The effect of catchment liming on bryophytes in upland Welsh streams with an assessment of the communities at risk.код для вставкиСкачать
AQUATIC CONSERVATION: MARINE A N D FRESHWATER ECOSYSTEMS, VOL. 4, 297-306 (1994) The effect of catchment liming on bryophytes in upland Welsh streams, with an assessment of the communities at risk S. M. WILKINSON and S . J. ORMEROD* Catchment Research Group, School of Pure and Applied Biology, University of Wales Cardiff, PO Box 915, Cardiff CFI 3TL, UK ABSTRACT 1 . The effects of catchment liming, used to restore acidified streams, are still only partially known. In particular, the impact on aquatic bryophytes remains unappraised. 2. Six upland streams were surveyed for aquatic bryophytes from 1987-93, and three of their catchments limed experimentally in 1987-88 so that the resulting changes in the abundance and distribution of common bryophyte taxa could be assessed. A n additional nine streams were surveyed during 1987-89, 1992 and 1993 using the same methods to give a wider indication of the bryophyte communities in streams where liming might occur. 3. Twenty-nine bryophyte species were recorded from the wetted margins of these 15 upland streams. Percentage cover by individual species varied markedly, and was highest in Nardia compressa (up to 71%). Cover by most other species was limited t o less than 5%. 4. Total bryophyte cover, and cover by Scapaniu undulata fluctuated from year to year so that no responses t o liming could be detected. However, cover by Nardia compressa declined significantly from 39% t 21 070 SD to 5% ? 2Vo SD in treated streams following lime addition. Reference streams showed no such change, and no other taxa increased in abundance to replace the lost Nardia in the 5 years after liming. 5 . Changes of this type, and their possible ramifications for invertebrates, will have to be considered where catchment liming is planned. INTRODUCTION Until recently, conservation of bryophytes in Great Britain had received scant attention, with many important sites only receiving protection indirectly in areas notified for other taxa and communities (Hodgetts, 1992). Riverine species are among those often dispersed in habitats outside areas formally notified for their conservation value, leaving some populations at risk from the wide range of environmental changes which occur in river systems. So far, however, there are few systematic data on how such changes affect bryophytes, even though many survey data have revealed marked differences in species richness and community composition between rivers of different conservation quality (e.g. Slater et a/., 1987). Aquatic bryophytes occur predominantly in fast-flowing streams, where suitable substrata provide attachment points and where there is sufficient free carbon dioxide in solution to permit photosynthesis (Hynes, 1970). Functionally they are important producers, fixing nutrients, accumulating detritus, and providing sites for periphyton or invertebrate colonization and shelter (Suren, 1991 ; Suren and Winterbourn, 1992; Steinman and Boston, 1993). This means that any change among aquatic bryophytes, in addition to direct conservation impacts, might have ecological consequences for other organisms. *Correspondence Received 5 January 1994 Accepted 1 April 1994 CCC 1052-761 3/94/040297-10 01994 by John Wiley & Sons, Ltd. 121.1 298 S. M. WILKINSON AND S. J . ORMEROD Some of the most widespread ecological changes affecting upland regions in recent decades have been due to acid deposition (Usher and Thompson, 1988). In this case, not only are there conservation ramifications arising from the initial problem, but also from some of the management options which have been suggested in the shorter term (Hildrew and Ormerod, in press). In particular, the spreading of limestone onto catchment wetlands as a treatment for symptoms has probably led to reductions in Sphagnum cover, with associated knock-on effects on other organisms (Mackenzie et al., 1990). At the same time, the need for such measures to protect fisheries is being increasingly emphasized. While there is widespread recognition that emission reductions would bring benefits over extensive areas, the scale of reduction required to protect and restore all sensitive habitats is liable to be substantial and costly (e.g. Reynolds and Ormerod, 1993). Moreover, neither the implementation of control technologies, nor the chemical recovery of affected systems, can be immediate (e.g. Ormerod et al., 1988), so that environmental managers need to consider the options available more readily at the local scale. Experimental applications of limestone have already been made to headwater catchments and surface waters in the British uplands to assess the effectiveness of liming (Diamond et al., 1992; Donald and Stoner, 1992). So far, most considerations of its effects have been restricted to fish and invertebrates. However, recent qualitative evidence suggested that there may be adverse changes among some aquatic bryophytes (Weatherley and Ormerod, 1987; Rundle et al., in press). Hitherto, however, there has been no quantitative assessment of such changes, nor any assessment of the wider communities that might be at risk. This paper investigates changes in the cover by aquatic bryophytes in upland Welsh streams draining a base-poor area which occurred following replicated experimental catchment liming. It also records the wider riparian bryophyte community potentially at risk from chemical change. STUDY AREA AND METHODS The second-order streams under study are located adjacent to the Llyn Brianne reservoir in the catchment of the upper River Tywi, mid-Wales (see Stoner et al., 1984; Rutt et al., 1989). All drain base-poor soils overlying shales, mudstones and grits of Ordovician and Silurian origin, giving rise to soft waters susceptible to acidification (total hardness 3.9-18.8mg CaC03 L-'; Rutt et al., 1989). Experimental