Regenerating montane conifers with variable retention systems in a coastal British Columbia forest: 10-Year results

Regenerating montane conifers with variable retention systems in a coastal British Columbia forest: 10-Year results

Forest Ecology and Management 246 (2007) 240–250 www.elsevier.com/locate/foreco Regenerating montane conifers with variable retention systems in a co...

754KB Sizes 0 Downloads 11 Views

Forest Ecology and Management 246 (2007) 240–250 www.elsevier.com/locate/foreco

Regenerating montane conifers with variable retention systems in a coastal British Columbia forest: 10-Year results A.K. Mitchell *, R. Koppenaal, G. Goodmanson, R. Benton, T. Bown NRCanada, Canadian Forest Service, Pacific Forestry Centre, 506 West Burnside Rd., Victoria, BC V8X 1M5, Canada Received 7 November 2006; received in revised form 4 April 2007; accepted 6 April 2007

Abstract As a component of the Montane Alternative Silviculture Systems (MASS) project, this study investigates limits on the growth of montane conifers resulting from silvicultural systems with varying amounts of overstory retention. In spring, 1994, Abies amabilis (amabilis fir) and Tsuga heterophylla (western hemlock) seedlings were planted in replicated blocks of clearcut and alternative systems with retention aggregated in patches (patch cut: PC) and dispersed at high density (shelterwood: SW) and low density (green tree: GT). In addition, fertilization (F) and vegetation control (V) treatments were applied alone and in combination (FV) in each silviculture system to test the extent to which growth limitations are related to nutrient availability and competing vegetation. While total tree height and volume in the CC, GT and PC were similar after 10 years, recent growth of both species tended to be greatest in the PC, which had more shallow-soil growing degree days than the CC, possibly due to earlier depletion of the spring snow pack. The growth response in the PC (in comparison to the CC) was most pronounced in the vegetation control treatments, but significant only for western hemlock. In the reduced light environment of the SW, height and volume after 10 years was markedly lower in both species compared to the more open CC, GT and PC systems. Both western hemlock and amabilis fir showed the greatest growth response (20–80% increase in height) to vegetation control alone and in combination with fertilization in all silvicultural systems, indicating the presence of below-ground (nutrient) resource limitations. Foliar nitrogen concentrations in both species were at deficient levels in all silvicultural systems, indicating reduced availability of nitrogen on the study site, and may be a precursor to growth check, a phenomenon of decreasing or variable annual height increment in conifer regeneration that has been observed following clearcutting on nitrogen-poor, mid- and high-elevation coastal sites in British Columbia. Crown Copyright # 2007 Published by Elsevier B.V. All rights reserved. Keywords: Montane Alternative Silviculture Systems; Variable retention; Conifer regeneration; Tsuga heterophylla; Abies amabilis; Vegetation control; Foliar N

1. Introduction Public pressure to preserve structural and functional attributes of old growth forests has led to the rapid implementation of variable retention silvicultural systems in coastal British Columbia (Mitchell and Beese, 2002). While a large portion of the future timber supply in British Columbia will come from coastal montane forests, long-term retention silviculture experiments established in the 1990s have only recently begun to provide science-based information to guide sustainable forest management practices (Mitchell et al., 2004a). The Montane Alternative Silvicultural Systems (MASS) project was established in 1993 in the coastal mountains of eastern Vancouver Island to provide a long-term comparison of old-growth

* Corresponding author. Tel.: +1 250 363 0786; fax: +1 250 363 6005. E-mail address: [email protected] (A.K. Mitchell).

retention-based silvicultural systems and conventional clearcut harvesting from biological, physiological, ecological and operational perspectives (Arnott and Beese, 1997; Beese and Bryant, 1999; Mitchell et al., 2004b). This paper follows on earlier reports (Mitchell, 2001; Mitchell et al., 2004b) from an ongoing study at MASS examining the effects of these retention systems on the productivity of conifer regeneration. In a climate characterized by a deep winter snow pack followed by a short growing season, silvicultural practices that alter the opening size and distribution of overstory retention may have potentially profound effects on understory microclimate (Spittlehouse et al., 2004), competing vegetation (Beese and Bryant, 1999) and nitrogen (N) availability (Prescott, 1997), all of which can exacerbate or alleviate limitations to growth of conifer regeneration (Mitchell, 2001). Of particular concern to forest managers is the potential for reduced or variable periodic height growth (commonly referred to as ‘‘growth check’’) of shadetolerant conifer regeneration following clearcut harvesting on

0378-1127/$ – see front matter. Crown Copyright # 2007 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.foreco.2007.04.036

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

mid- to high-elevation coastal sites that has been reported since the mid-1970s (Herring and Etheridge, 1976; Husted, 1982). Growth check typically occurs several years after harvesting on nutrient-poor sites, and may be related to reduced N availability and competition from ericaceous shrubs (Titus et al., 2006a,b). It has been postulated that some level of stand retention may benefit shade-tolerant regeneration by moderating the post-harvest microclimate; however, this has generally only been demonstrated on sites with extreme surface temperatures and moisture deficits (Seidel and Cooley, 1974; Keller and Tregunna, 1976), conditions not typical of those in the temperate coastal forests of British Columbia. A more likely influence of stand retention is the amelioration of early growing season conditions including snow pack and frost that affect the growth of regenerating conifers. This study reports on the 10-year growth, survival and foliar nutrition (N and S) of planted western hemlock (Tsuga heterophylla (Raf.) Sarg.) and amabilis fir (Abies amabilis Dougl. Ex Forbes) in response to conventional clearcut and alternative silvicultural systems with retention in patches and dispersed at high and low densities. We hypothesized that varying the density and distribution of overstory retention would affect above- and below-ground resource availability and subsequently the growth of conifer regeneration. Vegetation control and fertilization treatments were applied alone and in combination within each silvicultural system to test the extent to which growth limitations are related to competing vegetation and nutrient availability. 2. Methods 2.1. Study site The MASS site is located on eastern Vancouver Island (latitude 498550 N; longitude 1258250 W) south of Campbell River in the montane moist maritime variant of the Coastal

