The effect of soil water-table depth on root-plate development and stability of Sitka spruce DUNCAN RAY and BRUCE C. NICOLL Forestry Commission Research Agency, Northern Research Station, Roslin EH25 9SY, Scotland Summary Stability was tested on 46-year-old Sitka spruce trees (over 20 m tall) growing on gleyed soils. Trees with a range of root depths on peaty gley and surface-water gley soils were selected by studying their water regime. Trees were pulled over horizontally with a winch and the vertical displacement of the soil-root plate was measured as the load was applied. The resistive turning moment at the point of soil failure and the maximum resistive turning moment were measured. Thin plates, which developed over a shallow water table, had a greater surface area than thick plates that developed over deeper water tables. However, they also had a smaller soil mass and tended to shear in soil material of lower shear strength and were particularly flexible. Thicker plates sheared in stronger soil and were more rigid. Soil-root plate rigidity was shown to be a major factor affecting the contribution of soil resistance to overturning. Regression analysis revealed that the more rigid the plate the greater the applied turning moment at soil failure. Trees that develop rigid soil-root plates through adaptive growth of their structural roots will therefore also have improved stability. The results indicate that intensive drainage of peaty gleys will increase rooting depth and the resistance to overturning. Introduction Sitka spruce (Picea sitchensis (Bong.) Carr.) has been extensively planted in the uplands of Britain, particularly in the north and west. However, high rainfall in combination with soils of low hydraulic conductivity leads to frequent waterlogging except for short periods in the summer (Pyatt and Smith, 1983) when evapotranspiration is higher. Waterlogging causes anaerobic soil conditions (Boggie, 1977; King et al., 1986) which are unfavourable for the survival of growing Sitka spruce roots (Coutts and O littritute of Qurtercd Foretterx, 199S Philipson, 1978). This leads to shallow root plates which are particularly unstable in the windy British climate (Quine et al., 1995). Any silvicultural treatments that improve rooting depth, and hence stability, would lead to longer rotations with the benefits of increased production as well as improved forest biodiversity (Peterken et al., 1992; Ratcliffe, 1993). In this paper the stability and root development of trees in relation to water-table depth, using undrained and drained areas of a plantation on two types of soil are investigated. A Fomrry, Vot 71, No. 2 , 1998 170 FORESTRY previous study on the same site (Ray et al., 1992) showed that intensive drainage (10 m spacing) of peaty gley soil significantly lowered the water table by 0.07-0.15 m compared with the currently recommended wider drain spacing of 40 m (Pyatt, 1990). In this study the water table below trees was measured in two winter periods before the trees were pulled over and the soil-root plate characteristics measured. This allowed an assessment of the effect of drainage on water-table depth, and of the relationship between watertable depth and tree morphology and stability. Drains were installed on this site when the trees were 17 years old. Although the structural root systems of forest trees commonly develop from roots that have grown to be largest within the first 8-15 years (Deans, 1981; Courts, 1983a) of the tree's life, the root system remains highly responsive to changes in the soil environment. Courts (1983a) found small roots increasing in size as late as 18 years to become important structural roots, and Wagg (1967) found large changes in the root system architecture of white spruce in response to changing water tables and soil surface levels until the trees were over 70 years old. The authors therefore expected root systems on this site to have exploited, and developed into, progressively deeper soil as a result of a deeper water table following drainage. Coutts (1983b) suggested that shear would begin at the edge of a rigid soil-root plate and move rapidly towards the centre as a tree started to overturn. In such a system the calculated turning moment to fracture the soil-root plate was six times greater than the observed turning moment of flexible plates. Measurements of soil displacement while a turning moment was applied to a thin, flexible soil-root plate showed that soil sheared first under the plate close to the centre of the stem on the windward side, spreading quickly outwards towards the edge of the soil-root plate (Coutts, 1983b and 1986). There have been a number of studies demonstrating the effect of various silvicultural techniques on root development and tree stability. Ground preparation techniques have a major influence on the radial symmetry of a root system. For example Savill (1976) reported that shallow furrows 0.3 m deep could restrict the lateral growth of Sitka spruce planted on surface-water gleys. Restricted root spread, e.g. by spaced furrow ploughing, has been shown to reduce the stability of conifers, particularly when root depth is also restricted (Pyatt and Booth, 1973; Savill, 1976; Hendrick, 1989). Quine et al. (1991) found marked asymmetry in the distribution -if spruce roots when planted next to or close to a stump, whereas the roots of trees planted more than 0.7 m from a stump had better radial symmetry. In plantations on cold wet sites, some ground preparation is essential for successful tree establishment, and a mound provides a warmer, drier, less compact, well aerated, weed-free planting site in which soil nutrients are more accessible than in uncultivated sites. These factors combine to produce good establishment success (Binns, 1962; Taylor, 1970; Tabbush and Ray, 1989). In particular, the higher soil temperature gained by a mound early in the growing season (Ray and Anderson, 1990) will help initiate and stimulate root growth (Coutts and Philipson, 1987) at a time that is critical to the development and establishment of the structural root system (Deans, 1981; Coutts, 1983a; Coutts, 1987). Unlike continuous plough ridges with furrows at 4-m spacing, mounds are less restrictive to lateral root growth, producing trees with better radial root symmetry (Quine et al., 1991). In this study the stability of large, 46-year-old trees planted on spread turfs was examined. The 'spread turP (or 'Belgian turf method') was used extensively in the 1940s and 1950s, and involved excavating a shallow (approx. 