JUDGMENT

Decision

The predominant cause of the desiccation and subsidence that occurred was the oak in the ownership of the defendant highway authority who is, in consequence, fully liable for the damage with no reduction for a cause within the claimants’ ownership or as a result of any contributory negligence by them.

JUDGMENT

1. Introduction

1. This judgment is concerned with damage caused by tree roots. The claimants have owned and occupied no 56 Woodfield Road, Ealing, London, W5 since 1971. This house is semi-detached and its pair and mirror image is no 54. In the period from 1994 until at least October 1996 when monitoring ceased, these houses were damaged by subsidence, cased by tree roots. The present dispute arises because there are two separate potential causes of this damage, firstly a large oak tree located in the highway and therefore owned and under the control of the defendant highway authority, the second being a group of trees and other vegetation located in the front garden of no 56 and the adjacent front garden of no 58, the property also neighbouring no 56. This latter group was partly in the ownership and control of the claimants and partly in the ownership and control of the Malewskis’ next door neighbours at no 58. The dispute is, essentially, as to which insurer, being the household insurer of the claimants or the public liability insurer of the defendants, should pay for the underpinning and consequential repairs and the other consequential costs arising from the subsidence damage incurred between September 1997 and May 1998.

2. Woodfield Road consists of pairs of semi-detached two storey houses built in 1907 with solid masonry walls and timber floors beneath pitched slate roofs. No 56 has a rear annex which was probably constructed at the same time as the main structure but which is of different construction with a flat roof. Opposite no 50 Woodfield Road, and about 19m from the nearest point of the pair 54-56 Woodfield Road, is located the mature oak in question which is at least 160 years of age. The two houses were constructed of what are, by today’s standards, very shallow foundations consisting of concrete footings 0.5m in depth.

3. Both houses have small front gardens and longer rear garden and, in the front garden of no 56, in 1994 a purple plum and six Leyland Cypress trees and an ivy bush were located. In the road outside no 56 was an elderberry tree and adjacent to no 56 in the front garden of no 58 were located 4 Lawson Cypress trees. Although referred in passing in the report of the defendant’s arboriculturalist as a potential contributor to the desiccation process causing the subsidence in question, the elderberry tree located in the highway did not feature in the evidence as a potential contributor to that damage and I do not refer to it further.

4. The cracking in question was first noticed by Mr and Mrs Malewski in the late spring or early summer of 1994. This cracking did not appear significant at that time but, in June 1995, a major vertical crack opened up in the party wall at about its mid-point and, on 29 June 1995, Mr Malewski submitted a claim for its repair to his household insurers. This crack developed and further cracks and related damage opened up. All this damage was located in or close to the party wall and in the front bay located near by. In the same period, similar cracking and damage opened up in no 54, albeit such damage was not so extensive.

5. The claims arising from this damage were dealt with separately by the different insurers of the two properties but the repairs and underpinning were undertaken simultaneously as a joint scheme by the same contractor. Prior to that work being undertaken, the cracking was monitored in both properties and, for no 56, two separate site investigations were undertaken comprising the drilling of a borehole in the front garden on two occasions, August 1995 and December 1996 with the resulting soil samples being tested. A third, separate, site investigation was undertaken in no 54 in February 1996 and three boreholes, two in the front and one in the rear, were drilled and samples taken and tested.

6. The repairs essentially consisted of the construction of underpinning to the front walls and the party wall of nos 54 and 56. This underpinning was to varying depths down to 4m with the deepest lengths of underpinning being those: (1) at the front of no 54, (2) running for a short distance along the party wall adjacent to the front wall and (3) running for the further short distance from the party wall towards the left hand side of the bay of no 56. Significantly, the underpinning to the bay of no 56 closest to the trees in its front garden was to lower depths varying from 3.5 down to 2.0m.

7. Associated with these investigations were two separate culling exercises. Mr Malewski, on the advice of the engineers initially instructed who undertook the first site investigation in August 1995 and who designed the underpinning work, arranged for the removal of the six Leyland Cypresses at the front of the property as well as for the crown reduction of the purple leaf plum tree and for its thinning by 20 - 25%. The crown reduction reduced the height of this plum tree from about 8m to about 4m. Tree removal of two cypresses was also undertaken at the rear. This work was carried out in August 1995. Separately, the defendant arranged for the light crown reduction of the oak by about 10% and its medium thinning by about 20%. This work was carried out in about August 1995.