241

Western Hemlock zone (CWHmm2) (Green and Klinka, 1994). The elevation ranges from 740 to 850 m on a gentle slope (<20%) with a northerly aspect. The site is characterized by cool temperatures (5.4 8C annual mean), a short growing season (150 frost-free days) and snow cover (up to 130 cm) for 5 months of the year (Arnott and Beese, 1997). The pre-harvest stand was a multi-storied uneven-aged oldgrowth forest (overstory 200–800 years old) with gross merchantable (>17.5 cm DBH) volumes of 975– 1038 m3 ha1 dominated by western hemlock (43–44% of total basal area (BA)) and amabilis fir (23–29% of total BA), with varying amounts of western redcedar (Thuja plicata Donn. ex. D. Don) and yellow-cedar (Chamaecyparis nootkatensis (D. Don) Spach) (Arnott and Beese, 1997). Canopy heights of amabilis fir and western hemlock were 24–32 m, and those of western redcedar and yellow-cedar were 34–42 m (Beese et al., 1995). Understory vegetation in the predominant ‘HwBaPipecleaner moss site association’ is dominated by blueberry (Vaccinium alaskense and V. ovalifolium) and mosses (Rhytidiadelphus loreus, Hylocomium splendens and Rhytidiopsis robusta). 2.2. Silviculture systems The study includes four silvicultural system treatments that were completed in fall, 1993. Each silvicultural treatment was randomly assigned (with some restrictions in the case of the clearcut; see below) to three 9-ha replicate blocks (Fig. 1). The silvicultural systems were as follows: - Clearcut (CC): Operational constraints precluded unrestricted randomization of the clearcut treatment so all three 9-ha replicate ‘‘blocks’’ were located within a single 69-ha clearcut opening; however, for the purposes of analysis, the randomization was considered unrestricted (i.e., variation

Fig. 1. Montane Alternative Silviculture Systems (MASS) project layout of silvicultural treatments.

242

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

in growing conditions among the three blocks within the 69ha clearcut was assumed to be comparable to variation among the blocks to which the other three silvicultural treatments were assigned). - Green tree (GT): A minimum of 25 trees/ha (about 5% of original basal area), uniformly spaced, were retained within each 9-ha replicate block. The residual trees were selected for relatively even distribution and windfirmness, representing the entire stand profile. - Patch cut (PC): Half of each 9-ha replicate block was harvested in three 1.5-ha openings (120 m  125 m) alternating with uncut patches. The centre of the 1.5-ha openings were no more than two tree heights from a stand edge. - Shelterwood (SW): Approximately 25% of the original stand basal area (about 200 trees/ha) was retained in a dispersed pattern within each of the 9-ha replicate blocks. The retained trees represented the entire stand profile and were selected for yarding feasibility, safety, windfirmness and residual stand structure. 2.3. Experimental design The study was designed as a split-plot experiment with each of the four silvicultural systems (CC, PC, GT, SW) assigned to three whole 9-ha blocks (Fig. 1). Within each of the replicated whole blocks, points were randomly selected from a 30 m  30 m permanent grid and used to locate 12 sub-plots (12 m  16 m) inside a 1.0-ha (PC, cut area only) or 1.6-ha (CC, GT, SW) central ‘‘core’’ area buffered by three tree lengths (two tree lengths in the PC due to the smaller cut area) from the treatment edge. Each sub-plot was divided into four quadrants (split-plots) to which one of four post-planting treatments were randomly assigned: untreated (U), fertilization (F), vegetation control (V), and vegetation control with fertilization (FV). Within each quadrant (6 m  8 m), three seedlings each of amabilis fir and western hemlock (i.e., a total of six seedlings per quadrant) were assigned to six randomly located planting spots 2 m from each other and 2 m from the edges of the quadrant (Dunsworth and Arnott, 1995). As two of the sub-plots within each replicated whole block were dedicated to destructive physiological sampling in 1996, and a further three were dedicated to destructive sampling for seedling dry weights in the spring of 1997, measurement and analyses for seedling height and stem volume in year 4, 5, 7 and 10 were based on a total of seven sub-plots per replicated whole block. 2.4. Planting stock Container (PSB 415B, Beaver Plastics Ltd., Edmonton, Alta.) seedlings were grown using conventional cultural techniques at the Pacific Regeneration Technology Nuuchah-nulth Nursery at Port Alberni, BC, in 1993. Seedlings were held in cold storage (2 8C) at that company’s facility at Campbell River, BC, until required for planting in early May 1994. To ensure that size and physiological variability of planting stock was acceptable, 25 trees of each species

were sampled at time of planting to characterize initial height, root collar diameter, dry weight, root growth capacity and frost hardiness (for details see Dunsworth and Arnott, 1995). 2.5. Post-planting treatments The fertilizer treatment (F) consisted of a one-time application at planting of 24 g of slow-release fertilizer (Nutricote 16-10-10 to deliver 4, 1, and 2 g of elemental N, P and K, respectively, 180-day formulation) placed in a speared slot within 10 cm of the seedling at the time of planting. Most of the fertilizer was released in the first year with some residual release in the second year. Treatment to control competing vegetation (V), principally Vaccinium spp. and fireweed (Epilobium angustifolium), was applied from 1994 to 1997. Prior to treatment with Vision1 (glyphosate, 35.6% active ingredient, a.i.) a mechanical cleaning of the taller vegetation was conducted in mid-July, 1994. Herbicide was applied as follows: 1st application (Aug., 1994)—0.83 kg a.i./ha, 2nd application (June, 1995)—1.12 kg a.i./ha., 3rd application (July, 1995)—1.03 kg a.i./ha., 4th application (July, 1996)—0.73 kg a.i./ha., 5th application (July, 1997)—0.87 kg a.i. /ha. Seedlings in the FV post-planting treatment plots received both fertilizer and vegetation control treatments as described above. Untreated (U) seedlings had no fertilizer or vegetation control applied to them. 2.6. Growth measurements Seedlings were assessed for height (5 mm), and basal diameter at ground level (0.5 mm) following planting and at the end of the first (Oct., 1994), second (Oct., 1995), third (Oct., 1996), fourth (Oct., 1997), fifth (Oct., 1998), seventh (Oct., 2000) and tenth (May 25–June 2, 2004; pre-flush) growing seasons. Stem volume was calculated assuming a conical shape. 2.7. Foliar nitrogen and sulphur Current-year foliage of untreated (no post-planting treatment) trees of each species was collected from all four silvicultural systems at 11 years after planting (Oct., 2005). Earlier data are also presented of fall collections from the CC at 1 (late Aug., 1995), 2 (Sept., 1996), 3 (Sept., 1997), 4 (Sept., 1998) and 5 years (Sept., 1999) after planting. Foliar samples of three trees (of each species) collected from a randomly selected location within each of the three silvicultural system blocks (replicates) were bulked except at 1-year after planting when only two of the treatment blocks were bulk sampled. At 11 years after planting, sub-sampling was increased to four randomly selected gridpoint locations within each silvicultural system block. At the Pacific Forestry Centre (Victoria, BC) foliage was dried for 48 h at 70 8C, ground to pass through a 40mesh screen and assayed for total N and S (on a dry mass basis) by an automated combustion method (FP-228, Leco Corp., St. Joseph, Michigan).

