Short-term impact of forest soil compaction and organic matter removal on soil mesofauna density and oribatid mite diversity

Transcription

1 1136 Short-term impact of forest soil compaction and organic matter removal on soil mesofauna density and oribatid mite diversity Jeffrey P. Battigelli, John R. Spence, David W. Langor, and Shannon M. Berch Abstract: This study examines the short-term impact of forest soil compaction and organic matter removal on soil mesofauna, in general, and oribatid mite species, in particular. Both soil compaction and organic matter removal reduced the density of soil mesofauna. Stem-only harvesting reduced total mesofauna densities by 20% relative to uncut forest values. A combination of whole-tree harvest and forest floor removal with heavy soil compaction significantly reduced total soil mesofauna densities by 93% relative to the uncut forest control. Removal of the forest floor represents a substantial loss of habitat for most soil mesofauna. The forest floor apparently buffered the mineral soil by limiting both the impact of soil compaction and fluctuations in soil temperature and moisture. The relative abundance of Prostigmata and Mesostigmata increased with treatment severity, whereas that of Oribatida decreased. Species richness of the oribatid mite fauna was reduced as the severity of treatments increased. The number of rare oribatid species (those representing <1% of the total oribatid mite sample) decreased by 40% or more relative to the uncut forest control. Evenness also decreased as treatment severity increased. Oppiella nova and Suctobelbella sp. near acutidens were the dominant oribatid species in both the forest floor and mineral soil, regardless of treatment. Soil compaction and organic matter removal significantly impacted the density and diversity of soil mesofauna and oribatid mite fauna in the short term at these study sites. Résumé : Nous avons étudié l impact à court terme de la compaction des sols forestiers et du retrait de la matière organique sur la mésofaune du sol en général, et en particulier sur les espèces d acariens oribates. La compaction du sol et le retrait de la matière organique ont tous deux réduit la densité de la mésofaune du sol. La récolte par troncs entiers a réduit la mésofaune totale de 20 % en comparaison de la forêt intacte. La récolte par arbres entiers avec décapage de la couverture morte combinée à la compaction sévère des sols a réduit la mésofaune totale de 93 % par rapport au témoin non coupé. Le décapage de la couverture morte représente une perte substantielle d habitat pour la plupart des espèces de la mésofaune. La couverture morte semble jouer un rôle tampon en limitant à la fois la compaction et les fluctuations de température et d humidité dans le sol minéral. L abondance relative des Prostigmata et des Mesostigmata a augmenté avec la sévérité du traitement tandis que celle des Oribatida a diminué. La richesse spécifique de la faune acarienne oribate a diminué en relation avec l augmentation de la sévérité des traitements. Le nombre d espèces rares d oribates (ceux représentant <1 % de l échantillon total d acariens oribates) a diminué de 40 % ou plus en comparaison du témoin non coupé. L équitabilité a aussi diminué en relation avec l augmentation de la sévérité du traitement. Oppiella nova et Suctobelbella sp. near acutidens étaient les espèces dominantes d acariens oribates dans la couverture morte et dans le sol minéral quel que soit le traitement. La compaction du sol et le décapage de la matière organique ont eu un impact significatif à court terme sur la densité et la diversité de la mésofaune du sol et de la faune acarienne oribate dans ces sites d étude. [Traduit par la Rédaction] Battigelli et al Introduction Soil fauna are integral to humus formation (Klinka et al. 1981) and the release of important plant nutrients through mineralization of organic matter (Seastedt 1984; Setälä 1995). Knowledge of the diversity and distribution of soil organisms as well as an understanding their varied roles in soil processes might be applied to maintain high nutrient cycling and good plant growth within well-managed forest ecosystems (Marshall 1993). Furthermore, changes to faunal Received 18 March Accepted 31 October Published on the NRC Research Press Web site at on 27 May J.P. Battigelli 1,2 and J.R. Spence. 3 Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2E9, Canada. D.W. Langor. Canadian Forest Service, Northern Forestry Centre, St., Edmonton, AB T6H 3S5, Canada. S.M. Berch. British Columbia Ministry of Forests, Research Branch Laboratory, PO Box 9536, Victoria, BC V8W 9C4, Canada. 1 Corresponding author ( jeff@earthworksresearch.com). 2 Present address: Earthworks Research Group, 10 Naples Way, St. Albert, AB T8N 7E8, Canada. 3 Present address: Department of Renewable Resources, University of Alberta, Edmonton, AB T6G 2H1, Canada. Can. J. For. Res. 34: (2004) doi: /X03-267

2 Battigelli et al assemblages may be more readily detectable than changes in physical or chemical properties, making soil fauna useful indicators of site degradation or recovery processes (Garay and Nataf 1982). Development of silvicultural practices that would satisfy forestry objectives and at the same time favourably influence soil fauna to enhance soil fertility (Marshall 1989) should be of primary importance. In 1991, the British Columbia Ministry of Forests established the Long-Term Soil Productivity study (LTSPS) (Powers et al. 1990; Holcomb 1996) in conjunction with a similar research program widely employed in the United States (see Powers 1989). The LTSPS addresses effects of various levels of soil compaction and organic matter removal on long-term soil productivity in the Sub-Boreal Spruce (SBS) biogeoclimatic zone in British Columbia over a full rotation period ( years on SBS sites). Silvicultural practices, such as various timber-harvesting methods and site preparation activities, can modify a variety of physical and chemical properties in the soil, consequently affecting soil pore space, composition and amount of organic matter, forest floor and mineral soil temperatures, and soil moisture. Alteration of these properties, in turn, can adversely affect soil fauna density and diversity (Hill et al. 1975; Cancela da Fonseca 1990; Marshall 1993) as well as alter their living space and food supply (Shaw et al. 1991). Two features of the soil ecosystem, organic matter and soil porosity, can be related to all alterable soil properties that influence soil productivity (Powers et al. 1990). Organic matter supplies nutrients in soil, influences soil structure (Banerjee and Sanyal 1991), and provides living space for more than 80% of soil biota in the forest soil ecosystem (Wallwork 1970; Price 