Crown condition is assessed byestimating the 'reduction' in density of a tree's crownwhen compared either with an 'ideal' tree or with thebest 'local' tree in the vicinity of a plot. Since 1987,the crown condition data from the Forest ConditionSurvey have similarly been presented in terms ofreductions in density. However, rather thanconsidering the percentage of the crown which isabsent, it may be easier to visualise the condition oftrees on the basis of the percentage of the crownwhich is present when compared with a knownstandard. Therefore, although the methods used forassessing crown condition in 2002 remainedunchanged from those employed in previous years, theresults of the survey are presented here in terms ofcrown densities rather than crown density reductions.It should be noted that percentage scores for crowndensity and crown density reduction arecomplementary i.e. the sum of the two measures foran individual tree is always 100%. To compare thenumerical results presented here with those fromprevious surveys, percentage crown density can beconverted to percentage crown density reduction byapplying the following formula:
% crown density reduction = 100 - % crown density
The crown density results obtained using both theideal and local tree methods are presented in 10%classes in Table 1. The marked effect of using a localreference tree rather than an ideal tree as the basis forcomparison can be seen for all species. Much of thedifference can be accounted for by variations ingrowth habit between the reference photographs ofideal trees (Innes, 1990) and the trees in and aroundthe plots to be assessed, from among which a localreference tree is chosen. For example, young trees ofall species, but particularly Scots pine, tend to have amore open appearance (i.e. a lower crown density)than the older trees illustrated in Innes (1990). Some older oaks and spruces also have a naturally openstructure. The crown densities of trees like these aremuch lower when compared to an ideal tree than whenjudged against local trees of the same age and form.
Figure 1 shows the changes in crown condition thathave taken place since 1987. A 'downward' gradient inthis figure indicates a deterioration in crowncondition. Changes in condition compared to last yearwere minor in Norway spruce, Scots pine and beech,but marked reductions in crown density occurred inSitka spruce and oak. In spite of this, none of thechanges in crown density which occurred between2001 and 2002 were statistically significant: analysisof the time series for each species shows that relativelylarge fluctuations in crown density (more than 3% to4%) between years are required for the statistical 5significance of such changes to be established. Short-termfluctuations of such magnitude are relativelyinfrequent in all of the surveyed species except beech,the condition of which has been characterised by largeinter-annual changes in crown density over the 16 yearsurvey period. However, there is no evidence of along-term trend for deterioration or improvement inthe crown density of beech; similarly, no trend in thecrown density of either Sitka spruce or Norway spruceis apparent. In the case of Norway spruce, thisobservation indicates a reversal in the trend fordeterioration which has previously been reported(Hendry, Boswell & Proudfoot 2001, 2002).However, since Forest Condition Survey plots aresubject to normal silvicultural operations, clearfellingof old crops in poor condition has had an effect on theoverall score for the species by removal of those plotsdisplaying the lowest crown densities. As a result, theimprovement in the condition of Norway spruceindicated by the survey does not simply reflect a longtermchange in the health of the species per se but is atleast partly attributable to a change in the nature ofthe population which is being sampled.
Analysis of the 1987 to 2001 data indicates that astatistically significant deterioration in the crowncondition of both Scots pine and oak has occurredover the duration of the survey. However, the timeseries is of short duration and the indicated rates ofchange are small, with an apparent reduction in crowndensity of 0.24% per annum in Scots pine and 0.53%per annum in oak. For oak, the trend is heavilyinfluenced by the high crown density values recordedin the period 1987 to 1990 when the number of surveyplots of this species was relatively low and cautionshould therefore be excercised in the interpretation ofthis result. The magnitudes of past increases in crowndensity in both oak and Scots pine suggest that a singleyear of major improvement in condition could nullifythe current trends for either species.
Since 1991 the condition scores of Scots pine andNorway spruce have changed less than those of any ofthe other species. Although Scots pine deterioratedslightly in 2002 and the condition of Norway spruceshowed a minor improvement, neither of thesechanges was significant. Following a sharp decline in1997, the condition of Sitka spruce improved between1999 and 2001 but this recovery was completelyreversed in 2002. The crown density of oak alsodecreased this year, its condition now being as poor asin 1997. The major improvement in beech whichoccurred in 2001 was partially offset by a minorreduction in its crown density in 2002.
Table 1: Percentages of trees in each crown density class for five species in 2002. Each 10% class represents a reduction in crown density compared either to an 'ideal tree' (I), i.e. a tree with the maximum possible amount of foliage, or to a 'local tree' (L), i.e. a tree with full foliage under local conditions.
Figure 1: Changes in crown density since 1987 for five species surveyed annually. The crown density compared with that of an 'ideal' tree with a completely opaque crown is shown for each species.
Figure 2 shows the geographical variation in crowndensity for the five species assessed. Variation wasgreatest in oak, which continues to display a patternsubstantially unchanged since 1997 when data werefirst presented in this way (Redfern et al., 1998; 1999;2000). The condition of oak was poorest in northernand central Scotland, north-east England, Wales andEast Anglia, and best in southern England. Scots pinealso displayed a pattern which was similar to that ofprevious years, with crown density tending to behighest south of the Humber-Mersey line. Beech andNorway spruce showed no clear pattern. Sitka sprucewas poorest in southern and western England, but thisimpression is created by relatively few plots and thespecies shows considerable local variation.