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

2.8. Understory biomass

3. Results

In late July, 2005, exploratory sampling of the aboveground biomass of understory vegetation, excluding conifer regeneration, was conducted along transects in the CC and SW systems. All vascular vegetation was clipped at the soil surface from a total of 9 (SW) and 11 (CC) circular plots (0.5 m2) on two transects in SW block 2 and three transects in CC block 3. As such, only one replicate block of the CC and SW were sampled. Each transect began at a different MASS treatment core gridpoint and ran in a randomly determined bearing direction. Distance between plots ranged between 10 and 20 m. Shrub species (Vaccinium spp., Rubus spectabilis, Sambucus racemosa) were identified and collected separately. Herbaceous plants including Rubus pedatus, Cornus Canadensis, Clintonia uniflora, Achlys triphylla, Linnaea borealis, Tiarella spp., grasses, and ferns were combined for biomass determination. E. angustifolium was collected separately from other herbaceous plants because of its high percent cover on the study site and included in the biomass determination of herbaceous plants. The biomass samples were oven-dried in paper bags at 70 8C for 48 h and then weighed. Dried and ground samples were analyzed for N as described, and content was calculated on a kilogram per hectare basis.

3.1. Survival

2.9. Microclimate Weather stations with dataloggers (CR-10, Campbell Scientific Ltd., Logan, UT, USA) located in the centre ‘‘core’’ area of two replicate blocks of each silvicultural system measured PAR (1.3 m), air temperature (5 cm and 1.3 m), soil temperature (5 and 15 cm) and windspeed (3 m). Air and soil temperature data were transformed to 5 8C growing degree days (GDD). All climate measures were calculated as totals (PAR, GDD) or as daily averages (windspeed) over the growing season (Apr.–Sept.) for 10 years (1994–2003). GDD was determined for only 1996–2003 because of missing data. 2.10. Data analyses Differences in seedling growth were analyzed by analysis of variance (ANOVA) using restricted maximum likelihood (REML) estimation (Proc Mixed procedure, SAS Institute Inc., 1992). Each species was analyzed separately (i.e., for each species, the ANOVA was based on a split-plot model with silviculture replicate blocks and post-planting treatments corresponding to whole plots and split-plots, respectively). Planned contrasts of least-square means were used to test for main effects of silviculture systems (for each post-planting treatment) and post-planting treatments (for each silviculture system). Fisher’s LSD test was used to compare pairs of least-square means. Conifer foliar N (11 years after planting) was analyzed by one-way ANOVA. Tukey’s test was used to compare foliar N means among silvicultural systems.

243

Survival of amabilis fir and western hemlock trees after 10 years was highest in the GT (fir: 93%; hemlock: 92%) and PC (fir: 93%; hemlock: 91%), intermediate in the CC (fir: 88%; hemlock: 85%) and lowest in the SW (fir: 82%; hemlock: 82%). Differences in survival among post-planting treatments were generally small; the largest variation occurring in the SW, where survival was lowest in the untreated (75%) and vegetation control (76%) treatments for fir and hemlock, respectively. 3.2. Growth Silvicultural system and post-planting treatment main effects were significant for least-square means of 10-year height and stem volume of both western hemlock (Table 1) and amabilis fir (Table 2). There was no significant silvicultural system X post-planting interaction for height or stem volume of either species. 3.3. Silvicultural system comparisons Least-square means of 10-year (total) height and stem volume and 3-year (years 7–10) height increment are presented for western hemlock (Table 1) and amabilis fir (Table 2). Height and volume growth in both species was generally greatest in the CC, GT and PC and lowest in the SW, which had the most canopy retention. Differences in tree growth between the SW and the other silvicultural systems was most pronounced (P  0.05) in plots with vegetation control (V and FV treatments). Height growth in the SW lagged significantly 2–3 years after planting in both species and continued to under-perform compared to the other silvicultural systems, particularly in the treatments with vegetation control (Figs. 2 and 3). Compared to the other silvicultural systems with vegetation control (V and FV), total (10-year) growth of western hemlock in the SW was reduced in height by 24–35% and in stem volume by 52–58%. Similarly, growth of amabilis fir in the SW V and FV treatments was reduced in height by 26–42% and in stem volume by 42–64%. Total height of untreated western hemlock was also marginally lower (P = 0.053) in the SW compared to the other silvicultural systems, while height of untreated amabilis fir in the SW was not significantly lower. Total volume and height growth in the fertilizer only (F) treatment was not significantly lower in the SW for either species. There were no significant differences in total height or total volume of both species among the more open silvicultural systems (i.e., CC = GT = PC). However, the recent 3-year trend (years 7–10) in height increment of western hemlock with vegetation control (V and FV treatments) was significantly greater in the PC compared to the CC and GT. Amabilis fir in the PC showed a similar but not significant trend (compared to the CC and GT) of greater 3-year height increment. Recent (years 7–10) mean diameter growth (data not shown) of both species was also greatest in the PC but not significantly so.

244

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

Table 1 Least-square means (1 S.E.) of 10-year growth measures in western hemlock by silvicultural system and post-planting treatment Growth measure

Silvicultural system

Post-planting treatment Untreated

Fertilized

Vegetation control

Fertilizer + vegetation control

N

P

Total height (cm)

Clearcut Green tree Patch cut Shelterwood P

327.1 (29.3)a/z 328.8 (29.2)a/z 347.3 (29.1)a/z 229.7 (29.9)a/y 0.0527

347.2 365.0 414.4 309.0 ns

(29.4)a/y (29.0)ab/y (29.2)b/y (30.1)b/y

438.9 (29.8)b/z 396.7 (29.2)bc/z 445.8 (29.2)b/z 290.3 (30.4)b/y 0.0076

421.9 (29.7)b/z 422.3 (29.3)c/z 450.5 (29.5)b/z 319.9 (29.3)b/y 0.0313

212 229 229 205

<.0001 0.0019 0.0001 0.0037

3-Year height increment

Clearcut Green tree Patch cut Shelterwood P

122.4 (17.0)a/x 115.8 (17.0)a/x 136.1 (17.0)a/x 75.4 (17.4)a/x ns

130.1 121.1 162.3 102.9 ns

(17.1)a/x (16.9)a/x (17.0)a/x (17.4)ab/x

164.7 (17.3)b/y 155.3 (17.0)b/y 207.8 (17.0)b/z 104.9 (17.6)ab/x 0.0071

147.9 (17.2)ab/xy 168.0 (17.0)b/yz 193.8 (17.1)b/z 111.2 (17.0)b/x 0.0256

212 229 229 205

0.0182 0.0003 <.0001 0.0768

Stem volume (cm3)