0.3 m) 'turf drain', manually or with a plough, to provide 'cut turfs' which were then spread manually into two rows, providing planting positions at 1.8—2.0 m spacing. The authors believe this study is applicable to new planting sites cultivated using 'continuous acting' mounding machines as well as drain mounding and hinge mounding techniques, which are both currently used on wet peaty gley and surface-water gley restock sites (Forestry Commission, 1993). Experimental details Site The experiment was located at Crookburn in Kershope forest, north-east Cumbria, UK WATER-TABLE EFFECT ON STABILITY OF SITKA SPRUCE (200-245 m altitude, latitude 55°66'N, longitude 2°47rW). Exposure is moderate with a southeasterly aspect, with windthrow hazard class (WHC) (Quine and White, 1993) varies between 4 and 5. WHC increases with windspeed and exposure on a scale of 1 to 6. Soils Soils have developed a 3-m thick layer of glacial till, and vary from peaty gleys (humic stagnoorthic gley; Avery, 1990) on gentle slopes to surface-water gleys (stagno-orthic gley) on steeper slopes. The soils are of the Carter Association (Muir, 1956) and have a slowly permeable subsoil of clay-loam texture, and a saturated hydraulic conductivity ranging from 0.6 to 60 mm day"1 (Ray et al., 1992). Soils equivalent to those in the Carter Association are described by the Soil Survey of England and Wales as the Wilcox 1 Association (Jarvis et al., 1984). Silviculture Sitka spruce was planted in 1948 on spread turfs at a spacing of 1.8 m. The turfs were cut from a plough ridge turned out of shallow (~0.3 m deep) furrows at a spacing of approximately 3.6 m. The spread turf method provided a planting position similar to a mound, but with a shallow furrow between every two rows of trees. In 1965, a drainage experiment was installed on the site (Pyatt et al., 1985), with three drain spacing treatments of 10 m, 20 m and 40 m. Half of the experiment was felled between 1983 and 1985, and replanted with a mixtures experiment in 1986 (Ray et al, 1992). However, the tree pulling study was located in a 20-ha area within the unfelled half of the drainage experiment, and in an adjacent undrained stand of trees. The area of forest had not been thinned, apart from cutting racks to enable drainage in 1965. Selection of sample trees In 1991, 100 dominant and co-dominant trees were selected on a stratified-random basis, by dividing the area into approximately 1-ha squares and randomly choosing five of the largest trees in each square. This ensured that 171 sample trees were selected in undrained, moderately drained and intensively drained stands, growing on both peaty-gley and surface-water gley soils. A tree was rejected and a replacement selected: if it was next to a drain, where it did not have an adjacent furrow. Trees were also rejected if they were 'pumping'. A tree is described as 'pumping' when its soil-root plate has sheared from the sub-soil, allowing the plate to rock freely in the hollow. This rocking causes soil water under the tree to mix with the fine mineral material in the soil, forming a 'soupy' suspension that is pumped out of the hollow as the soil-root plate rocks. As the soil-root plates of 'pumping' trees have already sheared from the sub-soil, they are unsuitable for a study of plate bending during soil failure. Coutts (1983b and 1986) chose dominants in his tree pulling studies at Kershope forest, and the same methodology was followed in this study. The dominant trees may be the most vulnerable to wind damage, since they are taller with larger crowns than sub-dominant trees. Water-table study A borehole lined with 0.05-m diameter, 1-m long perforated plastic pipe, was installed in a position 1 m from the geometric centre of each tree on the side away from the furrow. The depth to the water table was measured each week for 8 weeks in November and December 1991. From the initial 100 trees, 50 were selected on the basis of the depth to the borehole water level below the ground surface, i.e. with a range of water-table depths, which were expected to produce a similar range of tree rooting depth. A further three boreholes were installed around each of these 50 study trees. Three were installed vertically, 1 m from the geometric centre of the tree, and a fourth was installed at an angle of about 60° close to the stem on the furrow side of the tree, to enable measurements of the water table directly under the tree stump. Although the water level in the three vertical boreholes was measured directly, it was necessary to calculate the water level in the angled borehole. The borehole water level in the four positions under each of the 50 study trees was measured for two, 10-week winter periods: period 1, 12 172 FORESTRY February-15 April 1992; period 2, 21 October-23 December 1992. Two trees were blown over by winter gales early in period 2, and were removed from the assessment. Felling The site was clearfelled during the period October 1993 to June 1994 by a mechanized harvester. The tops of the study trees were removed leaving a 3-m high stem. The machines on the site were excluded from the area within 4 