1975). Changes in the quality and quantity of organic matter input can directly influence biological activity. Soil porosity, determined by soil structure and texture, directly influences soil physical properties, such as aeration, water storage, infiltration, and flow (Childs et al. 1989), which can be adversely affected by soil compaction. Soil compaction also reduces soil pore space, which impedes the movement of soil organisms, since most do not actively burrow in the soil, but utilize existing channels and openings in the soil to move around (Wallwork 1970). Numerous studies have explored, separately, the impact of various harvesting practices and soil compaction on soil fauna communities. Harvesting methods such as clearcutting (Huhta et al. 1967, 1969; Vlug and Borden 1973; Blair and Crossley 1988), whole-tree and stem-only harvesting (Bird and Chatarpaul 1986, 1988), and selective logging (Hoekstra et al. 1995) reduce soil fauna densities and change community structure. Likewise, soil compaction caused by human trampling (Garay and Nataf 1982; Borcard and Matthey 1995) and animal (Vtorov 1993) and mechanical activity (Smeltzer et al. 1986; Kevan et al. 1995) also reduces density and diversity of soil fauna. To date, there has been no single study that has addressed the combined impact of organic matter removal and soil compaction on soil fauna. The LTSPS experiment provided an opportunity to examine the impact of organic matter removal and soil compaction on soil fauna. We report here on the short-term impact of soil compaction and organic matter removal on density and relative abundance of soil mesofauna and diversity of oribatid mite species. This will establish a baseline for monitoring long-term recovery of the soil fauna community throughout a full rotation period. Materials and methods Study site This study was conducted within the Sub-Boreal Spruce (SBS) biogeoclimatic zone in the central interior of British Columbia. The SBS has a continental climate with snowy, cold winters and short, warm, moist summers. Mean annual temperature ranges from 1.7 to 5 C, with temperatures below 0 C for 4 5 months of the year and above 10 C for 2 5 months. Mean annual precipitation ranges from 415 to 1650 mm, with snow accounting for approximately 25% 50% (Meidinger et al ). Climax tree species in the SBS are hybrid white spruce (Picea engelmannii glauca (Moench) Voss) and subalpine fir (Abies lasiocarpa (Hook.) Nutt.). Lodgepole pine (Pinus contorta Dougl. ex Loud. var. latifolia Engelm.), Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), trembling aspen (Populus tremuloides Michx.), and paper birch (Betula papyrifera Marsh.) are seral species common in maturing climax forests (Meidinger et al. 1991). The SBS is most commonly found on the rolling mountainous and plateau landscapes of the central interior of British Columbia with Luvisolic, Podzolic, or Brunisolic soils (Soil Classification Working Group 1987) developed on extensive and often deep deposits of coarse- to fine-loamytextured glacial till. Each site in the present study has deep, medium-textured soils, derived from morainal blankets, with average soil moisture and nutrients for the subzone (Banner et al. 1993; DeLong et al. 1993; Steen and Coupe 1997). Three sites, each located in one of three different subzones to cover the range of climatic conditions within the SBS, were established as follows: (i) Log Lake: near Prince George, B.C. (54 21 N, W); in the SBSwk subzone with wet and cool climate. (ii) Topley: near Topley, B.C. (54 37 N, W); in the SBSmc subzone with moist and cold climate. (iii) Skulow Lake: near Williams Lake, B.C. (52 20 N, W); in the SBSdw subzone with dry and warm climate. Plot selection, layout, and treatment application Between 1991 and 1993, nine treatment plots (40 m 70 m) were established at each site. Three 3.0 m 3.0m plots were also located at each site in an unharvested area (>1 ha in size) set aside as an uncut forest control (OM 0 C 0 ) adjacent to the treatment plots. The LTSPS treatments were a factorial combination of each of three levels of soil compaction (none (C 0 ), light (C 1 ), and heavy (C 2 )) with each of three levels of organic matter removal (stem-only harvest (OM 1 ), whole-tree harvest (OM 2 ), and whole-tree forest floor harvest (hereafter referred to as forest floor) (OM 3 )) (Holcomb 1996). Plots were harvested in winter after a compressed snow pack had been established. Treatment application continued on each site the first summer following harvest. An excavator removed slash from whole-tree-

3 1138 Can. J. For. Res. Vol. 34, 2004 harvested plots as well as the forest floor on whole-tree forest floor removal plots. A compaction plate mounted on the arm of an excavator was used to compact the soils on each plot. A2cmdeep depression was made for each light compaction treatment and a 4 cm deep depression was achieved on heavy compaction plots. For further information on treatment application, see Holcomb (1996). Sampling, sorting, and identification of soil fauna Samples for soil mesofauna were collected from only six plots at each installation: the uncut forest control (OM 0 C 0 ) and five treatment combinations: (i) stem-only harvest, no compaction (OM 1 C 0 ); (ii) forest floor harvest, no compaction (OM 3 C 0 ); (iii) whole-tree harvest, light compaction (OM 2 C 1 ); (iv) stem-only harvest, heavy compaction (OM 1 C 2 ); (v) forest floor harvest, heavy compaction (OM 3 C 2 ). All plots at each site were sampled in the same phenological window the first year after treatment application (1994 at Topley and Log Lakes and 1995 at Skulow Lake). Timing of sample collection was based on the following: bud burst on trembling aspen (spring); soil temperature of 10 C at 10 cm (summer); and leaf colour change in trembling aspen (fall). In addition to standardizing sampling times among sites, these indicators relate to biological activity in the soil and span the range of seasonal variation in distribution and life stages for soil fauna at these sites. For each sampling date, one soil core (4.5 cm in diameter and up to 10 cm in depth, depending on organic horizon and mineral soil depth) was removed from each of three randomly selected subplots (2.5 m 2.5 m in the treatment plots and 0.5 m 0.5 m in the control plots) in each of the six plots (downed logs or large rocks were avoided). The top 3 cm from both organic and mineral soil horizons were sampled from each core. Only mineral soil samples were collected from plots that had the forest floor removed (OM 3 ). Each subplot was sampled only once during the study to limit the impact of soil removal. Within 48 h of collection, samples were placed in a highgradient extractor for 1 week (Lussenhop 1971) with dataloggers monitoring