Clearcut Green tree Patch cut Shelterwood P

3383 3077 3045 1596 ns

4258 4185 6163 3373 ns

(968)a/y (955)a/y (960)b/y (987)a/y

7887 (982)b/z 7225 (963)b/z 7220 (960)b/z 3350 (997)a/y 0.0189

7832 (979)b/z 7953 (963)b/z 7728 (972)b/z 3707 (966)a/y 0.0165

212 229 229 205

<.0001 <.0001 0.0001 ns

Height:diameter ratio

Clearcut Green tree Patch cut Shelterwood P

63.0 (1.8)b/x 65.1 (1.8)b/xy 68.1 (1.8)b/yz 73.1 (1.9)b/z <.0015

61.7 (1.8)b/y 62.8 (1.7)ab/y 62.3 (1.8)a/y 71.9 (1.9)b/z <.0005

59.1 (1.9)ab/y 58.3 (1.8)a/y 62.7 (1.8)a/yz 66.1 (1.9)a/z <.0175

55.8 (1.9)a/x 58.6 (1.8)a/xy 62.7 (1.8)a/yz 65.6 (1.8)a/z <.0018

212 229 229 205

0.0220 <.01074 <.0407 <.0028

(964)a/y (962)a/y (958)a/y (981)a/y

Least-square mean comparisons of post-planting treatments (rows) and silvicultural systems (columns) are indicated by a,b,c and x,y,z, respectively. Means followed by the same letter are not significantly different (P > 0.05) (P values >0.05 < 0.10 are considered marginally significant) using Fisher’s LSD test.

Western hemlock stem volume after 10 years was on average 3.0- to 3.5-fold higher than amabilis fir across all silvicultural systems. Height:diameter ratio (HDR) of western hemlock was highest (P < 0.05) in the SW, generally intermediate in the PC

and lowest in the CC and GT. Height:diameter ratio of amabilis fir differed significantly between silvicultural systems only in the FV treatment, where it was higher in trees in the PC than in the CC and GT.

Table 2 Least-square means (1 SE) of 10-year growth parameters in amabilis fir by silvicultural system and post-planting treatment Growth measure

Silvicultural system

Post-planting treatment

N

P

315.3 (22.9)b/z 292.0 (23.0)c/z 335.0 (22.8)b/z 215.5 (23.4)b/y 0.0077

221 234 234 204

<.0001 0.0018 <.0001 0.0205

118.9 (14.3)b/z 115.5 (14.1)b/z 148.0 (14.2)b/z 71.9 (14.4)b/y 0.0120

118.7 (14.1)b/yz 116.7 (14.2)b/yz 147.5 (14.1)b/z 92.9 (14.4)b/y 0.0928

221 234 234 204

0.0007 0.0035 <.0001 0.0214

1038 (386)a/y 1173 (384)ab/y 1204 (383)a/y 853.1 (386)a/y ns

2352 (388)b/z 1957 (384)bc/z 2299 (388)b/z 858.0 (392)a/y 0.0405

3332 (385)c/z 2591 (386)c/z 2961 (382)b/z 1511 (391)a/y 0.0164

221 234 234 204

<.0001 0.0024 0.0002 ns

59.5 57.9 60.8 61.3 ns

55.0 56.9 61.9 56.9 ns

53.9 (2.2)a/y 56.1 (2.2)a/y 63.2 (2.1)a/z 58.3 (2.3)a/yz 0.0275

221 234 234 204

ns ns ns ns

Untreated

Fertilized

Vegetation control

Fertilizer + vegetation control

211.6 224.9 235.1 175.3 ns

284.9 (23.1)b/z 261.3 (22.9)bc/z 301.8 (23.2)b/z 175.1 (23.4)b/y 0.0042

(14.2)a/y (14.1)a/y (14.1)a/y (14.2)ab/y

Total height (cm)

Clearcut Green tree Patch cut Shelterwood P

176.5 183.0 189.8 124.9 ns

3-Year height increment

Clearcut Green tree Patch cut Shelterwood P

64.5 68.4 78.8 43.3 ns

Stem volume (cm3)

Clearcut Green tree Patch cut Shelterwood P

769.4 720.6 672.4 318.0 ns

Height:diameter ratio

Clearcut Green tree Patch cut Shelterwood P

55.4 59.8 60.5 58.8 ns

(23.1)a/y (22.9)a/y (22.8)a/y (23.8)a/y

(14.2)a/y (14.1)a/y (14.1)a/y (14.7)a/y (387)a/y (384)a/y (382)a/y (398)a/y

(2.2)a/y (2.2)a/y (2.1)a/y (2.3)a/y

73.5 79.4 93.9 61.3 ns

(23.0)a/y (22.9)ab/y (22.8)a/y (23.0)b/y

(2.2)a/y (2.2)a/y (2.2)a/y (2.2)a/y

(2.2)a/y (2.2)a/y (2.2)a/y (2.3)a/y

Least-square mean comparisons of post-planting treatments (rows) and silvicultural systems (columns) are indicated by a,b,c and x,y,z, respectively. Means followed by the same letter are not significantly different (P > 0.05) (P values >0.05 < 0.10 are considered marginally significant) using Fisher’s LSD test.

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

245

Fig. 2. Height (least-square means  1 S.E.) of western hemlock over a 10-year period in (a) untreated, (b) fertilized, (c) vegetation control and (d) fertilized and vegetation control treatments in clearcut and alternative silvicultural systems.

