). Standardized tests describe how these effects relate to the engineering properties of soils (McBride 1993). These tests also help to explain how individual forest soils are affected by tillage under changing moisture conditions (McNabb 1994). Soil organic matter alters the short-term response of soils to tillage, but, over the long term, organic matter stabilizes soil structure. The persistence of macropores (fractures) created by tillage depends on the stability of the associated clods, which are small remnants of the compacted soil matrix. Soil organic matter plays a key role in stabilizing clods and allows them to transform into soil aggregates that resist dispersion when wetted. Our understanding of the effect of different tillage depths on forest productivity is drawn largely from results in the United States (Andrus and Froehlich 1983; McNabb and Hobbs 1989), where soil and climatic conditions differ from those in many parts of British Columbia. For example, many undisturbed soils in Oregon are derived from volcanic materials and have low bulk densities (< 1000 kg/m 3 ) from the surface to a depth of up to 1 m (Froehlich and McNabb 1984). Although rainfall patterns throughout the region vary tremendously depending on site location relative to the major mountain ranges, climates are generally mild, and growing season soil temperatures encourage deep rooting of forest trees. While these climatic conditions are similar to south coastal areas of British Columbia, many areas of the province have lower temperatures and soils derived from glacial materials (Lavkulich and Valentine 1978). Effective rooting depths (the depth above which 90% of the fine roots are located) of only 30 cm may occur in glacial-derived soils in the interior because of naturally dense subsoils and low soil temperatures. Therefore, tillage of deeper soil layers where rooting is restricted by low temperatures is unnecessary unless other factors (e.g., the need for drainage) are involved. Two broad strategies for tillage include those that treat the entire degraded area to a specified depth and those that attempt to ameliorate only a portion of the area, perhaps with the intention of creating a planting spot for each tree seedling Extensive tillage Extensive tillage, as the term is used here, refers to techniques that till the entire disturbed area to a specified depth. Various implements are used, often pulled by, or attached to, crawler tractors. Excavators with site preparation rakes and other attachments are also used to extensively till degraded forest soils. In general, productivity rates for extensive tillage are higher and costs are lower for implements that move in one direction and complete the operation in a single pass. Stopping, reversing direction, and turning all reduce productivity. Winged subsoiler The use of winged subsoilers in forestry tillage was tested in Oregon before These implements are either mounted directly to the back of a crawler tractor or are pulled behind a crawler. They consist of two or three vertical shanks with a single wing or shoe attached to the bottom of each shank (Figure 2). As the implement is pulled through the soil at depth, the wings lift the soil, creating forces that exceed the soil shear strength and result in fracture. Several years of experimentation and development on winged subsoilers in the early 1980s resulted in a patented design for a self-drafting implement 14

22 FIGURE 2 Winged subsoiler tilling a landing at the Aleza Lake Research Forest, Prince George Forest District. The implement illustrated here was designed by Tilth Inc. of Monroe, Oregon, but the attachment to the crawler is not standard. manufactured by Tilth Inc. of Monroe, Oregon. This implement, which represents the most advanced subsoiler design currently available, is mounted directly to a crawler tractor, or mounted to a dolly that is pulled behind the crawler. This winged subsoiler has several advantages over other subsoilers lacking the self-drafting design. First, the self-drafting feature allows the implement to remain at constant depth even as the crawler climbs over mounds, logs, and other irregularities in the soil surface. Also, a constant amount of force is applied to lift the soil profile regardless of depth, so that fracture patterns and clod size are more consistent. It also incorporates a tripping mechanism, which allows the individual shanks to automatically release, ride over buried wood or rocks, and then return to position for continued tilling. A selection of wings is available for use in different soil types. Generally, wider wings are preferred for deeper tillage and for fine-textured soils. Several self-drafting subsoilers are available for use in British Columbia. Winged subsoilers have proven more effective than other implements for tilling forest soils (Andrus and Froehlich 1983). For example, in a single-pass operation, a winged subsoiler loosened 70% of the compacted volume of a clay loam soil, and 80 90% of a loam soil. The proportion of compacted soil loosened by rock rippers and brush blades was much lower (maximum 45%), and a disk harrow loosened only 20% of the compacted soil volume. The main advantage of the self-drafting winged subsoiler is its ability to maintain a lifting and loosening action when run at depth. Conventional rock rippers and brush blades produce a desirable soil condition when used for shallow tillage, but compress the soil and create well-defined slots when run below a critical depth especially in moist finetextured soils. Therefore, while the critical depth varies for different soil conditions and implements, it is generally near 15 cm, too shallow to effectively rehabilitate forest sites in Oregon, for instance (McNabb and Hobbs 1989). The disk harrows used were unable to penetrate the compacted soils below approximately 10 cm. Winged subsoilers have effectively loosened landings (Carr 1988b), skid roads (Andrus and Froehlich 1983), compacted areas resulting from brush piling (Davis 1990), and naturally compact forest soils (DeLong et al. 1990). 15