the temperature gradient during the extraction process. Mesofauna were collected into a 0.6% (w/v) picric acid solution and then transferred into glass shell vials with 70% ethanol by washing the contents of the collecting dishes with distilled water through a 50-µm sieve until no picric acid remained. All samples were initially sorted and counted under a dissecting microscope ( 16 to 40) with a fibre optic light source. Collembola were identified to family and Acari to suborder. Within the samples, several taxa other than Acari and Collembola occurred at low densities. These were pooled into one group, hereafter referred to Other Mesofauna. All mesofauna samples are stored with the British Columbia Ministry of Forests Research Branch Laboratory, P.O. Box 9536, Victoria, BC V8W 9C4. Oribatid specimens were sorted to morphospecies under a dissecting microscope, cleared with lactic acid, temporarily mounted on cavity slides, and identified to species under a compound microscope (Norton 1990). Dr. Valerie Behan- Pelletier (Agriculture and Agri-Food Canada, Eastern Cereal and Oilseed Research Centre, Ottawa, Ont.) checked species identifications. Classification of oribatid species follows Marshall et al. (1987). Numbers of individuals per taxon were recorded for each sample. Voucher specimens are deposited with Biodiversity Assessment and Evaluation Research Branch, Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Foods Canada, K.W. Neatby Building, Central Experimental Farm, Ottawa, ON K1A 0C6. Data analyses Mesofauna Density (number of individuals per sample) and relative abundance ((number of individuals per taxon/total individuals collected in the sample) 100) were used in analyses. Density values are presented as number of individuals/m 2 and relative abundances are presented as a percent of the total individuals/sample. All are pooled by season and are reported as mean ± standard error, unless stated otherwise. Only taxa representing 1% of all collected material were considered for further analysis. The number of zero scores associated with taxa representing <1% limited further analysis using these methods. Data were transformed to meet assumptions of normality before analyses; density data were log transformed (log 10 (X + 1), where X = actual count of individuals for a taxon) and relative abundance data were arcsine transformed (arcsin p, where p = relative abundance of the taxon). Variances were treated as equal in the analyses. Separate analyses were carried out on (i) all soil core data collected from each treatment plot combining the organic horizon and mineral soil data (hereafter referred to as combined); (ii) the organic horizon data (including only those treatment plots that retained some organic material (OM 0, OM 1, and OM 2 )); and (iii) the mineral soil data (consisting of only mineral soil samples collected from all treatment plots). A three-way mixed ANOVA (sites, treatments, seasons (3 6 3))wasused to compare mesofauna data. In the ANOVA procedure, test of the hypothesis that there was no difference in density or relative abundance among treatments was carried out using Type III MS site treatment as an error term. Test of the hypotheses that there were no differences among seasons and season treatment were tested using Type III MS site season treatment as an error term. If differences were found with ANOVA, then Tukey s studentized range test was used. Statistical significance was judged using the Type III sums of squares for all ANOVA procedures. Test statistics and probabilities are presented when differences are significant. All statistical tests in this study were conducted using the SAS General Linear Models (GLM) procedure (SAS Institute Inc. 1990) with α = 0.05 unless stated otherwise. Oribatid mites Two diversity indices, the exponential form of Shannon Wiener Index (N 1 ) (Hill 1973) and the reciprocal of Simpson s Index (N 2 ) (Hill 1973), as well as the average number of species/sample were determined for each sample. N 1 measures the number of very abundant species and is sensitive to changes in rare species abundance. It represents the number of equally common species required to generate the observed heterogeneity. N 2 measures the number of abundant species and is sensitive to changes in abundant

4 Battigelli et al Fig. 1. Comparison of mean density of total mesofauna among treatments. Capital letters at top of column indicate test among combined data. Lower case letters in columns indicate significant difference between organic (open) and mineral soil (solid), respectively. Different letters indicate significant difference according to Tukey s test (α = 0.05). species. Values for N 1 represent the number of equally common species that would produce the same diversity as the Shannon Wiener Index (Magurran 1988). Higher values for both indices indicate greater diversity. Units for N 2 and N 1 are number of species and thus are more easily interpreted than the Shannon Wiener Index (Hill 1973). Comparison of these diversity values among treatments followed the same procedure as for the density and relative abundance of the soil mesofauna. Mean N 1 and N 2 values ± SE are presented in a scatter plot to examine species evenness in each treatment plot (Hill 1973; Ludwig and Reynolds 1988). Points closer to the plot origin indicate lower species diversity and reduced evenness (increased dominance of a single species in the plot). Relative abundance of each oribatid mite species was plotted to examine the impact of organic matter removal and soil compaction on the species assemblage. Hågvar (1994) proposed that a lognormal distribution of relative abundance values for species was typical for nonstressed soil microarthropod populations. In stressed or disturbed environments, the relative abundance of some species will decrease while the relative abundance of a few species will increase, skewing the distribution of relative abundance values from lognormal (Hågvar 1994). Species accumulation curves for oribatid mite assemblages were calculated using rarefaction to compensate for differences in the number of individuals collected (Krebs 1989). Results Soil mesofauna With the combined data (organic horizon + mineral soil), densities of total soil mesofauna decreased significantly as treatment severity increased (F [5,10] = 11.79, p = ) (Fig. 1). While OM 1 C 0 reduced mesofauna densities by 10% 25% of control values, densities in treatment plots where the forest floor had been removed (OM 3 ) and heavy compaction had been applied (C 2 ) were significantly lower (by 80% 90%) than those of the uncut forest control and stem-only-harvested plots (Fig. 1). Density of total soil mesofauna was higher in plots that retained the forest floor, even in OM 1 C 2, than in plots where the forest floor had been removed (OM 3 ) (Fig. 1). Isotomidae (F [5,10] = 6.66, p = 0.006), total Collembola (F [5,10] = 8.18, p = 0.003), and Oribatida (F [5,10] = 10.72, p = 0.001) densities were significantly lower in OM 3 C 2 plots than in the uncut forest control, OM 1 C 0 and OM 2 C 1 plots (Table 1). Onychiurid densities were significantly lower on OM 1 C 2, OM 3 C 0, and OM 3 C 2 plots than on either the uncut forest control or OM 1 C 0 plots, whereas densities in the OM 2 C 1 plots were significantly higher than that in OM 3 C 2 plots (F [5,10] = 17.09, p = ). Mesostigmata (F [5,10] = 14.44, p = ) and total Acari (F [5,10] = 11.57, p = ) densities were lower in both the OM 3 C 0 and OM 3 C 2 plots than in either the uncut forest control or OM 1 C 0 plots, whereas densities in the OM 2 C 1 and