3.4. Post-planting treatment comparisons Both western hemlock (Table 1) and amabilis fir (Table 2) showed the greatest growth response (height and volume) to vegetation control (V) alone and in combination with fertilization (FV) under all silvicultural systems. Compared to the untreated trees, western hemlock in the V and FV treatments increased in height by 1.2–1.4 times and in volume by 2.1–2.6 times across all silvicultural systems. In amabilis fir, growth in the V and FV treatments increased by 1.4–1.8 times (height) and 2.7–4.8 times (stem volume) that in the untreated trees. In the SW, the volume response to vegetation control (V) and fertilizer (F) treatments alone and in combination (FV) was not significant in either hemlock or fir, although these post-planting treatments did increase the total height of both species. While the volume response to post-planting treatments was not significant in the SW, vegetation control and fertilizer treatments alone and in combination resulted in least-square mean stem volumes that were equal to (hemlock) or greater than (fir) the untreated trees of the other silvicultural systems. The response after 10

years to fertilization alone (F), a single application at time of planting, was generally intermediate between the untreated and the V, FV treatments, but only elicited a significant response in western hemlock in the PC (total height and stem volume) and SW (total height). In amabilis fir, the growth of fertilized-only trees was not significantly greater than the untreated trees. The combination of fertilizer and vegetation control seemed to have an additive effect (i.e., FV > V) on amabilis fir growth after 10 years, although the response was significant only for total stem volume in the CC. In western hemlock, there was no significant additive effect of fertilizer and vegetation control on growth (i.e., V = FV) in any silvicultural system. Post-planting treatment differences in HDR were only significant in western hemlock, being highest in the untreated trees of all silvicultural systems and lowest in the V and FV treatments. 3.5. Competing vegetation While measurements of above-ground understory biomass (excluding conifer regeneration) were limited to only one

246

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

Fig. 3. Height (least-square means  1 S.E.) of amabilis fir over a 10-year period in (a) untreated, (b) fertilized, (c) vegetation control and (d) fertilized and vegetation control treatments in clearcut and alternative silvicultural systems.

replicate block of the SW and CC, there were large differences in the understory biomass of these two silvicultural systems. Total understory biomass was an order of magnitude greater in the SW (10,039 kg/ha) than in the CC (999 kg/ha) (Table 3). Woody shrubs, mostly ericaceous Vaccinium spp., accounted for essentially all of the difference in biomass between the two silvicultural systems, while the biomass of herbaceous plants

was similar. Estimated total N content in understory competing vegetation in the SW (52.4 kg/ha) was 4.5 times greater than in the CC (11.7 kg/ha). As with biomass, almost all of the plant N in the SW was sequestered in ericaceous shrubs. In contrast, herbaceous plants, primarily fireweed, accounted for most of the N in competing vegetation in the CC. 3.6. Foliar nitrogen and sulphur

Table 3 Understory above-ground biomass and nitrogen content in clearcut and shelterwood systems Understory parameter

Vegetation type

Silvicultural system Clearcut

Biomass aboveground (kg/ha)

Herbaceous Shrub Total

Nitrogen content (kg/ha)

Herbaceous Shrub Total

a

552.0 477.4 1029.4

a

8.15 3.62 11.76

Shelterwood 447.6 9591 10038.7 7.51 44.89 52.4

Values represent the sub-sample mean from one replicate block.

Foliar N concentrations of conifer regeneration measured in untreated trees 11 years after planting did not vary significantly among silvicultural systems in western hemlock (0.66–0.81%) or amabilis fir (0.63–0.79%), although concentrations tended to be highest in the SW for both species (Table 4). Mean foliar sulphur concentrations also varied little among silvicultural systems in western hemlock (0.10–0.13%) and amabilis fir (0.08–0.11%). Foliar N concentrations of untreated trees in the CC declined 50% (western hemlock)–55% (amabilis fir) in the period from 2 years to 11 years post-planting (Fig. 4). A very similar pattern of decline was observed in the other silvicultural systems (data not shown).

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

247

Table 4 Foliar N and S concentrations (mean  1 S.E.) of untreated western hemlock and amabilis fir in clearcut and alternative systems Foliar element

Species

Silvicultural system

P

Clearcut

Green tree

Patch cut

Shelterwood

N (%)

Western hemlock Amabilis fir

0.746 (.031)a 0.672 (.022)a

0.662 (.057)a 0.690 (.050)a

0.766 (.023)a 0.627 (.028)a

0.809 (.080)a 0.788 (.079)a

ns ns

S (%)

Western hemlock Amabilis fir

0.110 (.005)a 0.086 (.001)a

0.096 (.007)a 0.087(.005)a

0.117(.003)a 0.082 (.001)a

0.131(.013)a 0.108 (.012)a

ns ns

Means followed by the same letter are not significantly (P > 0.05) different using Tukey’s test. P-values are for one-way ANOVA between silvicultural systems.

compared to the other silvicultural systems were less pronounced compared to near-surface GDD.

3.7. Microclimate Mean total growing season (Apr.–Sept.) light levels (PAR) at 1.3 m were highest in the CC, intermediate in the GT and PC, and lowest in the SW over 10 years (Table 5). Average daily windspeed was also highest in the CC, intermediate in the GT, and lowest in the PC and SW. Air (+5 cm) and soil (5 cm) 5 8C growing degree days (GDD) near the soil surface were markedly reduced in the CC relative to the other silvicultural systems. In contrast, the CC had more GDD farther from the surface (+1.3 m and 15 cm), although the differences

Fig. 4. Nitrogen concentrations (mean  1 S.E) of untreated current-year western hemlock and amabilis fir needles from the clearcut over an 11-year period after planting. Table 5 MASS growing season (Apr.–Sept.) climate (1994–2003) and growing degree day (GDD; 1996–2003) summary Silvicultural system Clearcut Green tree Patch cut Shelterwood 4667 Total PARa (mmol) 1.6 Windspeedb (m/s) 5 8C GDDa at 1.3 m 996 5 8C GDD at 5 cm 485 5 8C GDD at 5 cm (soil) 377 5 8C GDD at 15 cm (soil) 789 a b

3404 1.0 857 708 568 603

GDD and PAR are mean seasonal totals. Daily average for the 10-year period.

3269 0.7 857 656 622 657

1888 0.6 878 730 617 657

4. Discussion The response of 10-year-old planted western hemlock and amabilis fir varied with opening size and dispersal of stand retention, as well as with vegetation control and fertilization treatments. Ten-year survival and growth (height and stem volume) of both species in small patch cut (PC) openings (1.5 ha) with aggregated retention and in green tree (GT) openings (9 ha) with low-density dispersed retention was comparable to that in a conventional clearcut (CC) (69 ha) and supports our earlier findings that these two variable retention systems can be implemented on coastal montane sites without reducing early plantation performance (Mitchell et al., 2004a). Although total height and volume growth was similar among the CC, GT and PC systems, the later trend (years 7–10) in height (and diameter, data not shown) indicates that both western hemlock and amabilis fir are growing at a faster rate in the PC, although the differences are only significant in western hemlock height following treatment with vegetation control. Faster growth in the PC suggests that a balance of stand retention and opening size in the PC has a beneficial influence on regeneration microclimate while also allowing adequate light levels for growth. Stand edges likely had a greater influence on microclimate in the patch cuts, which were no more than two tree lengths to centre, than in systems with larger openings and less retention (CC and GT). While total seasonal light levels (PAR at 1.3 m) and air (1.3m) and soil (15 cm) 5 8C GDD was highest in the CC, air and soil 5 8C GDD closer to the soil surface (+5 and 5 cm) was markedly reduced in the CC compared to the alternative silvicultural systems. More near-surface (5 cm) soil GDD in the PC and SW (compared to the CC) is consistent with shallower snow pack or earlier spring snow pack depletion (Askin and Dragunas, 1995), and is likely related to the surrounding edges (PC) and residual trees (SW) intercepting snowfall and having a general insulating effect as indicated by average wind speeds that were half that in the CC. Conversely, colder soil rooting zone (5 cm) temperatures in the CC resulting from a lingering snow pack early in the growing season would be expected to delay or inhibit conifer growth (Lopushinsky and Max, 1990; Mitchell, unpublished data) and may account for the slower recent growth compared to the PC.