m of each study tree to avoid damage to soilroot plates. During the felling operation a further nine trees were blown over and had to be removed from the study. The blown trees were close to a harvester 'drift' and so were critically exposed along a newly created east facing edge during an easterly gale. Tree pulling The force required to pull over horizontally each of the 39 remaining trees was measured by a 10 tonne load cell connected to a data logger. A load was applied to each tree, across the load cell, by attaching one end of a steel cable to the tree at 2 m height and the other to a hydraulic, tractor-mounted winch. The load and time were recorded at 1.5 s intervals by a data logger during the pull. Every soil-root flate was pulled in a consistent manner by positioning the tractor 25 m away, anchoring the cable to the tree 2 m above the ground, returning the cable on to the drum of the winch at the same slow rate (0.02 m s~'), and pulling each soil-root plate away from the furrow. The angle of the cable from horizontal was also recorded. The side of the tree towards the pull direction, equivalent to the leeward side of a windthrown tree, is here described as the 'pull side', and the side away from the pull direction, equivalent to the windward side, is here termed the 'furrow side'. from the centre of the tree on the furrow side and the last 1.2 m on the pull side. Each sensor consisted of a displacement transducer mounted on a stand, which was made from a 0.3-m long piece of 0.18-m diameter plastic pipe (Figure 1). A 1-m long mild steel rod was located within a 0.04-m diameter hole, augered to a depth of 0.5 m through the top soil. The rod was then hammered into the subsoil, to a depth of approximately 0.95 m, leaving 0.05 m protruding. This ensured that the rod did not move when the soil-root plate was pulled. The cord of a displacement transducer was attached and the movement sensor was then placed over the steel rod. Two spikes (0.2 m long) fixed to the base Direction of pull Elevation Tree n/n nln n nn Furrow Movement sensor/ o o d o cj> o o o o i\ i m I\ I ^ ^ F Direction of pull ooQpqoooo Displacement trmoiducer Root plate movement and bending Stedrod Eighteen movement sensors were installed on the soil-root plate at 0.40-m intervals in two parallel transects, parallel to the line of pull. The first movement sensor was positioned 2.0 m ! Figure 1. Plan and elevation of movement sensor layout around a study tree, and detail of a movement sensor. WATER-TABLE EFFECT ON STABILITY OF SITKA SPRUCE of the stand secured the movement sensor firmly to the soil-root plate when pressed into position (Figure 1). Sensors were connected to a data logger which recorded the rime and displacement at 1.5-s intervals until the furrow side of the soilroot plate had been displaced by about 0.3 m. This limit ensured that transducers were not damaged by pulling the cord beyond the furthest point of travel. So/7 and soil-root plate measurements Having measured the load applied to the soilroot plate beyond the maximum value, movement sensors were disconnected and removed, and the plate was pulled through 90° and secured to allow safe working beneath it. Root plate measurements (Nicoll and Ray, 1996) were made of the width and depth along eight cardinal radii from the centre of the stem. The soil was described and sampled from a soil profile pit excavated next to each plate. The depth of each soil horizon was measured, and three soil-cores of known volume were taken from each of the mineral soil horizons around the edge of the plate, and bulked together. In peat, the core sampler caused too much compression. Instead, undisturbed blocks of peat (about 0.2 X 0.2 X 0.2 m in size) were cut from the soil pit, and sealed in plastic bags. The volume of each peat block was determined by a fluid displacement method (Pyatt and John, 1989). All samples were weighed fresh, before drying to constant weight at a temperature of 80°C. It was therefore possible to calculate the water content and the wet bulk density of each soil horizon at the rime of tree-pulling. Soil-root plates were too large and heavy to weigh directly; instead we estimated the mass of each plate by multiplying the volume of soil material in each horizon by its wet bulk density. A Pilcon Hand Vane Tester (English Drilling Equipment Co. Ltd, Huddersfield, UK) was used to measure the soil shear strength within the horizon exposed at the shear plane, and in the other soil horizons exposed by the soil-root plate. Six replicate readings were made in each horizon exposed, and exceptionally high readings, caused by the vane touching a stone or root, were discarded. Soil was then cleaned from each plate using a 173 jet of compressed air from an 'air-knife' (Briggs Technology Inc., Pittsburgh, US). The root systems were removed from the site and transported to the Northern Research Station for root architecture studies (Nicoll and Ray, 1996). The woody root systems were also too heavy to weigh fresh, and because there was a 2-3 week gap between tree pulling and root cleaning, it was necessary to calculate fresh root weight. Samples of fresh root from two plates were weighed immediately following tree pulling. The samples were dried at 80°C to constant weight, from which the water content and density of the wood were calculated. Later in the study the air-dry, cleaned roots were weighed, and again samples of root were taken and weighed, dried to constant weight at 80°C, and re-weighed as before. From the differences in the water content between the fresh roots, air dry roots, and the oven dried samples, the fresh weight of each root system could be calculated. Statistical analysis Data presented in Tables 1-3 were analysed using analysis of variance. Other relationships were explored using linear regression. Throughout this