2. The Damage

8. The relevant damage was caused by tree roots abstracting moisture from the predominantly clay soil located beneath the foundations of the two houses. This clay is London Clay, being stiff over consolidated clay which weathers to a characteristic reddish brown colour where it outcrops. Once there had been significant moisture abstraction, the clay became desiccated and shrank. This volume change caused withdrawal of support from the foundations, foundation movement and consequent superstructure damage, largely in the form of cracking in walls supported by the foundations. Overall, the relevant damage can be characterised, adopting the description used by Lord Cooke in the leading modern authority concerned with tree root induced damage, Deleware Mansions Ltd v Westminster City Council [2001] 3 WLR 1019, HL at paragraph 33, as “impairment of the load-bearing qualities of the sub-soil”.

9. Clay soils exist at a moisture content which is in equilibrium with the in-situ stresses in the ground, particularly the vertical overburden pressure. This moisture content can vary as a result of climatic wetting and drying of the soil close to the surface. The drying part of this cycle may be exacerbated by moisture extraction by tree roots and other vegetation. Moisture drying and wetting is normally confined to depths of no more than about 1 to 1.5m and the consequent desiccation and soil shrinkage will occur in the summer months and subsequent soil swelling in the winter months. Climatic soil movement is rarely highly localised nor deeper than 1.5m so that any localised or deep seated shrinkage is usually associated with vegetation moisture abstraction.

10. The principal damage was confined to the subsoil located beneath the front foundations, particularly at the point close to and beneath the right-angled junction between the front and party wall foundations on the 56 Woodfield Road side of that junction. The two boreholes taken at that location in July 1995 and December 1996 showed that the soil was severely desiccated down to a depth of 4 metres. This desiccation was revealed by the moisture content and soil suction test results carried out on samples down to below 4 metres in depth. These boreholes also revealed traces of root presence at depths of between 0.5 and 2.5m. These traces were identified and were predominantly oak roots but there were less significant traces of plum and ivy roots as well. The bulk of the roots, being all the oak roots and only a few of the remaining roots, were found in the band between 2.0 and 2.5m.

11. A further borehole taken in February 1996 at the front left hand corner of 54 Woodfield Road at a point close to the nearest point of the house-pair to the oak outside 50 Woodfield Road showed both a relatively higher soil moisture content and significantly less desiccation given the low suction test results from soil samples taken from the boreholes located in the front garden of no 56. However, this borehole was sunk during the winter month of February when any desiccation in the soil, which would have occurred in the summer and early autumn months of the previous year, would largely have been eliminated. Moreover, as Mr Daley, the claimants’ expert engineer, pointed out, the relatively lower moisture content of the samples from this borehole at the front of no 54 indicated that even at this location the underlying clay was susceptible to being desiccated.

12. The damage itself was predominantly manifested as cracks which widened and narrowed cyclically in the summer and winter of both 1995 and 1996 as the subsoil correspondingly shrunk whilst desiccated in the summer and swelled whilst moist in the winter. These cracks were confined to each side of the party wall, to the junction of the front door and party wall and the left hand side of the adjacent bay to no 56 and to both sides of the bay to no 54. These cracks gave rise to associated distortion of the plaster and brickwork and were agreed by both engineers to have been initiated and propagated by subsidence beneath the front of the party wall leading to a rotation of the party wall along an axis located in the landing at nearly mid-point along this load bearing structure such that large cracks, whose width narrowed from the top downwards, opened up. Mr Daley, the claimants’ engineer expert witness, suggested that this movement had, in lay terms, broken the back of this major structural feature of the house-pair.