248

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

In contrast to the other silvicultural systems, high levels of dispersed retention in the SW (25% of the original basal area) resulted in lower survival and growth (height and volume) of both western hemlock and amabilis fir. Similar findings of reduced height growth in the same SW system (compared to the CC) have also been reported for advance amabilis fir regeneration (Hawkins and Moran, 2003). Low understory light seemed to be the overriding factor responsible for lower growth in the SW, where total growing season irradiance was well below those in the other silvicultural systems. Sixty percent shade, a level that is exceeded in approximately onethird of the SW understory (data not shown), has been reported to reduce biomass production in both western hemlock and amabilis fir (Mitchell and Arnott, 1995). Other variable retention studies have also found reduced conifer growth in small logging-created openings of 0.1 ha or less (openings in the SW were mostly <0.1 ha) compared to larger openings (Gray and Spies, 1996; Coates, 2000; Huggard et al., 2005). Height:diameter ratios (HDR) of western hemlock, but not amablis fir, were highest in the SW (and PC), a typical response to shading that favours relatively more carbon allocation to height than diameter growth and is more pronounced in western hemlock than amabilis fir (Mitchell and Arnott, 1995). Vegetation control tended to enhance differences in growth between the more open silvicultural systems (CC, GT and PC) and the SW, where the response to alleviation of secondary below-ground limitations from the V and VF treatments was probably muted by the primary limitation (i.e., light). However, below-ground resource limitations may have also adversely affected growth of conifer regeneration in the SW, where the high density of dispersed retention was associated with much more understory biomass of competing vegetation, particularly ericaceous shrubs (i.e., Vaccinium sp.), compared to the CC. This resulted in four-fold more N sequestered in above-ground biomass by competing vegetation in the SW than in the CC and, therefore, at least seasonally unavailable to conifer regeneration, although this must be balanced against the contribution of Vaccinium and fireweed litter and subsequent N turnover to the forest floor (Kimmons et al., 2002). The much greater biomass of shrubs in the SW is due largely to less damage sustained from logging activity within the undisturbed clumps of retained trees (Beese and Bryant, 1999). The greater amount of N tied up in understory vegetation in the SW did not seem to have a treatment effect on conifer foliar N concentrations, which tended to be slightly higher in the SW compared to the other silvicultural systems where N uptake was diluted in larger trees. Above-ground biomass and N content of understory vegetation in the CC was far lower than that reported in a 10-year-old clearcut on a submontane CWH site on southern Vancouver Island dominated by fireweed, Vaccinium sp., and salal, Gautltheria shallon (Kimmons et al., 2002). The higher elevation and absence of salal on the MASS site likely accounts for much of this difference. Vegetation control alone and in combination with fertilizer more than doubled stem volume in hemlock and tripled (V treatment) to quadrupled (FV treatment) stem volume in fir across all silvicultural systems, although the volume response

was not significant in the SW. Amabilis fir showed a larger growth response (height and volume) to vegetation control treatments than did western hemlock in all silvicultural systems, relative to untreated trees. Vegetation control treatments generally reduced the time to free growing height (0.75 m for amabilis fir and 1.25 m for western hemlock in zonal CWHmm2; Forest Practices Code of British Columbia, 2000) by 1 year in both western hemlock and amabilis fir. The growth response to vegetation control underscores the effect of competing vegetation on below- and above-ground resources. Vegetation control would have increased the availability of nutrients, particularly N to regeneration, as well as available light in microsites where regeneration was overtopped by surrounding vegetation in the first 5–7 years after planting. After 7 years however, competition for light was much less of a factor, as regeneration height was greater than that of the surrounding vegetation. These results are confirmed by a related study from the same site where removal of competing understory vegetation in the CC and SW resulted in large increases in available mineral N and increased foliar N and growth of western hemlock seedlings measured one growing season after planting (Maynard et al., 2004). In a study on two Vaccinium-dominated, montane sites on Vancouver Island, Titus et al. (2006a) also observed increased N availability and an improved 5-year growth response to vegetation control in amabilis fir, western hemlock and yellow-cedar. Another related study on the MASS site found increased root collar diameter of advance amabilis fir following vaccinium removal in the CC, GT and SW, although N concentrations in foliage were unaffected (Hawkins and Moran, 2003). The response to fertilizer alone, applied only once at the time of planting, was most pronounced early in the study, as described in a previous paper (Mitchell et al., 2004b), but was diminished after 10 years in both species except in the SW and PC (hemlock only). The lack of, or inconsistent, response of these conifer species to fertilization has also been observed in other montane studies on Vancouver Island—results that are as yet largely unexplained given the poor nutrient status of these sites (Hawkins and Moran, 2003; Titus et al., 2006a). Nonetheless, there was an additive effect on the volume of amabilis fir when both fertilizer and vegetation control were applied (i.e., FV > F, V), indicating a cumulative response to the initial fertilizer treatment (at time of planting) and the longer-term increase in nutrient and light availability from vegetation control. Although there was an early (2 years post-harvest) but relatively small assart effect (elevated N mineralization and nitrification rates) that was most pronounced in the CC (Prescott, 1997; Titus et al., 2006b), foliar N of regenerating conifers after 11 years was similar among silvicultural systems. These findings are consistent with other studies of similar time since harvest (9–10 years post-harvest) that report little difference in foliar N concentrations between overstoryinfluenced and clearcut regeneration of both amabilis fir (Hawkins et al., 2002) and western hemlock (Kranabetter and Coates, 2004) regeneration. Nitrogen is regarded as limiting to growth in coastal forests of British Columbia and the Pacific Northwest (Prescott and Zabek, 1999) and has been positively