paper, differences are described as significant at the P < 0.05 level and as highly significant at the P < 0.01 level. Results Rainfall and water-table depth The mean annual rainfall of the site is 1400 mm (Anderson et al., 1990) giving a weekly rainfall of approximately 30 mm. Both assessment periods were slightly wetter than average, as was expected during the winter, with 319 mm of rain falling over the 10-week period (12 February-15 April 1992), and 371 mm of rain falling in the second 10-week period (21 October-23 December 1992). There was a large range in the mean winter borehole water level between the 50 trees (0.25-0.60 m). Table 1 shows the effect that an intensive drainage system had on the mean winter water-table depth, and how the drainage effect varied between soil types. Sample trees in 174 FORESTRY Table 1: Mean winter water-table depth (m) under sample trees growing on surface-water gley (peat thickness 20.10 m), shallow phase peaty gley (peat thickness >0.10 and 20.25 m) and deep phase peaty gley (peat thickness >0.25 and 20.45 m), and the effect of drainage on the water table within each soil type Soil type All trees Low intensity drains High intensity drains Surface-water gley Shallow phase peaty gley Deep phase peaty gley 46.7 42.8 35.9 45.2 35.7 33.0 48.0 48.0 46.0 Significant differences *, differences significant at P £ 0.05; n.s., not significant. a stand with 10 m or 20 m spaced drains are described as having a high intensity drainage system, whereas sample trees in stands with no drainage or with drains at 40 m spacing are described as having a low intensity drainage system. On the surface-water gley soils, the depth of the mean winter water table was not significantly different between trees in areas of high intensity and low intensity drainage. However, on the peaty gley soils the water table was significantly deeper with high intensity drainage than with low intensity drainage. The mean winter water table was deeper on surface-water gley soils than on peaty gley soils, differences approached significance (P = 0.07). 1.00 0.K 040 0.70 040 0.10 0.40 0J0 0-10 0.10 0.00 0-10 Tree dimensions in relation to site There was a slight, but not significant, inverse trend between increased above ground biomass and deeper winter water tables. Both crown width and crown depth increased with tree mass. Root:shoot ratio, i.e. fresh root weight divided by stem and canopy weight was significantly (negatively) related to mean winter water-table depth (Figure 2). Although the data show considerable scatter (r2 = 0.3), the regression coefficient was highly significant. Tree pulling An example of soil-root plate displacement, while a turning moment was applied about the base of a tree, is shown in Figure 3a. The horizontal positions represented on the jc-axis as negative numbers identify movement sensors positioned on the furrow side, whereas positive values identify movement sensors on the pull side. Individual lines on the graph represent the 0J6 040 044 0.40 0.41 0M I n n wlnttr wittr UbW dipt* (m) 041 040 Figure 2. The relationship between root: shoot ratio and the mean winter water-table depth for all study trees ( • , linear regression ). position of the soil-root plate at 1.5-s time intervals as the load was applied, causing the plate to move. On the pull side, the compressive forces bearing on the soil caused soil failure and compaction at 0.8 m from the stem. On the furrow side, the plate lifted causing bending followed by shearing from the underlying soil, initially at a point between -0.4 m and -0.8 m from the centre of the tree. Figure 3b shows the applied turning moment, and the change in the applied turning moment for the same tree plotted as a time series. In previous studies (Courts, 1986; O'Sullivan and Richie, 1993), the applied turning moment (M) was plotted against displacement (D). In this study, the load was applied slowly and continuously through time allowing M to be plotted against time (f) The WATER-TABLE EFFECT ON STABILITY OF SITKA SPRUCE 175 failed, and the position of the root plate and its curvature were measured, at the time of failure. ^ tkK root taDor* Trees in the present study were large and showed considerable buttressing of the structural roots close to the stem (Nicoll and Ray, 1996) which resulted in very little bending of the root plate closer than 0.5 m from the centre of the stem. Curvature was calculated as the reciprocal of the radius of a circle, whose arc fitted the coordinates of the movement sensors positioned at -0.4 m, -0.8 m and -1.2 m from the centre of the stem, at the time of soil failure. The authors were interested in the relationship between soilroot plate curvature and the resistive turning moment, and therefore to remove the effect of tree size (increased stem and root plate mass) on turning moment (Fraser and Gardiner, 1967) the resistive turning moment was divided by stem mass. Figure 4 shows a significant negative relationship between the normalized turning moment at soil failure plotted against curvature, although points had a large scatter about the regression line (r2 = 0.20). This regression shows that the more rigid the soil-root plate, the greater the applied turning moment at soil failFigure 3. (a) Soil-root plate displacement during tree- ure. Curvature was not significantly correlated pulling on one example tree, lines represent the plate with soil-root plate thickness, or rooting depth position at 1.5 s intervals ( • •) (P = 0.03). The mean turning moment at soil (b) The turning ( • • ) moment and change in failure for all of the test trees was 35 kNm (45.3 the turning moment (dM/df) ( • A) applied to Nm kg"1 normalized by dividing by stem the same tree. assumption is made in this study, that the peaks in dM/dt represent in sequence: soil failure, failure of the roots on the furrow side (windward roots), and hinge resistance. Tree pulling measurements were completed on 34 of the soil-root plates (five more trees were removed due to technical problems with the movement sensors). Thirteen of the study trees showed signs of pumping and were removed from the curvature analysis described below as it would not be possible to examine their curvature at the time of soil failure. From the data it was possible to find the applied turning moment and displacement of the plate at the point of soil failure represented by the first peak in dM/df (Figure 3b). For each tree calculations were made of when the soil — to • B i ' 8 -a B 0.02 0.0J 0.04 0.05 0M 0.07 0JM OJJ» 8otl-root ptata cunratar* (1An) Figure 4. Normalized turning (TM/stem weight) at soil failure against soil-root plate curvature at the moment of soil failure ( • , linear regression ). 