5 1140 Can. J. For. Res. Vol. 34, 2004 Table 1. Comparison of mean density ( 10 3 /m 2 ) ± SE of soil mesofauna from the organic horizon, mineral soil, and combined data among treatments. Control OM 1 C 0 OM 2 C 1 OM 1 C 2 OM 3 C 0 OM 3 C 2 Collembola Entomobryidae Organic 0.24±0.087b 0.71±0.204a 0.15±0.081b 0.10±0.058b Mineral 0.05± ± ± ± ± Combined 0.29± ± ± ± ± ± Hypogastruridae Organic 3.3± ± ± ±1.35 Mineral 0.8± ± ± ± ± ±0.223 Combined 4.0± ± ± ± ± ±0.223 Isotomidae Organic 20±2.7 17±3.2 16±3.2 17±6.5 Mineral 2±0.5 5±1.5 3±0.8 5±1.8 4±1.3 1±0.8 Combined 22±3.1a 22±3.4a 19±3.4a 20±7.6ab 4±1.3ab 1±0.8b Neelidae Organic 1.3±0.47a 0.24±0.154b 0.34±0.138b 0.02±0.024b Mineral 0.1± ± ± ± ±0.024 Combined 1.4± ± ± ± ± ±0.024 Onychiuridae Organic 13.7±4.48a 12.7±3.34a 5.8±1.72b 2.7±0.76b Mineral 3.0±0.86a 4.5±1.78a 1.1±0.38ab 1.1±0.36ab 2.0±0.63ab 0.3±0.19b Combined 16.7±4.66a 17.2±3.60a 6.9±1.87ab 2.7±0.81bc 2.0±0.63bc 0.3±0.19c Total Organic 40±6.0 33±5.8 24±4.5 23±6.8 Mineral 6±1.4 10±3.5 4±0.9 7±2.2 8±2.5 2±1.3 Combined 45±6.4a 44±6.2a 28±4.6a 25±8.0ab 8±2.5ab 2±1.3b Acari Oribatida Organic 62±7.2a 42±6.1ab 30±4.4ab 19±2.9b Mineral 18±6.8a 16±6.1ab 5±1.7abc 6±2.8bc 8±1.8bc 3±2.7c Combined 79±12.4a 58±10.3a 35±4.3ab 22±5.1bc 8±1.8ab 3±2.7c Prostigmata Organic 95±9.5a 76±10.1a 38±4.7b 34±5.3b Mineral 24±7.8 17±5.0 20±8.4 16±4.2 23±6.5 10±4.5 Combined 119±12.8a 93±11.7ab 58±8.4ab 33±4.4ab 23±6.5ab 10±4.5b Mesostigmata Organic 9±1.2a 7±1.1ab 4±1.0b 6±1.2b Mineral 2±0.5 2±0.5 1±0.5 1±0.6 3±1.1 2±1.1 Combined 11±1.2a 9±1.1a 5±1.1ab 4±1.2ab 3±1.1bc 2±1.1c Total Organic 166±13.0a 125±13.8ab 72±7.6bc 59±7.2c Mineral 43±11.6a 35±9.6ab 26±9.0ab 23±6.0ab 34±8.8ab 15±7.8b Combined 209±19.6a 159±18.0a 98±9.9ab 60±8.6ab 34±8.8bc 15±7.8c Other Mesofauna Organic 6.6± ± ± ±1.14 Mineral 1.9± ± ± ± ± ±0.32 Combined 8.5±1.81a 5.5±1.26a 4.7±1.12ab 4.2±1.50ab 3.6±1.35ab 0.7±0.32b Note: OM 1, stem-only harvest; OM 2, whole tree; OM 3, whole tree forest floor removal; C 0, no compaction; C 1, light compaction; C 2, heavy compaction. Values followed by a different letter are significantly different. Means are compared against each other after ANOVA using Tukey s test (α = 0.05). OM 1 C 2 plots were significantly higher than those of the OM 3 C 2 plots. Prostigmata densities were significantly lower on the OM 3 C 2 plots than on the uncut forest control (F [5,10] = 4.17, p = 0.03), whereas densities of Other Mesofauna were significantly lower on the OM 3 C 2 plots than those on either the uncut forest control or OM 1 C 0 plots (F [5,10] = 5.94, p = 0.01) (Table 1). Total mesofauna densities from the organic data differed significantly among treatment plots (F [3,6] = 13.01, p = 0.005), decreasing as the level of organic matter removal in-

6 Battigelli et al Fig. 2. Comparison of mean relative abundance (% of total) of soil mesofauna among treatments for organic and mineral data combined. Areas with different letters indicate significant difference among treatments (after Tukey s test, α = 0.05). Absence of letters indicates no significant difference. creased (Fig. 1). Densities in OM 1 C 0 plots did not differ significantly from densities in the control plots, but OM 1 C 2 and OM 2 C 1 significantly reduced soil mesofauna densities compared with the uncut forest control plots. Densities of Oribatida (F [3,6] = 10.51, p = 0.008), Mesostigmata (F [3,6] = 24.97, p = ), and total Acari (F [3,6] = 14.17, p = 0.004) followed a similar pattern as the total mesofauna values, with both total Acari and Oribatida densities being significantly higher in the stem-only-harvested plots than those in the OM 1 C 2 treatments (Table 1). Prostigmata (F [3,6] = 8.11, p = 0.016) and Onychiuridae (F [3,6] = 14.57, p = 0.003) densities were significantly higher in the control and OM 1 C 0 plots than in the OM 2 C 1 and OM 1 C 2 treatments. Neelidae (F [3,6] = 6.46, p = 0.026) densities were significantly lower in all three treatments plots than in the uncut forest control plot, whereas densities of Entomobryidae (F [3,6] = 8.29, p = 0.015) were significantly higher in the stem-only-harvested plots than in the other treatment plots and the control (Table 1). There were significant differences in total mesofauna densities in the mineral soil among the treatments (F [5,10] = 4.06, p = 0.03). However, Tukey s test results indicated that only densities in OM 3 C 2 were significantly lower than densities in the uncut forest control plot (Fig. 1). Densities did not differ significantly among the uncut forest control plots and the other four treatment plots. Densities of Onychiuridae (F [5,10] = 4.75, p = 0.017), Oribatida (F [5,10] = 7.25, p = 0.004), and total Acari (F [5,10] = 3.27, p = 0.05) differed significantly among the treatments and followed the same pattern as the total soil mesofauna densities with values in the OM 3 C 2 treatments being significantly lower than those in the uncut control. Both Oribatida and Onychiuridae densities were significantly higher in the OM 1 C 0 treatments than in the OM 3 C 2 treatments (Table 1). Overall, organic matter removal had a greater influence on densities than soil compaction, accounting for almost 60% of the variation in mesofauna density. Removal of forest floor (OM 3 plots) is key, since most soil fauna reside in this layer. Acari was the dominant taxon in each treatment plot, representing 65% 78% of the total mesofauna, whereas Collembola accounted for 18% 23%. Prostigmata was the most abundant acarine suborder, representing 42% 56% of the mesofauna, followed by Oribatida (15% 26%) and Mesostigmata (3% 8%). Isotomidae (4% 15%), Onychiuridae (1% 8%), and Hypogastruridae (1% 3%) were the most abundant collembolan families. This pattern of relative abundance is similar to that observed in the uncut forest control. The relative abundances of only two taxa, Onychiuridae and Oribatida, were significantly impacted by the treatments. Relative abundance of Onychiuridae, composed of true soil-dwelling collembolan species, was significantly lower in heavily compacted (C 2 ) plots than in plots with no