13. The movement of these cracks was monitored in the period July 1995 until November 1996 with 7 readings taken in this period for no 56 and 4 for no 54. These readings indicated that the two principal cracks in each property, 3 of which were located in or close to the party wall and the fourth, in no 56, located close to the junction between that wall and the front wall, had opened to between 1 and 4mm by July 1995, had remained static until the autumn of 1995, had then closed or significantly narrowed until the spring or early summer of 1996 and had then re-opened rapidly and either returned to their previous summer’s width or widened to an even greater extent. Since Mr Malewski first noticed the sudden and dramatic crack opening on his side of the party wall in early July, it is a fair inference that the crack cycle started in about July 1995 with a sudden and dramatic surge, that it then remained in the open stage of the cycle until the late autumn when the cracks virtually closed and the cycle was then repeated in 1996 with approximately the same times for opening and closing. A fifth crack, also located in the party wall of no 56, followed the same pattern but with opening and closing movements of about half the size of the other two cracks being monitored at no 56.

14. The other five cracks, two in no 56 and three in no 54, did not move appreciably during the opening and closing cycles in 1995 and 1996 but these must have initially opened at the same time in June 1995 when the other cracks initially opened. The relative lack of movement of these other cracks following initial opening can be explained by these cracks not being directly associated with or linked to desiccated subsoil so that they did not move sympathetically with corresponding movement in the subsoil. They were instead cracks formed consequentially on the initial cracking and structural movement in the party wall.

15. The initial opening up and subsequent cyclical movement of the cracks associated with the party wall can be linked directly to the unseasonal lack of rainfall in both 1995 and 1996. The initial opening up occurred in June 1995 when an unusually dry spell had been present for about 3 months. Two further very dry months followed in July and August 1995, followed by an exceptionally wet September, very dry months in October and November and then a very wet December. 1996 repeated, with less severity, this pattern with each month from March to July having less than usual rainfall with June being particularly dry. The autumn months were dryer than normal although not exceptionally so. The severity of the opening and closing amplitudes of the crack cycles can generally be linked to the pattern of unseasonal weather in those two years which must, in the summer months, have created conditions in the subsoil which were particularly susceptible to desiccation if tree roots were also present withdrawing further appreciable quantities of moisture.

16. The slight cracking observed by the Malewskis in 1994 was obviously superficial and not of a kind to give rise to any appreciable concern. The date when that first cracking initially opened up cannot be placed with precision since Mr Malewski did not note it down at the time but he stated in evidence that he believed that it was in the late spring or early summer and, given the opening and closing cycles that occurred in 1995 and 1996, I find that that initial date occurred in the same period in 1994, between late May and early July, that corresponded to the opening up that occurred in those subsequent two years.

17. The area around the oak is relatively flat. However, two level surveys were undertaken. The first, of no 54, was carried out as part of the site investigation of that property in February 1996, the second, of both properties, was carried out by the joint instruction of the parties on the recommendation of their engineer experts in January 2003. These level surveys, and particularly the second, revealed two remarkable features of the properties. The first is that the pair of properties has one high point, the left hand corner of no 54, and the second is that the rear annex of each property has clearly rotated away from the main structure so as to lie, at the rear wall, about 40 - 50mm lower than the rear of the main structure. The high point in the front left hand corner of no 54 is very localised with points close by to the front and side of no 54 being appreciably lower.

18. There is also a pronounced dip in the front wall on the no 56 side of the party wall which can be associated with the location of the desiccated subsoil located by the two boreholes in that location drilled in August 1995 and December 1996.

19. I must, finally, in this summary of the primary evidence, summarise the relevant features of the Oak, Plum and Cypresses that are the major contenders for causing the desiccation. It is a feature of all trees that they draw in moisture from their roots to feed the leaves which then lose much of this moisture by transpiration from their face. The tree roots migrate away from the trunk, in the case of some trees for distances of up to three times their height, and it follows that desiccation can occur up to 30 - 40m away from the trunk of a tall oak such as the one in question in this case. However, as is clear from the evidence obtained from published material put in evidence at the trial, 90% of oak tree root induced damage occurs at distances up to 18m. Once roots have developed, they remain in situ even if the tree is subsequently pollarded and remain available for moisture removal to feed the tree’s demands. These demands are, substantial for any tree and for some, such as oaks, are prodigious.