176 FORESTRY —1 0.2! 1 0.J0 1 1 1 I 049 0.40 0.49 0.90 MMti wlntar wtt«r tab!* diplh (m) I 1 0.99 0.80 Figure 5. Normalized turning moment (TM/stem weight) at soil failure against mean winter water-table depth ( • , linear regression ). weight). This was about half of the mean maximum turning moment of 70 kNm for all of the test trees (88 Nm kg"1 normalized). Figure 5 shows the relationship between the normalized turning moment at soil failure, and mean winter water-table depth, for the group of pulled trees that had not been pumping. Again there was considerable scatter about the regression line (r2 = 0.21), but Figure 5 does show a significant trend for trees growing in drier soil above a deeper water table to be more resistant to soil failure than trees growing in more waterlogged conditions. The relationship between the maximum turning moment, normalized by dividing by stem weight, and water-table depth was tested using the complete data set of 34 trees. The data exhibited a large amount of scatter about the regression line (r2 = 0.2), which was significant, showing that as water-table depth increased from 0.25 m to 0.55 m, the average normalized turning moment increased from approximately 60 Nm kg"1 to 90 Nm kg"1. Figure 6 shows regression lines fitted separately according to soil type. There was no significant difference between the deep and shallow phase peaty gley regressions, and when combined (as in Figure 6) the coefficient of regression was significant, showing that resistance to overturning increased significantly with water-table depth, from about 60 Nm kg"1 to 90 Nm kg"1 with a lowering of the water table from 0.3 m to 0.6 m. The relationship between resistive turning moment and water-table depth on surface-water gley soil appeared to show a good relationship, but the coefficient of regression was not significant, pos- i O20 OJB i i i i i 0J0 OJS 0.40 0.48 0.50 Mean wfcrtor water tabto depth (m) r i 0.66 000 Figure 6. Maximum normalized turning moment (max TM/stem weight) against mean winter watertable depth on peaty gley (if, linear regression line ) and surface-water gley ( • , linear regression ) soils. WATER-TABLE EFFECT ON STABILITY OF SITKA SPRUCE 177 040 — 0.79 0.70 049 £ 040 - ? 040 - £ c 0.49 - * 0.40 - • . ••'1:1 lint ojg OJO ojg OJO r i OJI OJO i i i i 049 0.40 0.49 040 M M D wlntar water U b k depth ( • ) i i 049 040 Figure 7. Rooting depth plotted against mean winter water-table depth ( • ) , showing the linear regression line ( ), one-to-one line ( ), and estimated relationship of soil-root plate depth ). sibly because of the small sample size (seven trees). Soil-root plate characteristics The centrally weighted rooting depth (mean thickness of soil-root plate along eight cardinal radii, measured at 0.5 m and 1.0 m from centre of the stem) was significantly correlated (Figure 7) with the mean winter water-table depth (P < 0.001). The linear regression line was above the 1:1 line, indicating that roots show a degree of tolerance to waterlogging in the winter months. Table 2 shows a summary of the data, arranged into three categories of peat thickness: surface water gleys—peat thickness less than 0.10 m, shallow phase peaty gleys—peat thickness 0.10-0.25 m and deep phase peaty gleys— peat thickness 0.25-0.45 m, and the least significant difference between means. Soil type varied with site slope angle, surface-water gleys occurred on the steeper slopes and deeper phase peaty gleys on gentle slopes and flat sites. The maximum resistive turning moment increased with plate thickness, and was almost twice as large for plates on surface-water gley soils (85 kNm) compared with plates on deeper phase peaty gleys (48.4 kNm), however, this difference was partly due to larger trees (stem mass—936 kg) with heavier soil-root plates (5.77 tonne) on the surface-water gley, compared with the deep phase peaty gley (stem mass—662 kg, soil-root plate mass—3.53 tonne). After normalizing the maximum resistive turning moment by dividing by stem weight, differences between soil types were not significant. Surface-water gley soil-root plates were thicker because of the deeper winter water table, and were heavier because they contained a greater volume of mineral soil with a larger wet bulk density than plates on peaty gley soils. Trees grew significantly taller on the surfacewater gley and the shallow phase peaty gley than on the deeper phase peaty gley (Table 2). Soil strength Mean soil strength, from the hand vane measurements, increased significantly in successively deeper soil horizons (Table 3). The mean strength of soil on the B horizon (51 kPa) was almost twice the strength of the A horizon material (27.2 kPa) and the peat (18.3 kPa). Discussion The relationship between the resistive turning moment and water-table depth, and between the resistive turning moment at soil failure and 178 FORESTRY Table 2: Summary of tree and root plate data sorted by soil type Factor/Soil type All trees Trees on deep phase peaty gley. Peat thickness >0.25<0.45 m (a) Trees on shallow phase peaty gley. Peat thickness >0.10<0.25 m Trees on surface-water gley. Peat thickness <0.10 m LSD , P £ 0.05 (1) (a)-(b) (2) (bHO (3) (aHc) (c) (b) 62.4 48.4 64.6 85.0 Max turning moment/ stem mass (Nm kg"1) 81.0 76.4 80.7 90.8 Water-table depth (m) 0.43 0.395 0.432 0.467 Site slope (°) 4.2 2.9 4.9 5.3 Rooting depth (m) 0.55 0.48 0.58 0.63 Surface area (m2) 6.13 6.53 5.93 5.84 Soil volume (m3) 3.35 3.10 3.42 3.71 Soil mass (tonne) 4.48 3.53 4.75 5.77 Hinge distance (m) 0.922 0.991 0.905 0.823 Tree height (m) 24.23 22.79 25.18 24.77 Diameter at breast height (m) 0.32 0.30 0.32 0.35 Crown width (m) 6.87 6.48 7.00 7.33 Crown depth (m) 10.29 10.24 10.77 9.6 Stem mass (kg) 770 634 801 936 M a x turning moment (kNm) *, differences significant at P £ 0.05; n.s., not significant. 