7 1142 Can. J. For. Res. Vol. 34, 2004 Table 2. List of oribatid mite species collected and identified from the Long-Term Soil Productivity study (LTSP) sites sampled during this study. Species PALAEOSOMATA Palaeacaridae Palaeacarus hysticinus Tragardh, 1932* ENARTHRONOTA Atopochthoniidae Atopochthonius artiodactylus Grandjean, 1948* Brachychthoniidae Brachychthonius bimaculatus Willmann, 1936* sp.* Liochthonius brevis (Micheal, 1888)* lapponicus (Tragardgh, 1910)* muscorum Forsslund, 1964* simplex (Forsslund, 1942)* tuxeni Forsslund, 1957* sp.* Neoliochthonius sp. near globuliferous* Sellnickochthonius immaculatus (Forsslund, 1942)* suecica Forsslund, 1942* sp. near suecia** Synchthonius crenulatus (Jacot, 1938)* elegants Forsslund, 1957* Verachthonius sp.* Pterochthoniidae Pterochthonius angelus (Berlese, 1910)* MIXONOMATA Euphthiracaridae Euphthiracarus cernuus Walker, 1965* Oribotritiidae Protoribotritia sp.* Phthiracaridae Phthiracarus longulus (C.L. Koch, 1841)* bryobius Jacot, 1931* DESMONOMATA Camisiidae Camisia lapponica (Tagadh, 1910)* spinifer (C.L. Koch, 1835)* Heminothrus longisetosus Willman, 1925* Neonothrus humicolous (Forsslund, 1955)* Platynothrus peltifer (C.L. Koch, 1893)* septentrionalis (Sellnick, 1944)* sp. near septentrionalis* Nothridae Nothrus anauniensis Canestrini et Fanzago, 1876* borussicus Sellnick, 1929* silvestris Nicolet, 1855* sp. near pluchelius* Table 2 (continued). Species Trhypochthoniidae Trhypochthonius tectorum (Berlese, 1896)* BRACHYPYLINA Astegistidae Cultroribula bicultrata (Berlese, 1905)* Cepheidae Cepheus corae Jacot, 1928* Ceratozetidae Ceratozetes cuspidatus Jacot, 1939** thienemanni Willmann, 1943** pacificus Behan-Pelleter, 1984* Dentizetes rudentiger Hammer, 1952* Neogymnobates luteus (Hammer, 1955)* Sphaerozetes sp.* Trichoribates copperminensis Hammer, 1952* Damaeidae Belba sp.* Epidamaeus articolous (Hammer, 1952)* koyukon Behan-Pelletier et Norton, 1985* sp. near coxalis* sp. near floccosus* sp. near kodiakensis* Eremaeidae Eremaeus brevitarsus (Ewing, 1917)** translamellatus Hammer, 1952* Eueremaeus sp. near marshalli* Galumnidae Pilogalumna sp. near tenuiclava* Gymnodamaeidae Gymnodamaeus sp. near taedaceus Paschoal, 1982* Hermanniellidae Hermanniella picea (C.L. Koch, 1839)* Liacaridae Liacarus bidentatus Ewing, 1918* Metrioppidae Ceratoppia quadridentata arctica Hammer, 1955* Mycobatidae Mycobates incurvatus Hammer, 1952* punctatus Hammer, 1955* Oppiidae Lauroppia sp.* Moritzoppia clavigera (Hammer, 1952)** Multioppia sp.* Oppia sp. 1* Oppiella nova (Oudemans, 1902)*** washburni (Hammer, 1952)* Quadroppia quadricarinata (Micheal, 1885)* Oribatulidae Jornadia n. sp.* Oribatula tibialis (Nicolet, 1855)* Phenopelopidae Eupelops sp. near plicatus*

8 Battigelli et al Table 2 (concluded). Species Polypterozetidae Polypterozetes sp.* Scheloribatidae Scheloribates pallidulus (C.L. Koch, 1841)* sp. 1** Suctobelbidae Suctobelba sp. 1* sp. 2* Suctobelbella sp. near acutidens** sp. near subcornigera* sp. 2** sp. 3* sp. 4* sp. 5* sp. 6* sp. 7* sp. 8* sp. 9* sp. 10* Tectocephidae Tectocepheus velatus (Micheal, 1880)* Undetermined sp. 1* Note: Total relative abundance from uncut forest control plots based on combined data: *, <1%; **, 1% 16%; ***, >16%. compaction (C 0 ) for the combined soil data (F [5,10] = 11.79, p < 0.005) (Fig. 2). Relative abundance of Oribatida was significantly lower in the forest-floor-harvested plots (OM 3 ) than in any other plot (F [5,10] = 6.63, p < 0.05) (Fig. 2). Conversely, the relative abundances of Prostigmata and Mesostigmata were slightly higher in the forest-floor-harvested plots than in the stem-only-harvested plots. Retention of the forest floor seems to have reduced the impact of soil compaction. Although the density of all taxa examined decreased, the shift in relative abundance indicated that Onychiuridae and Oribatida responded more negatively to disturbance than Prostigmata, which increased in overall relative abundance as treatment severity increased. Oribatid mites Eighty-seven species of oribatid mites belonging to 49 genera and 28 families were identified (Table 2). Average number of species/sample decreased as treatment severity increased (Fig. 3). Among treatment plots that retained organic matter, the mean number of species/sample in the OM 1 C 2 plots was significantly lower than that in the uncut forest control plots (F [3,6] = 6.22, p = 0.03). In the mineral soil, significantly fewer species/sample were collected from the OM 3 C 2 plots than from the uncut forest control and OM 1 C 0 plots (F [5,10] = 4.15, p = 0.03). With the combined data, the mean number of species/sample was significantly lower in the forest floor harvest (OM 3 ) plots and OM 1 C 2 plots than in the uncut forest control plots (F [5,10] = 27.07, p < ). There were no significant differences in N 1 or N 2 values from the organic horizon data among the treatments (data not presented). In the mineral horizon data (Fig. 4), only N 1 values differed significantly among the treatments (F [5,10] = 3.28, p = 0.05) with the OM 1 C 0 plots having a greater diversity than the OM 3 C 2 plots. In Fig. 5, N 1 and N 2 values calculated using the combined data differed significantly among treatments (F [5,10] = 11.64, p = and F [5,10] = 8.70, p = 0.002, respectively). Values for both OM 3 treatments were significantly lower than those for the other treatments. Evenness in the OM 3 plots was lower compared with the other treatment plots because of an increased dominance of a single species, Oppiella nova (Oudemans, 1902). As in the overall faunal analysis, changes to the diversity of oribatid mite species were influenced more by organic matter removal than by soil compaction. However, diversity values in treatment plots subjected to OM 3 C 2 were even lower than when each treatment was applied alone (i.e., OM 1 C 2 or OM 3 C 0 ). A substantial loss of rare oribatid mite species (those representing <1% of the total collected) changed the structure of the oribatid mite species assemblage (Fig. 6). For the combined data, 55 of 70 species (79%) collected in the uncut forest control were rare (Fig. 6a). Percentage of rare species decreased to 42% in the OM 3 C 2 plot (Fig. 6f) and to 0% in the OM 3 C 0 plot (Fig. 6e), whereas the proportion of species with a relative abundance >1% increased from 19% in the uncut forest control to 100% in the OM 3 C 0 plots. With the loss of rare oribatid species and increased dominance of O. nova, evenness gradually decreased as treatment severity increased. This can be seen in Figs. 4 and 5, with N 1 and N 2 values calculated for OM 3 C 2 plots being