23 While the research from Oregon suggests that the winged subsoiler is effective over a range of soil texture classes, the most consistent results in British Columbia have occurred on coarse-textured soils. McNabb (1994) described the winged subsoiler as only moderately effective for loosening fine-textured soils in west-central Alberta. Limited shatter was observed in the surface layers, primarily because the clay and clay loam subsoils were near field capacity at the time of tillage, and had low strength, even though the tillage was carried out after an extended period of dry weather. In contrast to the moderate levels of compaction observed for degraded soils in Oregon, McNabb (1994) showed that bulk density values in the surface horizons of the Alberta soils were very near their maximum attainable values. The tillage operation did not fracture the very strong surface horizons because the weak subsoils did not transmit the force of the subsoiler. Rock ripper Despite the results of Andrus and Froehlich (1983), which showed that winged subsoilers were favoured for single-pass operations, conventional rock rippers have been used successfully for landing rehabilitation in British Columbia. Rock rippers are tooth-like attachments that are directly mounted to the rear of a crawler and are lowered into the soil as the machine moves forward (Figure 3). Preliminary observations made on several sites in the Kamloops Forest Region (Figure 4) indicate that under appropriate soil conditions (e.g., coarse texture with moderate gravel content) and with careful operation, satisfactory results are obtained using rock rippers. A single-pass tillage operation with a rock ripper may not till the entire soil surface layer (Figure 5). Several passes may be necessary to ensure that the entire surface layer is loosened. However, the full range of sites where rock rippers give satisfactory results is not known, and information on the relative costs of rock rippers compared to other implements is not available. McNabb and Hobbs (1989) found that shallow tillage (20 cm) with a rock ripper did not improve the growth of ponderosa pine seedlings in compact, fine loamy soils on a hot, dry site in southwest Oregon. The techniques reduced the bulk density over only FIGURE 3 Rock ripper being used for forest soil tillage in the Penticton Forest District. The ready availability of rock rippers to logging operations in the province has encouraged their use in soil rehabilitation. 16

24 FIGURE 4 Excellent growth of lodgepole pine planted in 1990 on a landing tilled with a rock ripper (Kamloops Forest District, Jamieson Creek Forestry Road, km 33). FIGURE 5 Results of a single pass with a rock ripper on a medium-textured landing in the 100 Mile House Forest District. Approximately 30% of the surface layer was tilled. 17

25 10% of the area because the rippers were widely spaced and penetrated to a depth of only 20 cm. After five years, seedlings planted in the middle of the furrow created by the ripper were no larger than seedlings planted in the compact soil between the furrows. The authors concluded that the shallow cultivation depth was not sufficient to provide any significant advantage to the seedlings. Brush blade In the Cariboo Forest Region, a landing rehabilitation project described by Vyse and Mitchell (1977) used a brush blade to till to a depth of 12 cm. The authors observations of soil physical properties are similar to those provided by Andrus and Froehlich (1983), indicating little effect below that depth. At least two sites were visited during the summer of 1996 that were tilled with brush blades. Others were visited where the tillage method was not documented, but where conditions resembled those expected when using brush blades. In the Nakusp area, a landing on coarse-textured soil, which was rehabilitated with a brush blade in 1988, was observed with good stocking and growth of lodgepole pine. The use of a brush blade is limited not only by effectiveness, but also by the expected cost using this implement. For the operator to avoid travelling over the tilled surface, the machine must constantly reverse direction to till a small area between the front of the tracks and the blade. Brush blades are effective for spreading topsoil piles, however, and will likely find use in specific rehabilitation situations. Excavator Extensive tillage is also carried out with an excavator. Although the costs per hectare tilled are higher than those for winged subsoilers (Table 5), their use is justified in many situations because of: their versatility when equipped with various attachments; their low ground pressure; and their ability to gain access to abandoned roads while causing minimal soil disturbance. Excavators (Figure 6) have rehabilitated landings (Lawrie et al. 1996) and roads (Hickling et al. 1996), and although little is known of the soil physical conditions that resulted, depending on the prescription and the amount of money available for the work, these machines are capable of thoroughly loosening soil to any reasonable depth. Excavators are also able to replace topsoil, move debris, incorporate amendments, and break up large clods to prepare seedbeds. Compared to a winged subsoiler, more mixing occurs in surface soil layers that are tilled with excavators. A single ripper-tooth mounted on an excavator was used to till 50 landings during the summer of 1995 as part of an Forest Renewal BC funded project carried out by Pope and Talbot (Nakusp). Based on observations of several coarse-textured landings TABLE 5 Cost of various implements for extensive tillage of forest soils; all figures adjusted to 1996 costs ($140 per hour) using FERIC s estimates of machine-ownership costs (Lawrie et al. 1996) Productivity Cost Machine type Prescription (hr/1000 m 2 ) ($/1000 m 2 ) Source Winged subsoiler Deep tillage 85 Marsland 1994 a Winged subsoiler Deep tillage Walker and Horley 1991 b Winged subsoiler Deep tillage Lawrie et al c Excavator Shallow till Lawrie et al c Excavator Till + topsoil Lawrie et al c Rock ripper Deep tillage Ellen 1996 d Winged subsoiler Deep tillage Fewer 1992 e a 1993 program: costs in 1991 ($42) and 1992 ($54) were much lower, but concern was expressed about the quality of the work, so the 1993 program was carried out at lower machine speeds. b Total costs for the operation. c These costs represent tillage time only, not including delays. d Based on Forest Renewal BC cost estimate for 1994 work, p.17. e Average costs for landings and trails combined in 1991 (Kalum District), including repairs and delays. Costs declined in 1992 to $87 /1000 m 2. 18