16.0* 20.0* 20.6* 17.4 n.s. 21.7 n.s. 22.4 n.s. 0.08 n.s. 0.09 n.s. 0.10 n.s. 0.1 n.s. 1.1 n.s. 1.2* 0.07* 0.10 n.s. 0.10* 1.38 n.s. 1.73 n.s. 1.77 n.s. 0.89 n.s. 1.12 n.s. 1.14 n.s. 1.36 n.s. 1.69 n.s. 1.75* 0.15 n.s. 0.19 n.s. 0.19 n.s. 1.48* 1.83 n.s. 1.89* 0.03 n.s. 0.05 n.s. 0.05 n.s. 1.0 n.s. (2) 1.22 n.s. (3) 1.26 n.s. (1) 1.48 n.s. (2) 1.85 n.s. (3) 1.25 n.s. (1) 179* (2) 223 n.s. (3) 230* (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) (2) (3) (1) WATER-TABLE EFFECT ON STABILITY OF SITKA SPRUCE water-table depth, is fundamentally important to the stability of Sitka spruce planted on wet soils. Courts (1986) found that soil failure occurred after 0.002-0.005 m displacement at a mean turning moment of 6.3 kNm in plates pulled monotonically, whereas during cyclic loading, O'Sullivan and Ritchie (1993) found the soil-root plate was displaced on average 0.007 m at soil failure. Their displacement results were slightly less than those found in the present study, which showed a mean displacement of 0.010 m at soil failure. The estimated mean turning moment at soil failure in the present study was considerably larger than reported by Coutts (1986), and there are several possible reasons for this difference. First, there were differences in soil wetness between the studies; whereas the soil was quite dry at the time of pulling in this study, soil conditions were wet in Coutts' (1986) study. Second, in this study soil-root plates were larger; the trees were 10 years older than those in Coutts' study. Third, at the point of soil failure the elastic limit of the soil was exceeded. Coutts found a failure sequence beginning at a point close to the base of the tree, and spreading in the form of cracks mainly towards the windward side of the tree. There is likely to be some lag in the response of the soil-root plate to an applied turning moment, because the energy in the applied load is transmitted and dissipated first through the wood tissue of the stem and roots and then through the soil. In the present study soil-root plates resisted overturning in a dynamic way; the resistive turning moment responded to the continuous application of a load and, unlike the root plates in Coutts' (1986) study, were not given time to equilibrate with the applied turning moment while it was held constant in order to read instruments. Consequently, because the method in this study differed from earlier 'monotonic' tree pulling studies, the instruments measured a bigger load at soil failure than reported in earlier studies. All of the soil-root plates in the study sheared close to the maximum depth of the root system, some fine roots protruded less than 0.04 m from the base of the soil-root plate and all roots with secondary thickening were contained above the shear plane (Nicoll and Ray, 1996). Rooting depth corresponded with, but was slightly deeper than, the winter water table. It is likely that there is some minimum threshold for root plate thickness irrespective of winter water-table depth. If so the data might be distributed parallel to the 1:1 line for deeper water tables but would bend towards a threshold root plate thickness as shown in Figure 7. This study confirmed the assumption that root systems would develop into and fully exploit soil as the water table dropped after drains were installed on the site. Some plates sheared within the peat, others within the mineral soil; Anderson et al. (1989) also reported a similar soil-root plate shear zone pattern for Sitka spruce root plates. Many of the thinner plates on the deeper phase peaty gley soil sheared within the Eg horizon. This is an eluviated horizon from which the fine colloidal clay particles have been removed during some stage in the soil formation process, leaving a material of more sandy texture than the deeper clayey mineral horizons. Shear strength is commonly measured in place of soil tensile strength and although soil wetness was not taken into consideration, the data show that the Eg was somewhat weaker than the underlying Bg horizon (Table 3). In contrast, Anderson et al. (1989), as well as reporting smaller values of soil shear strength at this site (10-25 kPa) than in this study (18-51 kPa), found no difference in Table 3: Mean soil shear strength (kPa) measured with a hand vane tester and least significant difference (LSD) at P < 0.05. Soil horizon Soil shear strength (kPa) LSD (between value and one below) Peat A 18.3 27.2 42.6 51.0 5.6* 5.1* 7.7* *, difference significant at P < 0.05. 