20. Oaks are notorious initiators of desiccation and subsidence. Their propensity to cause damage was described in a text relied on by both arboriculturalists as:

“... deep-rooted on clay soils, potentially dangerous to plant near buildings on clay sub-soils ... [and] give the highest returns of reported damage.” (Cutler and Richardson: Tree Roots and Buildings)

It was the evidence of the claimants’ expert arboriculturalist, Dr Dobson, whose careful, authoritative and considered evidence I preferred to Dr Hope who gave evidence on behalf of the defendant, that oak rooting can cause desiccation down to about 4m and that the roots themselves are often found down to about 2.5m in clay. This is a characteristic of oaks not found in other rooting species such as plums and cypresses where desiccation is rarely found below 2m and where the rooting is at depths appreciably less deep than oak rooting. It is also clear from the evidence that plums and cypresses are considered as being of moderate to light risk which should not be planted nearer than, respectively, 3.5m and 6m to a building whereas oaks are severe risks which should not be planted nearer than 13m to a building.

21. The traditional way to curb and control a tree’s potential to cause desiccation in clay soils is to reduce the height of the crown and to thin the crown. It was the consensus of the expert written material adduced by Dr Dobson as support for his opinions that a height reduction of at least 30% is needed to produce an appreciable reduction in the moisture extraction propensity of a tree although lesser height reductions can sometimes be effective when undertaken in association with heavy thinning. It is to be noted that the reduction of the height of the oak in August 1995 was only 10% whereas the plum height reduction was about 50%. Furthermore, although the thinning of the Oak was described as medium, that of the plum was described as 20 - 25%.

22. In the light of that expert supporting evidence, I accept Dr Dobson’s opinion that the height reduction and thinning exercise carried out on the oak in August 1995 had no perceptible effect on its moisture removal propensity whereas the severe height reduction and thinning of the plum, which was in any case located at a safe distance from no 56, would have severely reduced that tree’s capacity to extract moisture through its roots. I do not accept Dr Hope’s riposte to this opinion that the plum grew back with such speed that its moisture reduction tendencies had been fully restored by 1996 since it did not increase in growth by anything like the 4m that had been culled from it in the few months between its having been culled in August 1995 and the onset of crack movement in about June 1996. It must also be borne in mind that the Leyland Cypresses located in the front garden of no 56 were removed at that time. These culling exercises produced what amounts to a control when the monitoring records are considered since the 1996 results can fairly be attributed to the oak alone whereas the 1995 results can be regarded as the result of the combined effect of the oak, the plum and the Leyland cypresses. The further cypresses at no 58 are too far away from the party wall to have been capable of having any significant influence on the cracking that actually occurred.

3. The Experts’ Views

23. The four experts, being the pair of engineers and the pair of arboriculturalists, produced three joint statements, two coming from the arboriculturalists and one from the engineers. The four are to be commended for having reduced the ambit of their disagreements so comprehensively. In the light of those statements, the differences which led Mr Daley and Dr Dobson for the claimants to implicate the oak and Mr Ainsworth and Dr Hope for the defendant to implicate the plum and the cypresses may be briefly stated.

24. All four experts took as their starting point that, before any relevant data was considered, the plum and the cypresses were the obvious candidates for being implicated as the originators of the relevant desiccation. They were located very close to no 56 whereas the oak was much further away. In any case, the oak had not caused damage to the part of the pair of houses nearest to it at their left hand corner but, if it was the originator of the relevant damage, it had nonetheless caused damage at points up to 4m further than that nearest point. Finally, as the defendant’s experts saw the situation, the oak would have caused subsidence at that near point if it was implicated yet that point was higher than any relevant point around the pair of buildings. All four experts did, however, accept that all relevant sources of evidence should be considered together and in the round.

4. The Evidence

4.1. General

25. The relevant evidence falls into six categories: (1) the nature and location of the damage and the contents of the remedial work; (2) the subsoil and its tree roots; (3) the crack monitoring records; (4) the level surveys; (5) the propensities of the relevant trees; and (6) similar fact evidence of damage to other neighbouring properties. It is helpful to make findings about each in turn.