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

correlated with tree growth and site index in these forests (Radwan and DeBell, 1980; Radwan et al., 1989; Kayahara et al., 1995; Kranabetter et al., 2003). Foliar N concentrations in both western hemlock and amabilis fir have been in decline since seedling establishment at the MASS site, falling about 50% from 2 years after planting, a period coincident with a post-harvest assart flush of nutrients on this site, to 11 years after planting. These N concentrations (western hemlock: 0.66– 0.81, amabilis fir: 0.63–0.79) are at the lower range reported for both species on several coastal sites in British Columbia and the Pacific Northwest (Radwan et al., 1989; Kayahara et al., 1995; Kranabetter et al., 2003), reflecting low rates of decomposition (Prescott and Zabek, 1999) and N mineralization (Grenon et al., 2004) on this site, compounded by an increasing sink of N in competing vegetation occurring with post-harvest colonization and development. Bradley et al. (2002) reported a similar decline in foliar N concentrations of amabilis fir 7 years after clearcutting on a chronosequence of submontane CWH sites and suggested that the increasing post-harvest development of understory N sinks, predominately fireweed, Vaccinium sp. and salal, may have contributed to reduced N uptake and growth of advance amabilis fir regeneration measured 8 years after clearcutting. Of particular concern to forest managers are recurrent problems of reduced or variable annual height growth of conifer regeneration (commonly referred to as growth check) observed several years after clearcutting at high elevations (Herring and Etheridge, 1976), often in association with N deficiency (Husted, 1982) and shrub competition (Titus et al., 2006a). While the increasing trajectory of height growth in both species in the CC, GT and PC systems indicates that growth check has not yet occurred on the study site, foliar N concentrations (western hemlock <0.8%, amabilis fir <0.7%) in these silvicultural systems were severely deficient for western hemlock (<0.95%, Ballard and Carter, 1985) and well below critical levels for true fir (<1.15%, Powers, 1983), suggesting that current growth rates may not be sustainable. 5. Management implications  Growth of montane conifers in the patch cut and green tree systems was comparable to that in the clearcut after 10 years and supports the view that aggregated and low-density dispersed retention systems can be implemented on coastal montane sites without reducing early plantation performance. Recent height growth (years 7–10) of western hemlock and amabilis fir tended to be greatest in the patch cuts, which may create a favourable microenvironment for shade-tolerant conifer regeneration.  In the shelterwood, low understory light contributed to poorer growth and delayed the time to free-to-grow height by 1 year in western hemlock and by up to 2 years in amabilis fir. These results suggest that lower densities of dispersed retention or grouped shelterwood would allow more light into the understory for conifer regeneration. Vigorous understory vegetation in the shelterwood, particularly ericaceous shrubs, may have also contributed to poorer growth.

249

 Control of competing vegetation more than doubled stem volume in western hemlock and about tripled stem volume in amabilis fir in the clearcut, green tree and patch cut systems, and reduced the time to free-grow height by about 1 year in both species. However, the purpose of the post-planting treatments (F, V, FV) used in this study was not to test application rates used operationally, but rather to study their effects on resource limitations to conifer growth.  Although there was no measurable evidence of growth rates declining after 10 years, very low foliar N concentrations in all silvicultural systems on this site may be a precursor to conifer growth check, which has been observed on other Npoor, high-elevation sites in coastal British Columbia over a similar period of time since harvest. Acknowledgements The authors gratefully acknowledge partial funding for this study provided by the Forest Investment Account, Forest Science Program of British Columbia. We also wish to acknowledge Bill Beese (Western Forest Products) and Glenn Dunsworth (Dunsworth Ecological Consulting) for their contributions as partners in the MASS project, and Dr. Brian Titus, Glenda Russo and Monique Keiran of the Pacific Forestry Centre (Canadian Forest Service) for their helpful comments during preparation of this manuscript. Thanks are also due to Amanda Nemec (International Statistics and Research Corp.) for her assistance with statistical analysis. References Arnott, J.T., Beese, W.J., 1997. Alternatives to clearcutting in BC coastal montane forests. For. Chron. 73, 670–678. Askin, R.W., Dragunas, V.P., 1995. MASS: snow hydrology pilot study. In: Arnott, J.T., Beese, W.J., Mitchell, A.K., Peterson, J. (Eds.), Montane Alternative Silvicultural Systems, Proceedings of a Workshop, Courtenay, BC, 7–8 June, 1995. Canada-British Columbia Forest Resource Development Agreement FRDA Report 238, pp. 113–122. Ballard, T.M., Carter, R.E., 1985. Evaluating forest stand nutrient status. B.C. Min. For., Land Manage. Rep. No. 29, 60 p. Beese, W.J., Bryant, A.A., 1999. Effect of alternative silvicultural systems on vegetation and bird communities in coastal montane forests of British Columbia, Canada. For. Ecol. Manage. 115, 231–242. Beese, W.J., Sandford, J., Toms, T., 1995. Montane alternative silvicultural systems (MASS) forest structure and natural vegetation dynamics. In: Arnott, J.T., Beese, W.J., Mitchell, A.K., Peterson, J. (Eds.), Montane Alternative Silvicultural Systems, Proceedings of a Workshop, Courtenay, BC, 7–8 June, 1995. Canada-British Columbia Forest Resource Development Agreement FRDA Report 238, pp. 113–122. Bradley, R.L., Kimmons, J.P., Martin, W.L., 2002. Post-clearcutting chronosequence in the B.C. Coastal Western Hemlock Zone: II. Tracking the assart flush. J. Sust. For. 14, 23–43. Coates, K.D., 2000. Conifer seedling response to northern temperate forest gaps. For. Ecol. Manage. 127, 249–269. Dunsworth, B.G., Arnott, J.T., 1995. Growth limitations of regenerating montane conifers in field environments. In: Arnott, J.T., Beese, W.J., Mitchell, A.K., Peterson, J. (Eds.), Montane Alternative Silvicultural Systems, Proceedings of a Workshop, Courtenay, BC, 7–8 June, 1995. Canada-British Columbia Forest Resource Development Agreement FRDA Report 238, pp. 48–68. Forest Practices Code of British Columbia, 2000. Establishment to Free Growing Guidebook, Vancouver Forest Region, revised ed., version 2.2.