26 FIGURE 6 Extensive tillage with an excavator equipped with a site preparation rake (Williams Lake Forest District). Excavators are often favoured for rehabilitation work because of their suitability for additional tasks such as culvert installation and removal, topsoil retrieval, and sidecast pullback. during the summer of 1996, the results appeared satisfactory. A similar implement (Figures 7 and 8) with wings attached to the ripper-tooth, was in use in the south Okanagan in Spot treatments and site preparation equipment Spot treatments are applied on their own or as part of treatment combinations. In British Columbia, applying spot treatments to rehabilitate soil has received limited use, although a pilot project carried out in the Cariboo Forest Region during 1996 used some of the treatments described in this section. Treatments that create individual planting and growing spots have potential advantages over extensive tillage methods. First, if only a portion of the total compacted soil volume is rehabilitated, energy requirements are reduced, and machines (e.g., excavators) that are expensive when applied to entire areas may become practical to rehabilitate a wider range of sites. Second, these types of treatments have a greater potential to influence surface relief by creating mounds or trenches. This may benefit sites where excess moisture, low temperatures, or drought adversely affect seedling establishment, or backlog sites where protecting the existing trees is desired. Throughout British Columbia, various implements are available to prepare spots, mounds, trenches, and other features, thereby addressing a range of growthlimiting conditions. The potential advantages of spot treatments are only realized, however, if the entire site s long-term productivity is returned to a level similar to other sites in the area. Because spot treatments do not till the entire degraded soil volume, growing roots will eventually encounter untreated soil. For most sites, this would probably reduce overall site productivity, but two possible mechanisms exist to maintain productivity. Spot treatments are successful when: 1. tree roots can penetrate the untreated soil when encountered; or 2. aboveground site productivity is unaffected by the roots inability to exploit the untreated soil. 19

27 FIGURE 7 Excavator with a single-tooth ripper attachment, used for loosening very rocky soils in the Penticton Forest District. FIGURE 8 Close-up of the ripper attachment illustrated in Figure 7, showing the wing attachments to enhance soil tillage (Penticton Forest District). 20

28 The ability of tree roots to proliferate through untreated soil depends on the extent to which soil conditions improve during the period after tree planting. While most authors agree that heavily compacted soils take several decades to recover (Wert and Thomas 1981; Froehlich et al. 1985), more rapid recovery is reported on some sites (Corns 1988; Reisinger et al. 1992). For site productivity to be unaffected by the loss of a significant amount of soil volume (e.g., case 2 above), tree roots occupying a restricted volume of soil must be capable of acquiring all the water and nutrients required. The most commonly used machine for spot treatment in British Columbia is an excavator, which can be equipped with various attachments, including site preparation rakes, thumbs, and spot-mixing heads (Figure 9). Mounding is commonly used for site preparation in the province, and is used to rehabilitate degraded soils as well. Mounding is especially suitable for winter landings, or rutted areas that occur on wet sites. Some examples of mounding were observed during visits to sites in the summer of 1996, but no information was available about the effects on productivity. In general, for spot treatments based on site preparation techniques, rehabilitation objectives should be based on a knowledge of the growth-limiting conditions and the effect of the treatment on those conditions. The effect of site preparation techniques on soil conditions affecting seedling growth was described by Orlander et al. (1990). Limited rooting volume resulting from spot treatment would show its effects at a young age for trees planted in small treated spots. In 1979, the Canadian Forest Service established a trial that evaluated auger planting as a method to break up compacted soils and improve seedling survival and growth on landings in north-central British Columbia. Seedlings were planted using a mattock, dibble, or chainsaw-powered auger, which created cm planting holes (Arnott et al. 1988). Bareroot lodgepole pine planted with the auger survived better than trees planted with a mattock, but auger planting had no FIGURE 9 Results of partial tillage with an excavator equipped with a silvatiller attachment (Horsefly Forest District). The silvatiller mixed the surface soil layers to a depth of approximately 30 cm. The darker strips were tilled, which loosened the surface soil and disturbed the grass sod. Untilled areas between the strips have a well-developed grass sod and more compact soils. 21