closer to the plot origin than the values for the other treatment plots. While the number of rare species decreased, identity of the abundant species remained, for the most part, consistent among treatment plots, and these abundant species represented an increasingly greater proportion of the oribatid assemblage. In the combined data (Figs. 6a 6f), O. nova was the dominant oribatid species in all plots followed by Suctobelbella sp. near acutidens. This pattern was similar in the mineral soil, with a reduction in the number of rare species (not illustrated). Oppiella nova again was the dominant species except in OM 1 C 2 plots, which were dominated by Ceratozetes thienemanni (Willmann, 1943), and OM 2 C 1 plots where Oppiella washburni (Hammer, 1952) was most common. By controlling for differences in overall sample size, rarefaction analysis of the oribatid mite species assemblage supports the conclusion that species have been lost. In the combined data (Fig. 7), species accumulation in the OM 3 C 2 plots was much lower than that in other treatment plots. The remaining treatment plots were characterized by similar species accumulation curves. However, species number on all treatment plots remained lower than the uncut forest control. Rarefaction analysis for the mineral soil data produced similar results (data not presented). Discussion Soil mesofauna respond to habitat disturbance in different ways. Huhta et al. (1967) documented three different responses of soil fauna communities to clear-cutting in Nor-

9 1144 Can. J. For. Res. Vol. 34, 2004 Fig. 3. Comparison of mean number of oribatid mite species/sample (±SE) among treatments. Points followed by similar letters are not significantly different (after Tukey s test, α = 0.05). way spruce forests in Finland: (i) a density increase immediately after cutting; (ii) a density increase followed by a decrease several years after cutting; and (iii) a density decrease immediately after cutting. In the present study, we found that stem-only harvesting without soil compaction slightly reduced the density of soil mesofauna immediately after timber harvesting; however, these values were not significantly different from densities in the uncut forest control. In fact, overall densities of three collembolan families, Entomobryidae, Isotomidae, and Onychiuridae, were slightly higher in stem-only-harvested plots than in uncut forest control plots. In the mineral soil alone, densities of Isotomidae and Onychiuridae as well as total Collembola, Oribatida, and Mesostigmata in stem-only-harvested plots did not differ significantly from those in the uncut forest control plots. Our results differ only slightly from other studies and suggest that soil compaction during timber harvesting can affect soil fauna. For example, 2 years after timber harvest in a temperate mixed conifer hardwood stand, both conventional (stem-only) and whole-tree harvesting reduced microarthropod communities 56% 68% (Bird and Chatarpaul 1986). In a Finnish study, densities of both Collembola and enchytraeids were higher in cut stands of Norway spruce than in uncut control stands immediately after timber harvesting, but mite (Acari) densities in cut stands were about half that of the control area (Huhta et al. 1967, 1969). We found that OM 1 C 2 decreased densities of most soil mesofauna taxa by 50% or more relative to the uncut forest control, which was similar to reductions observed by Bird and Chatarpaul (1986) and Huhta et al. (1967, 1969). This could be related to the harvesting method employed in these studies. While hand-only felling was used in the present study, there is no mention in the other papers of how harvesting was carried out. The use of machines for harvesting could account for compaction on those older studies and make their results more similar to the compactedharvested treatment plots in our study. Generally, fungi dominate most forest soils and bacteria are most prominent in agricultural soils (Wallwork 1970), and most mites and springtails in forest soils feed on fungal biomass. However, tree harvesting and soil compaction can increase bacterial diversity and biomass in forest soils (Marshall 2000), thereby altering food resources (i.e., fungi) available to soil mesofauna. Declines in oribatid diversity and density have been related to the loss or delayed development of fungi in organic matter (Huhta et al. 1967). An increase in bacterial biomass could increase the density of microbial feeding species, such as members of the oribatid mite family Suctobelbidae (Ryabinin and Pan kov 1987), altering the overall structure of a mesofaunal assemblage. In our study, relative abundance of Suctobelbidae was higher in all treatment plots (OM 1 C 0 20%, OM 3 C 0 21%, OM 2 C 1 27%, OM 1 C 2 18%) than in the uncut forest control (15%) except for the OM 3 C 2 plot (2%) (J. Battigelli, unpublished data). Loss of organic matter can also increase fluctuations in soil temperatures, light exposure at the soil surface, and evaporation rates (Huhta et al. 1967; Shaw et al. 1991). Heat

10 Battigelli et al Fig. 4. Comparison of mean N 1 and N 2 values (±SE) among treatments for oribatid mite species (mineral data only). Difference shadings (solid vs. open) indicate significant differences among N 1 values following Tukey s test (α = 0.05). sum values (the sum of daily average temperatures) calculated for six of the nine treatment plots at each site from September 20, 1996, to September 20, 1997, were higher on plots with greater compaction and more organic matter removal (Kranabetter and Chapman 1999). Although there were no significant differences in soil temperature among treatments, temperature fluctuation increased with treatment severity, with soil temperatures going below freezing in winter and above air temperature in summer (M. Kranabetter, B.C. Ministry of Forests, Northern Interior Forest Region, unpublished data). While most soil mesofauna taxa are able to tolerate some variation in soil temperature and moisture, some taxa are unable to adapt to large fluctuations in these soil properties. For example, Siepel (1996) documented a decrease in densities of oribatid species intolerant to drought after forest harvesting. Some oribatid species avoid drought conditions by moving deeper into the soil profile (Wallwork 1970; Siepel 1996.) However, increased soil bulk density due to soil compaction can limit vertical movement in the soil profile, even for smaller species such as O. nova and Suctobelbella species. In the present study, fine fraction soil bulk density was higher in the heavy compaction plots than those plots with no compaction (1.49 g/cm 3 versus 1.25 g/cm 3, p < F = 0.06, α = 0.1; M. Kranabetter, unpublished data). We found densities of mites and springtails