250

A.K. Mitchell et al. / Forest Ecology and Management 246 (2007) 240–250

Gray, A.N., Spies, T.A., 1996. Gap size, within gap position and canopy structure effects on conifer seedling establishment. J. Ecol. 84, 635–645. Green, R.N., Klinka, K., 1994. A field guide to site identification and interpretation for the Vancouver Forest Region. B.C. Min. For., Nanaimo, B.C., Land Manage. Handbook No. 28. Grenon, F, Bradley, R.L., Joanisse, G., Titus, B.D., Prescott, C.E., 2004. Mineral N availability for conifer growth following clearcutting: responsive versus non-responsive ecosystems. For. Ecol. Manage. 188, 305–316. Hawkins, B.J., Moran, J.A., 2003. Growth responses of Abies amabilis advance regeneration to overstory removal, nitrogen fertilization and release from Vaccinium competition. For. Sci. 49, 799–806. Hawkins, B.J., Koppenaal, R., Mitchell, A.K., 2002. Proximity to a stand edge influences growth of advance and planted Pacific silver fir seedlings. Northwest Sci. 76, 221–229. Herring, L.J., Etheridge, D.E., 1976. Advanced amabilis fir regeneration in the Vancouver Forest District. B.C. For. Serv. and Can. For. Serv. Joint Rep. No. 5. Huggard, D.J., Arsenault, A., Vyse, A., Klenner, W., 2005. The Opax Mountain Silvicultural Systems project: Preliminary results for managing complex, dry interior Douglas-fir forests. B.C. Min. For. Ext. Note 72, 16 pp. Husted, L., 1982. The relationship between foliar and soil chemistry, growth parameters, and variable height growth in advance regeneration of amabilis fir. M.Sc. Thesis. Dept. of For., Univ. of BC, Vancouver, BC, 123 p. Kayahara, G.J., Carter, R.E., Klinka, K., 1995. Site index of western hemlock (Tsuga heterophylla) in relation to soil nutrient and foliar chemical measures. For. Ecol. Manage. 74, 161–169. Keller, R.A., Tregunna, E.B., 1976. Effects of exposure on water relations and photosynthesis of western hemlock in habitat forms. Can. J. For. Res. 6, 40–48. Kimmons, J.P., Martin, W.L., Bradley, R.L., 2002. Post-clearcutting chronosequence in the B.C. Coastal Western Hemlock Zone: III. Sinks for mineralized or dissolved organic. N. J. Sust. For. 14, 45–68. Kranabetter, J.M., Coates, K.D., 2004. Ten-year postharvest effects of silviculture systems on soil-resource availability and conifer nutrition in a northern temperate forest. Can. J. For. Res. 34, 800–809. Kranabetter, J.M., Banner, A., Shaw, J., 2003. Growth and nutrition of three conifer species across site gradients of north coastal British Columbia. Can. J. For. Res. 33, 313–324. Lopushinsky, W., Max, T.A., 1990. Effect of soil temperature on root and shoot growth and on budburst timing in conifer seedling transplants. New For. 4, 107–124. Maynard, D.G., Mitchell, A.K., Titus, B.D., Moran, J., 2004. Relationships between soil nutrient availability and seedling growth after two different harvesting and vegetation control treatments on a coastal montane site. Poster presented at the Variable Retention Forestry Science Forum, April 21–22, Nanaimo, BC. Mitchell, A.K., 2001. Growth limitations for conifer regeneration under alternative silvicultural systems in a coastal montane forest in British Columbia, Canada. For. Ecol. Manage. 145, 129–136.

Mitchell, A.K., Arnott, J.T., 1995. Effects of shade on the morphology and physiology of amabilis fir and western hemlock seedlings. New For. 10, 79–98. Mitchell, A.K., Vyse, A., Huggard, D.J., Beese, W.J., 2004a. Long-term silviculture experiments contribute to science-based forest management in British Columbia’s public forests. For. Snow Landsc. Res. 78, 139–150. Mitchell, A.K., Dunsworth, B.G., Arnott, J.T., Koppenaal, R., Benton, R., Goodmanson, G., Bown, T., Sandford, J., 2004b. Growth limitations of planted conifers regenerating under montane alternative silviculture systems (MASS): seven-year results. For. Chron. 80, 241–250. Mitchell, S.J., Beese, W.J., 2002. The retention system: reconciling variable retention with the principles of silvicultural systems. For. Chron. 78, 397–403. Powers, R.F., 1983. Forest research fertilization in California. In: Ballard, T.M., Gessel, S.P., Stanley, P. (Eds.), IURFO Symposium on Forest Site and Continuous Productivity, pp. 388–397. USDA For. Serv. Gen. Tech. Rep. PNW-163. Pacific Northwest Forest and Range Experimental Station. Portland, OR. Prescott, C.E., 1997. Effects of clearcutting and alternative silvicultural systems on rates of decomposition and nitrogen mineralization in a coastal montane forest. For. Ecol. Manage. 95, 253–260. Prescott, C.E., Zabek, L.M., 1999. Decomposition and nitrogen mineralization in forests of British Columbia: effects of forest management practices. In: Meurisse, R.T., Ypsilantis, W.G., Seybold, C. (Eds.), Proceedings of Pacific Northwest Forest and Rangeland Soil Organism Symposium, Corvallis, OR, March 17–19, 1998. U.S.D.A. For. Serv. Pac. Northwest Res. Sta. Gen. Tech. Rep. PNW-GTR-461, pp. 124–136. Radwan, M.A., DeBell, D.S., 1980. Site index, growth, and foliar chemical composition relationships in western hemlock. For. Sci. 26, 283–287. Radwan, M.A., Murray, M.D., Kraft, J.M., 1989. Growth and foliar nutrient concentrations of Pacific silver fir. Can. J. For. Res. 19, 1429–1435. SAS Institute Inc., 1992. SAS/STATR User’s Guide. SAS Institute, Cary, NC. Seidel, K.W., Cooley, R., 1974. Natural reproduction of grand fir and mountain hemlock after shelterwood cutting in central Oregon. USDA For. Serv. Res. Note PNW-229. Pacific Northwest Forest and Range Experimental Station. Portland, OR. Spittlehouse, D.L., Adams, R.S., Winkler, R.D., 2004. Forest, edge and opening microclimate at Sicamous Creek. B.C. Min. For. Res. Report. 24. 43 p. Titus, B.D., Prescott, C.E., van Niejenhuis, A., Hogg, K., Barker, J., 2006. Fertilization and Vegetation Control to Reduce Conifer Growth Check on Vaccinium-Dominated Cutovers at High Elevations in Coastal BC: 5-Year Results. Forest Investment Account, Forest Science Program (British Columbia) Final Report, Project No. 16R1607. 54 p. Titus, B.D., Prescott, C.E., Maynard, D.G., Mitchell, A.K., Bradley, R.L., Feller, M.C., Beese, W.J., Seely, B.A., Benton, R.A., Senyk, J.P., Hawkins, B.J., Koppenaal, R., 2006b. Post-harvest nitrogen cycling in clearcut and alternative sivicultural systems in a montane forest in coastal British Columbia. For. Chron. 82, 844–859.