4.2. The nature and location of the damage and of the contents of the remedial work

26. Both engineers accepted that the damage to both properties had been caused by a rotation of the party wall caused by subsidence of the footings located at the front of that wall where it abutted the front wall. Thus, the subsidence, and its associated desiccation where extremely localised and were confined to that particular junction. Mr Daley pointed out that the remedial work was obviously designed by the engineer on the premise that it was to provide against oak tree desiccation and not plum tree or cypress tree desiccation. This was because sections of the underpinning went down to 4m, a depth only required if oak tree desiccation was being guarded against, and because the location of those particular lengths was at the points nearest to the oak and furthest away from the vegetation in the front garden of no 56. Indeed, the footings nearest the removed cypresses in no 56 or the remaining cypresses in no 58 were not underpinned at all by the remedial work.

27. The absence of damage at the left hand corner of no 54 is no more than an indication that the oak tree was not the cause of the desiccation in question. It is clear from the evidence that that indication could be countermanded by the other available evidence. Desiccation is particularly susceptible to localised topographical features and temperature differences, water courses, variations in the level of the water table, localised leakages and other features. It is also clear that apparent anomalies can occur whereby desiccation from an oak tree occurs at points further away from the tree without desiccation and consequent damage occurring nearer to the same tree. In this case, the clay subsoil nearer to the oak at the front of no 54 is obviously susceptible to desiccation but it did not become desiccated yet the evidence clearly points to oak-induced desiccation beneath the party wall further away. There has therefore obviously been an anomaly of the kind that I have referred to which saved the nearer locations from becoming desiccated at a time when the more remote location became desiccated.

4.3. The subsoil and its tree roots

28. The samples of tree roots found in the two boreholes drilled in no 56’s front garden were found at random. The presence of tree roots taken by themselves is not indicative of desiccation or that those roots have caused desiccation but the type and depth of the roots and their association with desiccation are highly significant if taken together. Firstly, the great majority of the root samples that were found were identified as being oak rootings. Secondly, most of the roots and all of the oak roots were found at depth, the majority between 2.0 and 2.5m. Thirdly, the soil was heavily desiccated down to 4.0m. All these factors taken together leads to the conclusion that the desiccation at that location was wholly or predominantly caused by oak rooting and, indeed, only occurred because of a coincidence of two very dry summers and an unculled oak in the vicinity which had particularly high moisture demands.

29. This conclusion arises because desiccation at depths below 2.0m cannot ordinarily be caused by other than oak roots, because of the overwhelming presence of oak roots at those deeper points at the relevant location and because the clay was heavily desiccated at depths where there was significant oak tree root presence. All this evidence taken together confirmed that that deep-seated desiccation was indeed caused by oak roots.

4.4. The crack monitoring records

30. I have already made the necessary findings about the crack monitoring records. In summary, the height reduction and thinning measures taken in respect of the oak in August 1995 would have made no difference to the desiccation propensities of the oak whereas the removal of the cypresses and the heavy height reduction and thinning measures taken in respect of the plum in the same month would have eliminated their desiccation propensities during the following summer. Notwithstanding these measures, the extent of desiccation that occurred in 1996 was similar to, or greater than, the extent of desiccation that had occurred in 1995 as was shown by the fact that the crack cycle in 1996 was similar to, or greater than, the equivalent cycle in 1995. Thus, the inevitable conclusion is that the predominant influence in causing desiccation in both 1995 and 1996 was the oak since the same amount of cracking occurred when the only desiccating influence was the oak as had occurred when there were also available the other potential influences of the plum and the cypresses.

4.5. The level surveys

31. The level surveys provided the defendant both with the high point of the buildings and the high point of its case. They showed, somewhat surprisingly, that no damage or subsidence had been caused to the part of the buildings nearest to the oak and that that point had not subsided which, if the oak was an influence at all, was the opposite to what was to be expected.

32. Two conclusions may, however, be made about the surveys. Firstly, the movements in the footings that the disparate levels revealed as having occurred since the buildings were constructed in 1907 must have occurred prior to 1971, save for those movements that are directly associated with the movement in the party wall and the other localised associated movement that occurred in 1995 and 1996. This is because any other movement could not have occurred without sympathetic movement in the superstructure yet the Malewskis had not observed any such movement since they acquired their house in 1971. Mr Ainsworth, the defendant’s engineer expert, accepted that this conclusion was correct for the obvious movement that had occurred to the rear annex but he was not prepared to make a similar concession for the movement that must have occurred at some stage close to the left hand corner of no 54. This inflexibility was, it became clear, engendered by Mr Ainsworth’s appreciating that his conclusion that the oak was not causative of the damage would be fatally weakened if it became clear that the movement close to the right hand corner of no 54 had occurred historically and before 1971 rather than occurring at the same time as the structural movement that did occur in 1995 and 1996.

33. Secondly, had the vegetation in the front and to the side of no 56 been the cause of the subsidence that actually occurred, that vegetation would also have caused subsidence in the right hand corner of no 56 since that corner was the closest part of the buildings to the cypresses within both nos 56 and 58 that would have been the predominant cause of the damage had the vegetation and not the oak been the culprit. Thus, the absence of damage to the locations closest to each source of desiccation would not appear to rule out either the oak or the other vegetation as desiccation-causing candidates since, if that absence ruled out either candidate, it would rule out both candidates and leave the damage with no source at all.

34. It follows that the unexpected levels revealed by the survey occurred more than 30 years ago and, moreover, those levels neither implicated nor eliminated the oak, the plum or the cypresses. In other words, the subsidence around the party wall occurred in footings which otherwise had reached their present levels some years previously. The level surveys therefore neither added to nor detracted from any of the other evidence and the levels that they revealed may be set aside and should not influence the conclusion as to the cause of the desiccation. The surprising absence of damage in the right hand corner of no 54, despite it being the nearest part of the buildings to the oak, is, as I have already found, an unexplained anomaly.

4.6. The propensities of the relevant trees

35. Although located about 20 - 21m away from the oak, the area that was heavily desiccated could have been brought to that state by the oak in question since oaks of about 14m in height can, on occasion, cause damage to buildings founded in clay up to 30 - 40m away and oak roots can desiccate soil down to about 4m and such oak roots can readily occur at depths down to 2.5m. However, plums and sycamores cannot cause desiccation at depths below 1 - 1.5m and do not root with any vigour much below that depth either. Furthermore, neither of the plum nor the cypresses were located near enough to the affected footings of no 56 to fall within the ordinary danger zone for such trees. Finally, neither the plum nor the cypresses could have caused damage in 1996 after the height reduction of the former and the removal of the latter in 1995, yet significant damage occurred in 1996 as evidenced by the movement in the cracks during the summer months of that year.

4.7. Similar fact evidence of damage to other neighbouring properties

36. It is only of marginal significance but it is clear that nos 46, 48, 50 and 52 Woodfield Road have all suffered from subsidence caused by desiccation of the underlying clay as a result of the action of the oak opposite no 50 in the period 1995 - 2000. These properties were located nearer to the oak and were not susceptible to other desiccating influences. However, the fact that subsidence damage was caused by the oak to the four neighbouring properties to nos 54 and 56 shows that this particular oak is a prime contender for implication as the cause of similar damage to nos 54 and 56.

5. Conclusion

37. When all the evidence is considered together, it is clear that the cause of the relevant desiccation is the oak and that the other trees and vegetation contended for by the defendant did not, or did not materially, cause this damage. The level surveys relied on by the defendant in its attempt to implicate those other trees provides no support to its case and the other evidence, particularly the borehole evidence, the rootings found in the subsoil, the monitoring records and the nature and location of the damage and its associated remedial work, all point irresistibly to the oak as being the culpable sole or predominant source of the desiccation beneath and close to the party wall.

38. It follows that the defendant is liable in nuisance for the damage that has occurred and it is agreed that the claimants should recover damages in the sum of £55,246.00 and interest and that each claimant should additionally recover £1,500 in general damages for inconvenience. Of the sum of £55,246.00, a sum of £1,000 plus interest should be paid to the Malewskis since they have funded this sum themselves as being the excess on their household policy.

HH Judge Thornton QC

Technology and Construction Court

28 March 2003