were, in some cases, an order of magnitude lower in the heavy compaction plots compared with those plots with no compaction. Plots in which mineral soil was still covered by forest floor had higher densities of soil mesofauna than those plots where forest floor had been removed. The forest floor may act as a buffer, reducing the impact of compaction in the mineral soil and limiting the increase in bulk density. Donnelly and Shane (1986) showed that mulch applied prior to compaction reduced the degree of compaction. As density and diversity of soil mesofauna decreased in the present study, structure of the mesofaunal assemblage also changed. For example, both density and relative abundance of oribatid mites declined as treatment severity increased. While densities of Mesostigmata and Prostigmata also decreased, relative abundance of both taxa increased as treatment severity increased. Oribatid mites were more sensitive to changes in their environment and decreased in number at a greater rate than Prostigmata or Mesostigmata. Because of their rapid decline in both density and diversity in disturbed habitats, oribatid mites may be useful biological indicators of changes in the soil habitat (Behan-Pelletier 1999). In the present study, we found that tree harvesting and forest floor removal reduced oribatid mite species richness and changed the structure of the oribatid mite species assemblage. Bird and Chatarpaul (1986) also found that relative abundance of several oribatid taxa differed significantly after harvesting but observed no change in the taxa present. In contrast, Siepel (1996) concluded that forest harvesting (a low frequency disturbance) reduced the number of oribatid species more than did natural disturbances, such as cyclical wildfires, or seasonal changes. He argued that species could

11 1146 Can. J. For. Res. Vol. 34, 2004 Fig. 5. Comparison of mean N 1 and N 2 values (±SE) among treatments for oribatid mite species using combined data. Different shapes (N 1 ) and shadings (N 2 ) indicate significant differences following Tukey s test (α = 0.05). accommodate natural disturbances through the development of life-history tactics or specific physiology. More than 70% of the oribatid species collected in the uncut forest control were rare, each representing <1% each of the total oribatid fauna collected. Loss of primary habitat by harvesting whole trees and removing forest floor reduced the number of rare species by half and increased the proportion of species with a relative abundance >1% from 19% in the uncut forest control to 100% in forest floor harvest plots. The loss of rare species reduced the diversity and evenness of the oribatid assemblage as treatment severity increased. In the wake of widespread logging, these results suggest reasons for concern about the biodiversity of soil mesofauna, especially because of the fact that soil fauna are integral to humus formation (Klinka et al. 1981) and nutrient mineralization (Seastedt 1984; Setälä 1995). Overall, O. nova was the most abundant oribatid mite species in this study and numerically dominant in uncut forest control plots and in all but two treatment plots. This species is parthenogenetic and has a long life span. Coupled with high fecundity and diverse feeding habits, O. nova is able to rapidly occupy a diverse range of habitats in large numbers (Ryabinin and Pan kov 1987). Soil pore size can restrict faunal movement and influence relative abundance and species composition (Whitford 1996). Small-bodied mites such as O. nova and Quadroppia quadricarinata (Micheal, 1885) and members of the genera Scheloribates and Suctobelbella are able to use the smaller pore spaces resulting from compaction. Larger species such as Ceratozetes cuspidatus Jacot, 1939 and Nothrus borussicus Sellnick, 1929 would have difficulty moving through soils with higher bulk densities (Wallwork 1983). Thus, greater densities or proportions of species from families with taxa of small physical size such as Brachychthoniidae, Tectocepheidae, and Oppiidae can indicate a recent disturbance (Behan-Pelletier 1999). In our study, the relative abundances of both Suctobelbidae and Tectocepheidae were higher in all the treatment plots than that in the uncut control. Soil compaction studies have shown reductions in the density of soil fauna. The abundance of total fungi, bacteria, nematodes, and arthropods in a mixed northern hardwood forest in northwestern Vermont was reduced for up to 2 years after compaction (Smeltzer et al. 1986). Vehicle tracks significantly reduced densities of soil arthropods in the high Arctic tundra (Kevan et al. 1995). Even walking can reduce the diversity and density of soil fauna. In deciduous forest soils near Paris, France, for example, soil compaction by human trampling reduced density and diversity of several mesofaunal groups, including Oribatida (especially Brachychthoniidae), by more than 50% (Garay and Nataf 1982). In Sphagnum mosses, human trampling dramatically reduced the number of oribatid mite species (Borcard and Matthey 1995). Compaction by grazing cattle in European pastures also reduced oribatid density and species diversity (Siepel 1996). Recovery of the soil fauna community may occur more rapidly in compacted soils than on sites with varying levels

12 Battigelli et al Fig. 6. Comparison of relative abundance of oribatid mite species among treatments using the combined data. The following names have been abbreviated: Oppiella nova, O.nov;Suctobelbella sp. near acutidens, S.sna; Moritzoppia clavigera, M.cla. Fig. 7. Rarefaction diversity analysis of oribatid mite species using the combined organic and mineral data. of organic matter removal. Bird and Chatarpaul (1986) suggested than faunal densities take years to return to preharvest levels. In a mixed hardwood stand in southwestern North Carolina, differences in mean density and relative abundance of microarthropod groups were still apparent between clear-cut and uncut sites 8 years after harvest (Blair and Crossley 1988). However, in a mixed northern hardwood forest in northern Vermont, there were no detectable differences in fungal, bacterial, nematode, or arthropod populations 5 years after soil compaction (Smeltzer et al. 1986). The impact of compaction on earthworms was of shorter duration than that of harvesting in a hardwood forest in Missouri (Jordan et al. 1999). Furthermore, after the elimination of compaction by feral pigs in Hawaiian rain forests, soil microarthropod densities nearly doubled and biomass of soil microarthropods increased by two and half times over 7 years (Vtorov 1993). Overall, we found that soil compaction and organic matter removal significantly reduced the density and diversity of soil mesofauna in the short term. Although lower densities may not be permanent, changes to the overall structure of the soil fauna assemblage may be long lasting (Marshall 2000). For example, oribatid densities remained at less than 50% of control values 7 years after clear-cutting in Norway spruce stands in Finland (Huhta 1976). The oribatid assemblage in our study showed a substantial decrease in density and diversity, while maintaining the same dominant species on all treatment plots. The most striking aspect of the re-