179 180 FORESTRY the soil shear strength between peat and mineral horizons. However, they did report an increase in the normal stress of a soil-root plate as the proportion of mineral-peat soil increased in the shear zone. From Anderson et al. (1989) and from the present data, it appears that an improvement in stability should result from deeper winter water tables, and hence deeper root plates, because the base of the plate will contain soil of higher shear strength, and because a larger volume of soil of higher bulk density will be incorporated within the plate. The effect of soil wetness on soil strength must also be considered in predicting stability, since soil strength is inversely related to soil water content. This will have the effect of causing the soil strength component of tree stability to change seasonally, with trees most vulnerable during winter. Soil failure is an important threshold event after which some trees begin to 'pump', before overturning in subsequent high winds. Certainly soil failure reduces the bending moment that a tree can resist (O'Sullivan and Ritchie, 1993), and Courts (1986) suggested that after soil failure a small space will form below a plate because it will not return completely to its former position once the force is relaxed. In wet soils, such gaps will quickly fill with water, and soil surrounding these saturated areas is probably more likely to liquefy, because tree sway creates an oscillating pressure which will be effectively transmitted through the increased volume of water-filled pores and cavities. Such conditions appear to be required for liquefaction failure of the soil (Rodgers et al., 1995) and the onset of pumping. Some trees may however recover from this destabilization if warmer and drier conditions, which stimulate active root growth, follow the period of soil disturbance. The repeated rocking of the plate in its cavity, represents a considerably less stable condition for the tree; wind energy is dissipated in the rocking motion rather than in stem bending, and the repeated stress of rocking may progressively weaken windward and leeward roots, thus further reducing anchorage (O'Sullivan and Ritchie, 1993). These trees are therefore vulnerable to windthrow at lower wind speeds with a shorter return period (Quine, 1994). The hypothesis that soil-root plate stiffness is an important contributor to tree stability (Coutts, 1983b; Deans and Ford, 1983), is confirmed by the significant negative relationship between soil-root plate curvature and the turning moment at soil failure (Figure 4). It may be expected that soil-root plate thickness would be the most important factor affecting plate stiffness. However, the regression of soil-root plate curvature at soil failure on soil-root plate thickness was not significant, possibly due to the complicating additional effect of structural root development. There was a significant negative relationship between root:shoot ratio and water table depth (Figure 2). This indicates that trees with thin soil-root plates growing on shallow water-tables either invest more biomass in woody structural root systems than deeper rooted trees, or that shallow rooted trees have a lower nutrient budget, are less productive and therefore support smaller crowns. In a followon study, root systems of trees in this study were found to have developed cross sectional shapes that would resist bending, with the shallowest roots having the greatest 'adaptive' development (Nicoll and Ray, 1996). Roots had developed I-beam and T-beam shapes (as described by Rigg and Harrar, 1931), that would have enhanced the rigidity of the shallowest soil-root plates (Nicoll and Ray, 1996). The large variation between tree characteristics and turning moments in this study (seen as a wide scatter around regression lines) results both from differences in soil conditions, even between neighbouring trees, as well as genetic variation in tree morphology. Nicoll et al., (1995) described substantial variation in root:shoot ratio and structural root system architecture between shallowly rooted Sitka spruce clones, and suggested that some clones would be particularly vulnerable to windthrow. In an adjacent experiment, Ray et al. (1992) showed that drains spaced closer than 40 m had no effect in lowering the water table on steeper sloping surface-water gley soils, but on the gently sloping peaty gley soil drains spaced at 10 m produced a small (0.07-0.15 m), but significant, lowering of the water table compared with drains at 40 m. The data presented here also show that drainage can promote deeper water tables and allow deeper soil-root plates to develop, even though drains were excavated WATER-TABLE EFFECT ON STABILITY OF SITKA SPRUCE when the crop was 17 years old. A conservative estimate of the improvement in stability resulting from deeper water tables from Figure 6 is 1.5 Nm kg"1 increase in the resistive turning moment for every 0.01 m lowering of the water table. Therefore, for a 47-year-old Sitka spruce • tree growing on a peaty gley soil, 24 m high with a diameter at breast height of 0.31 m and a mean stem weight of 750 kg, intensive drainage will improve the rooting depth by about 0.10 m, increasing the resistive turning moment (the tree stability) by 20 per cent from 50 kNm to 60 kNm. The authors aim to continue this analysis to investigate if the increase in the resistive turning moment of trees in drained stands will significantly improve the rotation length of the crop. This will permit a cost-benefit analysis to examine the economic benefits of intensive drainage on peaty gleys. 181 Coutts, M.P. 1983a Development of the structural root system of Sitka spruce. Forestry 56, 1-16. Coutts, M.P. 1983b Root architecture and tree stability. Plant Soil 71, 171-188. Coutts, M.P. 1986 Components of tree stability in Sitka spruce on a peaty gley soil. Forestry 59, 173-197. Coutts, M.P. 1987 Developmental processes in tree root systems. Can. ]. For. Res. 17, 761-767. Coutts, M.P. and Philipson, J.J. 1978 Tolerance of tree roots to waterlogging. I. Survival of Sitka spruce and lodgepole pine. New Phytol. 80, 63-69. Coutts, M.P. and Philipson, J.J. 1987 Structure and physiology of Sitka spruce roots. Proc. Royal Soc. Ed. 93B, 131-144. Deans, J.D. 1981 Dynamics of coarse root production in a young plantation of Picea sitchensis. Forestry 54, 139-155. Deans, J.D. and Ford, E.D. 1983 Modelling root structure and stability. Plant Soil 71, 189-195. Forestry Commission 1993 Forests and Water Guidelines, 3rd edn. HMSO, London. Fraser, A.I. and Gardiner, J.B.H. 1967 Rooting and stability in Sitka spruce. Forestry Commission Bulletin No 40. HMSO, London. Acknowledgements Hendrick, E. 1989 The effect of cultivation method on the growth and root anchorage of Sitka spruce. We wish to thank the following for considerable help Ir. For. 46, 19-28. during the fieldwork phase: David Clark, staff at the Jarvis, R.A., Bendelow, V.C., Bradley, R.I., Carroll, Kielder TSU Field-station, the NRS Workshop, Forest D.M., Furness, R.R., Kilgour, I.N.L. and King, S.J. Enterprise at Kielder Forest District, the Forest 1984 Soils and their use in northern England. Soil Enterprise Mechanical Engineers at Kielder and Survey Bull. No 10. Soil Survey of England and Dumfries, and the Bush TSU Field-station. In addiWales, Harpenden. tion, thanks to the six students involved in this study: Simon Parkin, Andrew Cameron, Elaine Washington, King, J.A., Smith, K.A. and Pyatt, D.G. 1986 Water and oxygen regimes under conifer plantations and Tobias Kerzenmacher, Mairi Nichol and Alice native vegetation on upland peaty gley soil and Hague. Thanks also to Alvin Milner for statistical deep peat soils. ; . Soil Set. 37, 485-497. advice, and to Barry Gardiner for his critical Muir, J.W. 1956 The soils of the country round appraisal of an earlier draft. Jcdburgh and Morebattle—Sheets 17 and 18. Memoirs of the Soil Survey of Great Britain. Department of Agriculture for Scotland, HMSO, References Edinburgh. Anderson, C.J., Campbell, D.J., Ritchie, R.M. and Nicoll, B.C. and Ray, D. 1996 Adaptive growth of Smith, D.L.O. 1989 Soil shear strength measuretree root systems in response to wind action and site conditions. Tree Physiol. 16, 891-898. ments and their relevance to windthrow in Sitka spruce. Soil Use Manage. 5, 62-66. Nicoll, B.C., Easton, E.P., Milner, A.D., Walker, C. and Coutts, M.P. 1995 Wind stability factors in Anderson, A.R., Pyart, D.G. and Stannard, J.P. 1990 tree selection: distribution of biomass within root The effects of clearfelling a Sitka spruce stand on systems of Sitka spruce clones. In Wind and Trees. the water balance of a peaty gley soil at Kershope M.P. Coutts and J. Grace (eds). Cambridge forest, Cumbria. Forestry 63, 51-71. University Press, Cambridge. Avery, B.W. 1990 Soils of the British Isles. CAB International, Wallingford. O'Sullivan, M.F. and Ritchie, R.M. 1993 Tree stability in relation to cyclic loading. Forestry 66, 69—82. Binns, W.O. 1962 Some aspects of peat as a substrate for tree growth. Ir. For. 19, 32-55. Peterken, G.F., Ausherman, D., Buchenau, M. and Foreman, R.T.T. 1992 Old-growth conservation Boggie, R. 1977 Water-table depth and oxygen conwithin upland conifer plantations. Forestry 65, tent of deep peat in relation to root growth of 127-144. Pinus contorts. Plant Soil 48, 447-454. 182 FORESTRY Pyatt, D.G. 1990 Forest drainage. Forestry Commission Research Information Note No 196. Farnham, Surrey. Pyatt, D.G. and Booth, T.C. 1973 Crop stability: tree pulling. Forestry Commission Report on Forest Research 1973. HMSO, London, 72-74. Pyatt, D.G. and John, A.L. 1989 Modelling volume changes in peat under conifer plantations. J. Sod Sci. 40, 695-706. Pyatt, D.G. and Smith, K.A. 1983 Water and oxygen regimes of four soil types at Newcastleton forest, south Scotland. J. Soil Sci. 34, 465-482. Pyatt, D.G., Anderson, A.R., Stannard, J.P. and White, I.M.S. 1985 A drainage experiment on a pcaty-gley soil at Kershope forest, Cumbria. Soil Use Manage. 1, 89-94. Quine, C.P. 1994 An improved understanding of windthrow—moving from hazard towards risk. Forestry Commission Research Information Note No 257, Farnham, Surrey. Quine, C.P. and White, I.M.S. 1993 Revised windthrow hazard classification. Forestry Commission Research Information Note No 230. Forestry Commission, Farnham, Surrey. Quine, C.P., Burnand, A.C., Coutts, M.P. and Reynard, B.R. 1991 Effects of mounds and stumps on the root architecture of Sitlca spruce on a peaty gley restocking site. Forestry 64, 385—401. Quine, C.P., Coutts, M.P., Gardiner, B.A. and Pyatt, D.G. 1995 Forest and wind: management to minimise damage. Forestry Commission Bulletin No 114. HMSO, London. Ratcliffe, P.R. 1993 Biodiversity in Britain's Forests. Forestry Authority, Forestry Commission, Edinburgh. Ray, D. and Anderson, A.R. 1990 Soil temperature regimes of mounds on gley soils. Forestry Commission Research Information Note No 168, Forestry Commission Research Division, Farnham. Ray, D., White, I.M.S. and Pyatt, D.G.P. 1992 The effect of ditches, slope and peat thickness on the water regime of a forested gley soil. Soil Use Manage. 8, 105-111. Rigg, G.B. and Harrar, E.S. 1931 The root systems of trees growing in sphagnum. Am. ]. Bot. 18, 391-397. Rodgers, M., Casey, A., McMenamin, C. and Hendrick, E. 1995 An experimental investigation of the effects of dynamic leading on coniferous trees planted on wet mineral soils. In Wind and Trees. M.P. Coutts and J. Grace (eds). Cambridge University Press, Cambridge. Savill, P.S. 1976 The effects of drainage and ploughing of surface water gleys on rooting and windthrow of Sitka spruce in Northern Ireland. Forestry 49, 133-141. Tabbush, P.M. and Ray, D. 1989 Effects of rough handling and microsite on the establishment of Sitka spruce on clearfelled sites in Britain. Forestry 62 (Supplement), 289-296. Taylor, G.G.M. 1970 Ploughing practice in the Forestry Commission. Forestry Commission Forest Record No 73, HMSO, London. Wagg, J.W.B. 1967 Origin and Development of White Spruce Root Forms. Canada Department of Forestry and Rural Development, Forestry Branch Departmental Publication No 1192. Received 8 October 1996
© Copyright 2025 Paperzz