Factors affecting crown condition in 2002
The autumn of 2001 was generally warm and wet,with frequent heavy rainfalls and a record highaverage October temperature. Mild, wet conditionsalso predominated over the winter months withFebruary being abnormally warm. Although April andMay were largely dry and clear, minimumtemperatures rarely approached damaging levels andthere were few instances of spring frost damage totrees. Towards the end of April, severe squalls incertain parts of the country led to localised branchbreakage and windblow. Rainfall was higher thanaverage over the summer months, with July beingparticularly wet, and as a result no drought damage toforest trees was reported. Continuing warm weatherover the remainder of the growing season providedconditions which were generally good for tree growth.
The decrease in the crown density of Sitka sprucerecorded in 2002 was of sufficient magnitude toreverse the improvement in condition which thespecies had displayed between 1999 and 2001 (Figure1). This deterioration was almost entirely due towidespread defoliation by the green spruce aphidElatobium abietinum, with attacks being reportedfrom 43 of the 63 survey plots. However, damage wasoften confined to the lower areas of the crowns of thesurveyed trees and was therefore not as severe asmight have been anticipated from inspection of individual trees and shelter belts located near to plots,which were often completely defoliated.
In common with previous cases of decline in 1995 and2000 (Hendry, Boswell & Proudfoot, 2001), thedeterioration in the condition of beech recorded in2002 was largely associated with heavy masting. Mastproduction was noted on 78% of the surveyed treesand was recorded as heavy (common or abundant) in58% of the population. Levels of insect damage tobeech in 2002 were broadly similar to those recordedin 2001, although the incidence of severe damage wasslightly elevated due to heavy attack by the leafmining insect Rhynchaenus fagi in 17 plots.
Changes in the crown densities of Norway spruce andScots pine this year were minor, thus continuing apattern of little variation in condition which has heldsince 1993. In 2002, Norway spruce was defoliated byElatobium abietinum but it is less susceptible to attackthan Sitka spruce and damage was slight. Distortionof the branching pattern of trees caused by the budblight fungus Cucurbitaria piceae was observed inplots in northern England and southern Scotland butthis had little influence on the crown density ofaffected individuals. The slight deterioration in Scotspine noted this year was attributable both to elevatedlevels of infection by the needle pathogenLophodermium seditiosum and to abundantproduction of male flowers on the 2002 shoots. Theloss of foliage infected by L. seditiosum was reflectedby a reduction in the percentage of trees retainingneedles for 3 or more years from 62% in 2001 to47% in 2002.
The deterioration of oak which occurred in 2002completely reversed the improvement in conditionwhich had occurred over the previous two years, andthe crown density of the species is now as low as atany time during the 16 year survey period. However,this decline cannot be related to serious climaticinjury, or to a major outbreak of a particular pest ordisease in 2002, since the incidence of abiotic andbiotic damage to trees was broadly similar to thatrecorded in 2001. Analysis of the changes in crowndensity which occurred between 2001 and 2002 in the1955 trees common to both surveys indicates that1009 trees deteriorated in condition. Of these, 68%displayed a reduction in crown density of only 5 to 10%whilst 86% displayed a reduction of 5 to 15%. Theoverall change in the condition of oak since last year istherefore attributable to a minor reduction in the crown density of many trees rather than a majorchange in the condition of a few trees. Since theclasses which are used to record abiotic and bioticdamage in the Forest Condition Survey are broaderthan those used for the recording of crown density, aminor change in the crown density of a tree would notnecessarily be reflected by a change in the damageclass to which it was allotted. Thus, a small butgeneral increase in the level of insect and fungaldamage to oak in 2002 may account for the recordedchange in condition.
Insect damage was present in 81 of the 86 oak plotsassessed in 2002 but was severe in only 9 plots wheredefoliation by the winter moths Operophtherabrumata and Erranis defoliaria had occurred. CentralScotland was particularly affected and it is noticeablethat the highest concentration of plots in poorcondition was located here (Figure 2). Branch breakresulting from the severe winds in April was recordedin 7 plots in northern England and Scotland but noother significant climatic damage was noted. Althoughnot directly damaging, the warm and wet conditionswhich prevailed in spring and early summer are likelyto have favoured the development of powdery mildew(Microsphaera alphitoides) which was more widespreadand severe this year than in the recent past.
Rainfall was well distributed throughout the 2002growing season and tree growth was generally good.High winds in April caused localised damage to all ofthe surveyed species but no other forms of climaticinjury were important this year. Changes in conditionwere minor in Norway spruce, Scots pine and beechbut marked reductions in crown density occurred inSitka spruce and oak. The slight deterioration of Scotspine this year was largely due to an increase in theincidence of the needle cast fungus Lophodermiumseditiosumand continued heavy male flowering onmany trees. Crown density in Norway spruce hasdisplayed only minor fluctuations since 1991 andremained virtually unchanged in 2002. A slight declinein the condition of beech was largely associated withheavy mast production and is not necessarily anindication of ill health. The sharp decrease in crowndensity displayed by Sitka spruce was due towidespread defoliation by the green spruce aphidElatobium abietinum. However, attack was oftenconfined to the lower areas of the crowns of the surveyed trees and damage was therefore not as severeas recorded after the previous outbreak of Elatobiumin 1997. The deterioration in the condition of oakwhich occurred this year was attributable to a minorreduction in the crown density of many trees ratherthan a notable increase in the incidence of attack bypests or pathogens.