13 1148 Can. J. For. Res. Vol. 34, 2004 sponse was the loss of rare oribatid species from the treatment plots. If soil mesofauna of sub-boreal forests behave similarly to other studies cited previously, we predict a continued decline in mesofauna density without significant change in the pattern of relative abundance over the first 5 years after treatment. We expect faunal densities and community structure to recover more slowly on plots where the forest floor has been removed than those where the forest floor remained. Furthermore, since oribatid mites are poor dispersers (Siepel 1996), we expect oribatid density and species diversity to remain low and be dominated by small, parthenogenetic species, such as O. nova. It is essential to establish baseline data on the density, diversity, and structure of the soil fauna assemblage against which to compare changes in recovery of the soil fauna (Behan-Pelletier 1999; Linden et al. 1994). The long-term nature of the LTSPS provides an opportunity to monitor the recovery of the soil fauna assemblage in the SBS over the full timber rotation period of years. This biological data can be integrated with other monitored soil properties, both physical and chemical, to measure the overall response of the forest soil ecosystem to disturbance. Acknowledgements Canada British Columbia Partnership Agreement on Forest Resource Development (FRDA II) and Forest Renewal British Columbia (FRBC) grants provided funding for this study. Thanks to C. Woon, M. Lavigne, R. Trowbridge, A. Macadam, M. Kranabetter, W. Chapman, P. Sanborn, and D. Dunn for their assistance during the research. Thanks also to the two anonymous reviewers for their comments. Dr. V. Behan-Pelletier (AAFC) assisted with identifying the oribatid mites. References Banerjee, S., and Sanyal, A.K Oribatid mites as bioindicator of soil organic matter. In Advances in management and conservation of soil fauna. Edited by G.K. Veeresh, D. Rajagopal, and C.A. Viraktamath. Oxford and IBH Publishing Co., New Delhi. pp Banner, A., MacKenzie, W., Haeussler, S., Thomson, S. Pojar, J., and Trowbridge, R A field guide to site identification and interpretation for the Prince Rupert Forest Region. British Columbia Ministry of Forests, Victoria, B.C. Land Manage. Handb. 26. Behan-Pelletier, V.M Oribatid mite biodiversity in agroecosystems, role for bioindication. Agric. Ecosyst. Environ. 74: Bird, B.A., and Chatarpaul, L Effect of whole-tree and conventional forest harvest on soil microarthropods. Can. J. Zool. 64: Bird, B.A., and Chatarpaul, L Effect of forest harvest on decomposition and colonization of maple leaf litter by soil microarthropods. Can. J. Soil Sci. 68: Blair, J.M., and Crossley, D.A., Jr Litter decomposition, nitrogen dynamics and litter microarthropods on a southern Appalachian hardwood forest 8 years following clearcutting. J. Appl. Ecol. 25: Borcard, D., and Matthey, W Effect of a controlled trampling of Sphagnum mosses on their oribatid mite assemblage (Acari, Oribatei). Pedobiologia, 39: Cancela da Fonseca, J.P Forest management, impact on soil microarthropods and soil microorganisms. Rev. Ecol. Biol. Sol. 27(3): Childs, S.W., Shade, S.P., Miles, D.W.R., Shepard, E., and Froehlich, H.A Soil physical properties: importance to long-term forest productivity. In Maintaining the long-term productivity of Pacific Northwest forest ecosystems. Edited by D.A. Perry, R. Meurisse, B. Thomas, R. Miller, J. Boyle, J. Means, C.R. Perry, and R.F. Powers. Timber Press, Portland, Ore. pp DeLong, C., Tanner, D., and Jull, M.J A field guide for site identification and interpretation for the southwest portion of the Prince George Forest Region. British Columbia Ministry of Forests, Victoria, B.C. Land Manage. Handb. 24. Donnelly, J.R., and Shane, J.B Forest ecosystem responses to artificially induced soil compaction. I. Soil physical properties and vegetation. Can. J. For. Res. 16: Garay, I., and Nataf, L Microarthropods as indicators of human trampling in suburban forests. In Urban Ecology. The 2nd European Ecological Symposium, Berlin, 8 12 September Edited by R. Borknkamm, J.A. Lee, and M.R.D. Seaward. Blackwell Scientific Publications, Boston. pp Hågvar, S Log-normal distribution of dominance as an indicator of stressed soil microarthropod communities? Acta Zool. Fenn. 195: Hill, M.O Diversity and evenness: a unifying notation and its consequences. Ecology, 54(2): Hill, S.B., Metz, L.J., and Farrier, M.H Soil mesofauna and silvicultural practices. In Forest Soils and Forest Land Management. Proceedings of the 4th North American Forest Soils Conference, Laval University, Québec, August Edited by D. Dernier and C.H. Winget. Les Presses de L Université Laval, Québec, Que. pp Hoekstra, J.M., Bell, R.T., Launer, A.E., and Murphy, D.D Soil arthropod abundance in coast redwood forest: effect of selective timber harvest. Environ. Entomol. 24(2): Holcomb, H The Long-Term Soil Productivity study in British Columbia. Canadian Forest Service and British Columbia Ministry of Forests, Victoria, B.C. FRDA Report 256. Huhta, V Effects of clear-cutting on numbers, biomass and community respiration of soil invertebrates. Ann. Zool. Fenn. 13: Huhta, V., Karppinen, E., Nurminen, M., and Valpas, A Effect of silvicultural practices upon arthropod, annelid and nematode populations in coniferous forest soil. Ann. Zool. Fenn. 4: Huhta, V., Nurminen, M., and Valpas, A Further notes on the effect of silvicultural practices upon the fauna of coniferous forest soil. Ann. Zool. Fenn. 6: Jordan, D., Li, F., Ponder, F., Jr., Berry, E.C., Hubbard, V.C., and Kim, K.Y The effects forest practices on earthworm populations and soil microbial biomass in a hardwood forest in Missouri. Appl. Soil Ecol. 13: Kevan, P.G., Forbes, B.C., Kevan, S.M., and Behan-Pelletier, V Vehicle tracks on high Arctic tundra: their effects on the soil, vegetation and soil arthropods. J. Appl. Ecol. 32: Klinka, K., Green, R.N., Trowbridge, R.L., and Lowe, L.E Taxonomic classification of humus forms in ecosystems of British Columbia. British Columbia Ministry of Forests, Victoria, B.C., Canada. Kranabetter, J.M., and Chapman, B.K Effects of forest soil compaction and organic matter removal on leaf litter decomposition in central British Columbia. Can. J. Soil Sci. 79: