Forestry Advance Access originally published online on January 27, 2009
Forestry 2009 82(2):199-210; doi:10.1093/forestry/cpn054
| ||||||||||||||||||||||||||||||||||||||||||||||||||
Side shelter on lowland sites can benefit early growth of ash (Fraxinus excelsior L.) and sycamore (Acer pseudoplatanus L.)
Forest Research, Forestry Commission, Alice Holt Lodge, Farnham, Surrey, GU10 4LH, UK
* Corresponding author. E-mail: ian.willoughby{at}forestry.gsi.gov.uk
 |
Summary |
|---|
 Top Summary IntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
The influence of side shelter on the growth of newly planted ash (Fraxinus excelsior L.) and sycamore (Acer pseudoplatanus L.) was investigated at a lowland site in Britain. Although tatter flag analysis classified the site as ‘very sheltered’ in comparison to upland sites, after three years, both species benefited to some degree from the provision of shelter, with height increment being improved by up to two to four times. The most effective shelter was provided by a wall of straw bales. Plastic netting also reduced exposure, but required regular maintenance in winter months. Our work suggests that on equivalent sites in southern Britain, where mean daily tatter rates of more than 2.4 cm2 day–1 (equivalent to windiness scores of greater than 10.3) occur, without side shelter, severe suppression of height growth in ash and sycamore is likely, and that exposure to wind is likely to be a significant cause of this reduction in growth. Currently, using artificial shelter materials on a large scale on lowland sites solely to improve early tree growth is unlikely to be cost-effective in most cases. However, the provision of side shelter, particularly through use of nurse species, is likely to become an increasingly important silvicultural consideration in the future.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
In a forestry context, the term shelter can be used to describe the protection of trees from exposure to potentially damaging environmental factors, in particular wind, but also frost, sun and snow. The specific effects of shelter will vary depending on the climatological and topographical characteristics of any particular site, as well the characteristics of the protection itself. Although the main benefit of shelter is often considered to be a reduction in wind velocity, this can also be accompanied by significant changes in air temperature, solar radiation, turbulence, relative humidity, soil moisture and soil temperature. Provision of shelter will usually reduce the mixing of air, increase air temperatures around plants on sunny days and possibly reduce night-time temperatures (Gardiner et al., 2006
).
Exposure can affect tree physiology in several ways. Firstly, wind may cause direct mechanical damage to the leaves and result in branch and stem snap, and windthrow (Quine et al., 1995). Wind can also cause stunting of plant growth and poor form (Kozlowski et al., 1991), increased biomass allocation to roots at the expense of shoots (Ritson and Sochacki, 2003) and changes in the biomechanical properties of the xylem leading to reduced timber quality (Kozlowski et al., 1991). The high evaporative demand in exposed sites also puts plants at increased risk of dehydration (Heiligmann and Schneider, 1974). A plant's response is often to close the stomata, triggered either by the physical shaking of the leaves (Kozlowski et al., 1991) or by tissue dehydration (Davies et al., 1974), which results in decreased daytime evapotranspiration rates and controlled water loss (Dixon and Grace, 1984). As stomata close, the rate of photosynthesis reduces (Kozlowski et al., 1991), and hence growth of exposed plants can be reduced.
Given these physiological effects, it is perhaps unsurprising that silvicultural texts often advise that the provision of side shelter can benefit tree establishment, particularly with broadleaved species (Evans, 1984; Kerr and Evans, 1993). Methods used by managers to provide shelter to newly planted trees include encouraging natural infill from tree and shrub regeneration (Evans, 1984), establishing mixtures of broadleaves and conifers (Kerr et al., 1992; Olsthorn et al., 1999; Horgan, 2004) or the use of individual tree shelters (Potter, 1991).
Although the effects of exposure and wind on establishment (Lines and Howell, 1963; Rees and Grace, 1980; Thompson, 1984) and long-term growth and stability (Miller et al., 1987; Quine and White, 1993; Quine et al., 1995) have been extensively studied in the uplands of Britain, by comparison there appears to have been little objective work carried out on the benefits of side shelter for newly planted trees on the warmer, drier, lowlands (<250 m) of southern Britain. Therefore, in the work reported here, the aim was to quantify the extent to which two broadleaved species could benefit from side shelter in what, by comparison with upland sites, might be generally thought to be a relatively sheltered lowland setting. A further aim was to investigate the practicality of using a range of potential shelter materials to help improve tree growth and form.
|
Materials and methods |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
The experiment took place within a deer- and rabbit-proof fenced enclosure within Thetford Forest, Norfolk, UK (52° 20' N, 0° 42' E), which receives an annual average of 600 mm of rainfall, has an average of 1761 growing degree-days above 5°C and an annual average soil moisture deficit of 216 mm (data from Pyatt et al., 2001
). Soil type according to Mackney et al. (1983) is a typical well-drained brown calcareous sand, in the Methwold association. The site is 50 m above sea level, has a slope of <5° and a southerly aspect. Using the Detailed Aspect Method of Scoring (Quine and White, 1993) for measuring the exposure of a site gave a topex (topographical exposure) value of 8 and a total windiness score of 9.16, which would equate to a Windthrow Hazard Classification (Miller, 1985) of 2. An area of mature Corsican pine (Pinus nigra ssp. laricio Maire) was clearfelled in 1995 to leave 
9 ha of open ground, which prior to the establishment of the experiment was predominantly covered in low growing mixed grasses, along with some areas of bracken (Pteridium aquilinum (L.) Kuhn) and bramble (Rubus fruticosus L., agg.). The experiment was located within this felled area such that there were no mature trees within 200 m of any aspect, except for a stand of mature beech (Fagus sylvatica L.) between 10 and 50 m from the southern and eastern boundaries of the experiment. In addition, a row of dead stumps and brash piled 1.5–2 m high was located 40 m to the west of the experiment. The experiment consisted of three rectangular blocks, 68 m x 3m, aligned in an approximate north–south direction, 25 m apart—see Figure 1. Each block consisted of five randomly assigned shelter plots, each of which was further randomly split into two tree species sub-plots. This gave three replicates of five shelter treatments per species, making 30 sub-plots in total. Each sub-plot was planted with two rows of trees at 1 m x 1 m spacing, giving 10 trees per sub-plot. Each shelter plot was 12 m x 3 m, with a 2-m buffer between treatment plots. Shelter treatments were established immediately after planting trees such that they completely surrounded one plot (i.e. two species sub-plots), at a distance of 1 m from the outermost trees. The following shelter treatments were used:
- S0: Control, no shelter.
- S1: Shelter provided by one layer of 50 per cent porosity ‘Rokolene’ plastic windbreak netting, 1.8 m high (manufactured by Growing Technologies, Derby, UK). Netting was supported by a top-wire slung between stakes and the stakes were held upright using tie-backs.
- S2: Shelter provided by two layers of 50 per cent porosity ‘Rokolene’ plastic windbreak netting, separated by a gap of 40 cm, 1.8 m high, supported as described above.
- S3: Living shelter provided by poplar (Populus sp.) trees, using 25 cm cuttings planted at 1m spacing in February 1996, at the same time as the experimental seedlings.
- S4: Shelter provided by a solid wall of straw bales piled 1.8 m high by 0.4 m wide, supported by stakes to increase stability.
- S1: Shelter provided by one layer of 50 per cent porosity ‘Rokolene’ plastic windbreak netting, 1.8 m high (manufactured by Growing Technologies, Derby, UK). Netting was supported by a top-wire slung between stakes and the stakes were held upright using tie-backs.
). Two-year-old nursery transplants were planted in February 1996, and all trees protected by 20-cm-tall plastic vole guards (Tubex, Aberdare, UK). All plots were kept completely weed-free throughout the life of the experiment through repeat applications of the herbicides glyphosate and propyzamide, following the techniques outlined in Willoughby and Dewar (1995).
|
The experiment was maintained for three years. Annual winter assessments of tree height, stem root collar diameter and angle of lean (the angle between an hypothetical straight line joining the tip of the tree with its base and vertical) were made on all trees. In addition, the survival, height and diameter of the poplar cuttings used to form the shelter in treatment S3 and the height of the straw bales (treatment S4) were assessed in March 1999 after three years of exposure.
A tatter flag was positioned on a post 1.5 m above ground level in the centre of each shelter plot to assess windiness and relative exposure. The installation and monitoring of the 15 flags was carried out as described by Mackie and Gough (1994). Flags were changed at approximately two-month intervals for a period of three years. Monthly mean wind speed and direction was calculated from records taken daily at 9 am at the Thetford Forest main meteorological recording station situated 15 km to the North of the site (52° 27' N, 0° 40' E).
The effects of the shelter treatments on height and diameter increment, angle of tree lean and mean daily tatter rate were analysed using analysis of variance (Genstat, 1993), and standard errors of differences of means were generated. Least significant difference or T tests were then carried out at the P 
0.05 level, again using Genstat. Species were analysed separately, and combined. Fisher's exact test was used to analyse tree survival (Genstat, 1993), as few plots suffered mortality, and on those that did there was generally a low number of dead trees, which meant that analysis using a more elaborate parametric modelling approach, such as generalized linear modelling, was inappropriate. The relationship between height increment and mean daily tatter rate was investigated using a mixed model with blocks as a random effect. This is equivalent to fitting mean daily tatter rates as a covariate in an analysis of variance without including the treatment term. The mixed models were fitted by restricted maximum likelihood analysis (REML) using Genstat. The R2 values were calculated using a likelihood ratio approach where R2 = 100(1 – exp(dev – dev0)/n)), ‘dev’ being the deviance from the full model and ‘dev0’ the deviance from the constant only model but using the same projection as the full model (Welham and Thompson, 1997).
|
Results |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
Shelter effects
Initial problems of sagging of the plastic netting were solved by stretching wire between the supporting stakes, however, gales caused damage that required replacement netting and additional staking to be used on three occasions during the life of the experiment. The straw bale treatment required less maintenance, although some damage resulting from gales in February 1997 needed to be repaired. Mean bale height after three years of exposure was 1.5 m, a 17 per cent loss in height over the three-year life of the experiment. Growth of the poplar cuttings was very poor—after three years the average survival was only 34 per cent, with an average height of 1.4 m, although this varied considerably between blocks.
Mean daily tatter rate (cm2 day–1) varied significantly (P < 0.001) between shelter treatments (Table 1). The highest mean daily tatter rates were seen in the unprotected control plots and the plots surrounded by poplar cuttings, which were not significantly different from each other. A single layer of plastic netting provided significantly more shelter than the unprotected control, and adding a second layer further significantly improved the effect. However, the straw bales provided the best shelter with the lowest mean daily tatter rate of 1.8 cm2 day–1, significantly lower than that of double layer of plastic netting. Overall, using tatter flag data, the site could be classified according to Miller et al. (1987) as ‘very sheltered’, with all mean daily tatter rates below 4.0 cm2 day–1.
|
Data for monthly mean wind speed and direction (Table 2 and Figure 2) show that the prevailing winds in the area tended to be from the south to south east, except in June to August, when prevailing winds were from the south west. Exposure was generally lowest between November and January period, but remained reasonably consistent during the remainder of the year.
|
|
Growth effects
Looking at the experiment as a whole, the shelter treatments had a significant effect on tree growth, but there was no significant interaction between species, indicating that both ash and sycamore responded to the individual treatments in a similar way. The greatest growth occurred in the most sheltered plots. Some wind snapping of leaders occurred, which is reflected in the increment data. Overall, angle of lean was generally less in the straw bale and plastic netting treatments, but there were no significant differences on an individual species basis.
When the species were analysed separately, for ash the analysis of variance indicated there was no overall significant difference in tree growth between treatments (Table 3). However, three-year height increment in the straw bale treatment was four times that of the control plots, and a T test indicted a significant difference for this specific contrast of what, at the outset of the experiment, were anticipated as being the most and least sheltering materials (i.e. an a priori test). There were no significant differences in diameter increment, although growth in the straw bale treatment was apparently twice that of the control plots. There were no significant differences in angle of lean or survival between treatments.
|
With sycamore there was no difference in three-year height or diameter increment between the straw bale, double or single layer of plastic netting treatments, but all three of these shelter types resulted in significantly larger height and diameter increments than the living shelter or control plots (Table 3). There were no significant differences in angle of lean or survival between treatments.
Relationship between growth and exposure
REML analysis indicated good evidence of a linear relationship between exposure as measured by mean daily tatter rate and three-year height increment of sycamore. Figure 3 gives the best fit model which accounts for 60.9 per cent of the variation in sycamore height increment after three years where:
 |
The relationship was less clear for ash. Figure 3 gives the best fit model which accounts for 27.1 per cent of the variation in ash height increment after three years where:
 |
|
|
Discussion |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
Growth effects
Tatter flags were used in our work as numerous experiments over the last 50 years have shown them to be a simple, robust and inexpensive method of assessing relative levels of exposure (Quine and White, 1993; Mackie and Gough, 1994). A close correlation has been demonstrated between tatter rates and total wind speed as measured by anemometers (Miller et al., 1987; Quine and White, 1993), and also with the growth of trees (Thompson, 1984). Other environmental factors, particularly rainfall, can also affect tattering to a lesser extent (Rutter, 1967), although Savill (1974) suggested that this might actually make the use of tatter flags a better measure of exposure than a simple assessment of wind speed. Although it would have been preferable, if possible, to have recorded micro-meteorological factors such as wind speed, temperature, humidity and solar radiation to determine the precise nature of the growth response to the provision of shelter, there is nevertheless good precedent for the use of tatter flags as giving a good indication of site exposure.
Given a total windiness score (Quine and White, 1993) of only 9.16, in comparison with many British upland sites, the risk of wind damage to trees on our experimental site would normally be considered very low. With mean daily tatter rates in the control plots of 2.19 cm2 day–1, the site is classified as ‘very sheltered’ according to Miller et al. (1987). However, despite the sheltered nature of the site, both ash and sycamore benefited from the provision of additional side shelter. Besides the reduction in wind exposure, the sheltered plots are also likely to have benefited from an increase in air temperature and growing season length, and a reduction in solar radiation (Heiligmann and Schneider, 1975), although the relationship between these factors was not tested specifically in our work. Although both species responded similarly to the different treatments, the actual magnitude of the growth response varied. Those ash trees growing in the straw bale plots, the most effective shelter treatment, put on four times as much height increment and twice as much diameter increment than those growing in the control plots. This adds weight to the view that young ash trees can require side shelter to achieve good early growth (Evans, 1984). Straw bales also provided the biggest growth response for sycamore, giving a 2.6 times improvement in height increment and a doubling in diameter increment compared with the control. Sycamore grew faster in the most exposed conditions, which fits with the assertion that the species is relatively tolerant of exposure (Evans, 1984).
As noted earlier, there appears to be little objective evidence in the literature regarding the effects of providing artificial side shelter on the growth of broadleaved trees in lowland Britain, despite this being a common silvicultural recommendation (Evans, 1984). However, in USA, Heiligmann and Schneider (1975), did report that, on a less exposed site than that tested in our work, average wind speeds were reduced by around 65 per cent by the provision of shelter, which resulted in a doubling of black walnut (Juglans nigra L.) shoot dry weight over an 18 week period. For conifers in the British uplands, Thompson (1984) reported that the provision of shelter on an exposed upland site reduced mean daily tatter rates from 7 cm2 day–1 to around 2.5 cm2 day–1, resulting in a trebling of Sitka spruce (Picea sitchensis (Bong.) Carr.) shoot dry weight over a five-year period. Our work suggests that tree growth would also be increasingly curtailed with greater exposure, in particular due to the effects of wind as measured by tatter flag degrade (Figure 3), in what might otherwise be assumed to be a relatively sheltered, lowland, Northern European context. Due to limited data, caution should be exercised in using our models to predict the exposure levels at which growth increment might be reduced to zero. However, it is fair to assume that, on similar sites with other conditions such as soil moisture availability being equal, mean daily tatter rates of more than 2.4 cm2 day–1 (equivalent to windiness scores of greater than 10.3) would be likely to result in a severe suppression of height growth in ash and sycamore. Although tree survival was not significantly reduced in our experiment, as exposure levels increase, so would direct mechanical damage to stems and branches (Quine et al., 1995), which may lead to increased mortality. In addition, slow-growing, wind-suppressed trees are likely to remain vulnerable for far longer to damage and death from a range of other biotic and abiotic factors (Willoughby et al., 2004).
Shelter effects
The amount of shelter achieved from a windbreak can be sensitive to its orientation (Wilson and Flesch, 2003; Gardiner et al., 2006), with the best protection usually provided when the shelter material is perpendicular to the wind direction. Prevailing winds in our experiment were from the south to south east, except during the middle of the growing season, when winds were from the south to south west. In both cases, prevailing winds were not perpendicular to the shelter screens (which were aligned north-south); however, given the clear shelter effect of some of the treatments, it is likely that the small size of the plots overcame any orientation effect that might be more apparent with larger windbreak areas.
Of the materials investigated, straw bales offered the most shelter, reducing mean daily tatter rates by 18 per cent and total windiness score by 6.3 per cent. Straw has the advantage of being a sustainable natural material, and in the case of afforestation situations, bales may be readily available on site—around 3–4 ha of ground would be required to produce sufficient straw for a 100-m run of a 1.8-m tall bale wall. Bales will biodegrade naturally and there is no need to collect and dispose of any waste material once trees are established. In our experiment, the bale wall was stable and required little maintenance over the three-year period, but the technique could be less practical in more exposed conditions. If straw bales need to be purchased, cost will depend on current market conditions and the weight of straw per cubic metre. Overall cost for a 100-m run of a 1.8-m tall wall, including straw, posts and labour, is likely to be in the region of £350 (£3.50 m–1). However, our experiment utilized relatively small plot sizes, where the trees received total, all round protection for the straw bales. As a non-porous material, a 1.8-m wall of straw bales is only likely to offer protection for trees up to 28 m from the barrier, with the maximum protection being obtained up to 9 m from the shelter (Gardiner et al., 2006). This would leave 70–90 per cent of a square 1-ha site unprotected. Additional straw bale walls could be constructed at intervals within the plantation, but this is unlikely to be economic. Hence, the use of straw bales as a shelter material may be best suited to situations where the material is already readily available on site, or for the establishment of narrow plantations, shelter belts and hedgerows on open, exposed lowland sites.
Porous plastic netting, of which Rokolene is but one proprietary example, is used widely in horticultural production to provide shade and shelter, to improve crop yields and quality (Baxter, 1986; R. Jinks, personal communication). In our experiments, the Rokolene treatments reduced mean daily tatter rates by 7 per cent and total windiness score by 3.7 per cent. Adding a second layer of material reduced mean daily tatter rates by 12 per cent and total windiness score by 6 per cent. Being a porous material, wind speed is likely to be reduced for a much greater distance (Gardiner et al., 2006), for a 1.8-m high wall perhaps up to 54 m may benefit, which for a square 1-ha plantation leaves around 45 per cent of the area unprotected. However, in our work Rokolene netting also required significant maintenance inputs to repair damage caused by gales, which may make it less practical for very exposed, isolated forestry locations. In these situations, one of the more robust plastic shelter materials on the market might be more appropriate. All conventional plastic netting is currently derived from non-sustainable hydrocarbons, may only remain intact for two to three years, and will require plastic waste to be collected and disposed of at the end of its useful life. A porous plastic netting such as Rokolene is likely to cost £400–600 for a 1.8-m high 100 m run (£4–6 m–1), including posts, construction and maintenance. Such an investment would be difficult to justify in forestry situations without additional evidence of long-term volume gains or reduced rotation lengths. However, where a fence is already being erected to protect against browsing damage, it could cost as little as £140 (£1.40 m–1) to add a 1.8-m high 100 m run of 60 per cent porosity plastic netting, including the cost of additional strengthening (Trout and Pepper, 2006) of supporting struts that would be required.
Where individual tree shelters are used instead of fencing to prevent mammal browsing, these can in theory also offer some protection from wind. However, they are not ideal for very exposed sites as shelters and stakes can degrade and break, and trees can become stunted and abraded when they emerge from the protective environment of the shelter (Nixon, 1994). Tree shelters can offer useful protection on less exposed lowland sites, on smaller scale plantings where fencing is not economic.
The site conditions in our work were probably not optimal for the poplar clone used in treatment S3, which may account the lack of shelter it afforded. Although some poplar clones can grow very rapidly and offer good shelter (Baxter, 1986), the best growth will only often be realized on the most fertile, well-drained alluvial soils. White poplar (Populus alba L.) is a slower growing species, but tolerant of a wide range of site conditions and higher levels of exposure (Jobling, 1990), and may have been a more appropriate choice for our experiment, particularly if established in advance of the main planting.
Being porous, living windbreaks offer their maximum shelter for a distance up to 10 times the height of the trees, with some degree of benefit being obtained for up to 30 times the height of trees (Gardiner et al., 2006). They need cost no more to establish than an equivalent area of the plantation they are protecting. Species that have been suggested as appropriate for shelter belts include willows (Salix spp.), alders (Alnus spp.), sycamore and beech or conifers such as Scots pine (Pinus sylvestris L.), Austrian pine (P. nigra ssp. nigra Arnold) and additionally in the lowlands, poplars (Baxter, 1986). Sycamore proved to be relatively tolerant of exposure in our experiment, so they may be a good future choice for windbreaks. Although conifers will offer more shelter during the winter period when most damaging gales occur, deciduous species can offer suitable protection when crop trees are in active growth. For maximum shelter effect, living windbreaks should be established before the main crop. The provision of shelter is often cited as one of the advantages of establishing mixed species stands (Kerr et al., 1992), and conifers such as European larch (Larix decidua Mill.) and Norway spruce (Picea abies (L.) H. Karst.) have also been recommended as for use in intimate mixtures with ash or oak (Quercus petraea (Matt.) Liebl.; Q. robur L.), to provide early side shelter to the broadleaved species (Evans, 1984). However, with intimate mixtures, care must be taken to select species with comparable growth rates on a site-by-site basis (Kerr and Evans, 1993).
|
Conclusions |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
The results from our work suggest that even on apparently sheltered lowland sites, broadleaved species such as sycamore and ash can be significantly affected by exposure. Angle of lean of adjacent young trees may not be a good indicator of likely future growth suppression arising from exposure, as apparently unaffected straight trees may in fact still be suffering from severe growth suppression. However, Ecological Site Classification software (Pyatt et al., 2001
) can give an estimate of exposure for individual sites throughout Britain. Alternatively, where time permits, tatter flags give a close correlation with wind speed (Mackie and Gough, 1994) and can provide a relatively simple method of more accurately judging exposure of a specific local site. The economic justification for providing specific types of shelter would depend on the magnitude and importance attached by managers to the predicted growth gains, which will in term depend on the tree species used, the initial exposure of a particular site and the proportion of the site likely to benefit. Currently, unless shelter could be provided cheaply by means of a nurse species, or through the utilization of surplus straw bales, the early growth gains demonstrated by our work would by themselves probably not justify the additional cost involved in installing artificial barriers. However, with the onset of climate change, consideration of the effects of exposure on tree growth are likely to become more important in the future. Although currently only relatively small (1–2 per cent) increases in summer winds are predicted for the southern lowlands of Britain, these are likely to be accompanied by significant increases in temperature and decreases in precipitation (Broadmeadow, 2002). This may make the threat of drought, growth suppression and death of young, broadleaved trees particularly acute in exposed lowland sites in the south east of Britain. The provision of side shelter to help reduce evapotranspiration and summer drought is therefore likely to become an increasingly important silvicultural consideration in the future. If shelter is judged to be necessary and cannot be provided by other trees, our work suggests that straw bales might provide an effective, sustainable method of protecting narrow planting schemes such as hedgerows. The use of a suitably robust 60 per cent porous plastic netting may be an option for providing shelter from prevailing winds for sensitive broadleaved species, in situations where fencing is already planned to protect against browsing mammals, but some strengthening of fences would be required, particularly in exposed upland settings. More research is required on different site types, materials and nurse species to refine our initial models of the effects of exposure on broadleaved trees in the lowlands.
|
Conflict of Interest Statement |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
None declared.
|
Acknowledgements |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
We are grateful to Dave West, Dave Hendrie, John Lakey, Paul Turner, Darryl Watts and colleagues at Thetford Field Station, who established, maintained and assessed the experiment; Geoff Morgan, Jane Poole and Sophie Hale for assistance with analysis and Bruce Nicoll, Bill Mason, Palle Madsen, Helen McKay and a further anonymous referee for their constructive comments on earlier versions of the draft.
|
References |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsDiscussionConclusionsConflict of Interest StatementAcknowledgementsReferences |
|---|
Baxter SM. Windbreaks for horticulture. In: MAFF/ADAS Booklet (1986) 2280. Alnwick, Northumberland: MAFF Publications.
Broadmeadow M. Climate change: Impacts on UK forests. In: Forestry Commission Bulletin (2002) 125. Edinburgh: Forestry Commission.
Davies WJ, Kozlowski TT, Pereira J. Effects of wind on transpiration and stomatal aperture of woody plants. In: Mechanisms of Regulation of Plant Growth, Bulletin—Bieleski RL, Ferguson AR, Cresswell MM, eds. (1974) 12. Royal Society of New Zealand, 433–438.
Dixon M, Grace J. Effects of wind on transpiration of young trees. Ann. Bot. (1984) 53:811–819.
Evans J. The silviculture of broadleaved woodland. In: Forestry Commission Bulletin (1984) 64. London: HMSO.
Gardiner B, Palmer H, Hislop M. The principles of using woods for shelter. In: Forestry Commission Information Note (2006) 81. Edinburgh: Forestry Commission.
Genstat. Genstat 5 Release 3 Reference Manual (1993) Oxford: Clarendon Press.
Heiligmann R, Schneider G. Effects of wind and soil moisture on black walnut seedlings. For. Sci. (1974) 20(4):331–334.
Heiligmann R, Schneider G. Black walnut seedling growth in wind protected microenvironments. For. Sci. (1975) 21(3):293–297.
Horgan T. Shelter is key to successful broadleaved silviculture. For. Br. Timber (2004) 33(6):14–18.
Jobling J. Poplars for wood production and amenity. In: Forestry Commission Bulletin (1990) 92. London: HMSO.
Kerr G, Evans J. Growing broadleaves for timber. In: Forestry Commission Handbook (1993) 9. London: HMSO.
Kerr G, Nixon CJ, Matthews RW. Silviculture and yield of mixed species stands: the UK experience. In: The Ecology of Mixed Species Stands of Trees—Cannell MGR, Malcolm DC, Robertson PA, eds. (1992) Special Publication Series of the British Ecological Society Number 11, 35–51. Blackwell, Oxford.
Kozlowski TT, Kramer PJ, Pallardy SG. The Physiological Ecology of Woody Plants (1991) New York: Academic Press.
Lines R, Howell RS. The use of flags to estimate the relative exposure of trial plantations. In: Forestry Commission Forest Record (1963) 51. London: HMSO.
Mackie AL, Gough PL. Using tatter flags to assess exposure in upland forestry. In: Forestry Commission Research Information Note (1994) 249. Edinburgh: Forestry Commission.
Mackney D, Hodgson JM, Hollis JM, Staines SJ. Legend for the 1:250,000 Soil Map of England and Wales (1983) Soil Survey of England and Wales, Rothamsted, Harpenden.
Miller KF. Windthrow hazard classification. In: Forestry Commission Leaflet (1985) 85. London: HMSO.
Miller KF, Quine CP, Hunt J. The assessment of wind exposure for forestry in upland Britain. Forestry (1987) 60(2):179–192.
Nixon CJ. Effectiveness of treeshelters in upland Britain. Q. J. For. (1994) 88(1):55–62.
Olsthorn AFM, Bartelink HH, Gardiner JJ, Pretzsch H, Hekhuis HJ, Franc A, eds. Management of Mixed-Species Forest: Silviculture and Economics (1999) IBN Scientific Contributions 15. IBN-DLO, Wageningen, Holland.
Potter MJ. Treeshelters. In: Forestry Commission Handbook (1991) 7. London: HMSO.
Pyatt G, Ray D, Fletcher J. An ecological site classification for forestry in Great Britain. In: Forestry Commission Bulletin (2001) 124. Edinburgh: Forestry Commission.
Quine CP, White IMS. Revised windiness scores for the windthrow hazard classification. The revised scoring method. In: Forestry Commission Research Information Note (1993) 230. Edinburgh: Forestry Commission.
Quine C, Coutts M, Gardiner B, Pyatt G. Forests and wind: Management to minimise damage. In: Forestry Commission Bulletin (1995) 114. London: HMSO.
Rees DJ, Grace J. The effects of wind on the extension growth of Pinus contorta (Douglas). Forestry (1980) 53(2):145–153.
Ritson P, Sochacki S. Measurement and prediction of biomass and carbon content of Pinus pinaster trees in farm forestry plantations, South-Western Australia. For. Ecol. Manag. (2003) 175(1):103–117.[CrossRef]
Rutter N. Tattering of flags at different sites in relation to wind and weather. Agric. Meteorol. (1967) 5:163–181.[CrossRef]
Savill PS. Assessment of the economic limit of plantability. Irish For. (1974) 31(1):22–35.
Thompson DA. The influence of shelter and weeds on early growth of Sitka spruce and lodgepole pine. Forestry (1984) 57(1):1–16.
Trout R, Pepper H. Forest fencing. In: Forestry Commission Technical Guide (2006) 2. Edinburgh: Forestry Commission.
Welham SJ, Thompson R. Likelihood ratio tests for fixed model terms using residual maximum likelihood. J. R. Stat. Soc. B (Stat. Methodol.) (1997) 59(3):701–714.
Willoughby I, Dewar J. The use of herbicides in the forest. In: Forestry Commission Field Book (1995) 8. London: HMSO.
Willoughby I, Evans H, Gibbs J, Pepper H, Gregory S, Dewar J, et al. Reducing pesticide use in forestry. In: Forestry Commission Practice Guide (2004) 15. Edinburgh: Forestry Commission.
Wilson JD, Flesch TK. Wind measurements in a square plot enclosed by a shelter fence. Bound.-Layer Meteorol. (2003) 109(2):191–224.
Received 9 April 2008.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||