manipulation Three stream catchments were treated with CaC03 between autumn 1987 and summer 1988 (C2, C5 and L4, the limed group; see Ormerod et al., 1990 for treatment details), while three adjacent streams were monitored as acidic reference sites (Cl, C3, L1; acid group). Cover and species composition of bryophytes in the wetted zone of each stream were assessed during July and August for 1987-93 following the method of Rutt et al. (1989): a 0.5 x 0.5 m quadrat divided into 25 sub-quadrats (0.1 x 0.1 m) was placed randomly on the left, right or mid-channel at 20 points, on a stratified random basis, over a 200m reach of stream. The number of sub-quadrats occupied by each taxon was recorded and from this their percentage cover, and the total percentage cover of all bryophytes, were derived. Other taxa observed in the 200m reach but not within the quadrats were assigned a nominal 0.01 To cover value. Voucher specimens were collected for subsequent identification and verification in the laboratory. Using the same quadrat method, the physical composition of each stream bed was assessed for the percentage cover of seven defined types of substratum (Table 1). In addition, calcium, aluminium, conductivity, pH and dissolved organic carbon (DOC) were measured as weekly spot samples throughout the period (see Weatherley and Ormerod, 1987, for details). However, only data from the 12-month post-liming period are used here, these providing a typical period from which subsequent years have not departed strongly (Rundle et a/., in press). LIMING AND AQUATIC BRYOPHYTES 299 Table 1. Substratum categories used in habitat survey (size categories according to Wentworth Scale). Category Size (mm) Bedrock Boulders Cobbles Pebbles Gravel Sand Silt/mud > 256 > 64-256 > 16-64 >2-16 > 0.0625-2 C0.0625 General bryophyte distribution Nine additional streams in the Llyn Brianne area were surveyed for bryophytes and physico-chemical variables, using the same methods, during 1987-89, 1992 and 1993 in order to provide data on typical bryophyte communities from a wider array of upland streams. Analysis Prior to any statistical analysis, all data were transformed (loglo ( x + 1) or arcsin square root (percentage cover) ) to normalize distributions. Percentage cover by all bryophytes, and by selected common species, were compared between streams, groups of streams (limed versus unlimed) and years using a crossed multifactorial analysis of variance (ANOVA) based on the general linear model (GLM, Ryan et al., 1985). Correlations between cover values and physico-chemical variables were calculated to investigate possible relationships (Spearman-ranked correlations; Sokal and Rohlf, 1981). RESULTS General bryophyte distribution and abundance Twenty-nine bryophyte species, including four unconfirmed, were recorded during the survey from the 15 streams (Table 2). Percentage cover by each species within the wetted margin vaned markedly between streams: most were present at very low cover, with only five species covering more than 10% of any stream bed. Nardia compressa exhibited the greatest cover in any one stream (71Yo), followed by Rhynchostegium riparioides (440/0),Polytrichum commune (26Vo), Scapania undulata (23%), and Hygrohypnum ochraceum (16%). Streams fell into three well-defined groups according to pH (‘acid’: mean pH = 5.23 f 0.38 SD, n = 9; ‘circumneutral’: mean = 6.79 f0.43, n = 3; and ‘limed’: mean = 6.12 f0.67, n = 3). Nine taxa were common to all (Table 2, pre-treatment data for limed streams omitted), with N . compressa occurring on 40 of 51 (78%) sampling occasions in acid streams, but on only 6 of 15 (40%) sampling occasions in circumneutral, and 6 of 18 (33%) in limed streams (post-liming). Five taxa were restricted to acid streams alone, a further eight occurred solely in circumneutral streams, whilst another seven were present in both circumneutral and acid streams. Nardia cover tended to decrease in pebble-rich streams, although this trend was not formally significant (Table 3). Instead, cover by Nardia was correlated negatively with stream pH (mean r = - 0.7197, ~ ~ 0 . 0 1 ; Table 3), and positively with aluminium concentration (mean r = 0.6255, p<0.05). Experimental streams: liming effects Acid streams contained significantly less dissolved calcium than limed streams (acid mean = 1.59 _+ 0.73 SD mg L- l , limed mean = 4.61 f 1.69 mg L- l ; one-way ANOVA, Fl,5= 10.25, p = 0.003). Otherwise, 300 S. M. WILKINSON AND S. J . ORMEROD Table 2. Major brvouhyte species recorded from 15 upland streams in mid-Wales, 1987-93. Speciesa =Nardiacompressa S. Gray ‘Scapania undulata Dum. ‘Sphagnum lescurii Sull. ePellia epiphylla Corda eHyocomium armoricurn Wijk & Marg. ‘Polytrichum commune Hedw. CRacomitriumaciculare Brid. dSphagnum papillosurn Lindb. dSphagnum cuspidatum Ehrh. ex Hoffm. ‘Dichodontium pellucidurn Schimp. CHygrohypnumochraceum Loeske ‘Rhynchostegium riparioides Card. dMnium hornum Hedw. dBrachytheciurn rivulare B., S. & G. dCinclidotusfontinaloides P. Beauv. ‘Gymnocolea inflata Durn. dDicranum scoparium Hedw. dHomalothecium sericeum B., S. & G. eDicranum majus Sm. dRacornitrium aquaticum Brid. CThamnobryumalopecururn Nieuwl. dLeucobryum glaucum Angstr. cNardia scalaris S . Gray R hytidiadeiphus loreus War nst . dThuidium tamariscinum B., S . & G. Preferred habitat Aquatic Aquatic Semi-aquatic Semi-aquatic Semi-aquatic Semi-aquatic Aquatic Semi-aquatic Semi-aquatic Semi-aquatic Semi-aquatic Aquatic Moorland Aquatic Semi-aquatic Semi-aquatic Moorland Moorland Moorland Semi-aquatic Semi-aquatic Moorland Moorland Moorland Moorland 9 acid streams 3 circumneutral streams 3 limed streamsb (15 samples) (18 samples) (51 samples) mean mean mean occurrence cover occurrence cover occurrence cover (Too) (70) (To) (Too) (To) (To) <1 <I 78.4 68.6 51.0 35.3 33.3 17.6 17.6 13.7 7.8 5.9 5.9 3.9 3.9 3.9 2.0 2.0 2.0 2.0 <I 5.9 10.8 3.0 <1 <I <I <I <1 <I <1 <I <1 <1 <I <I <I <I - - - - - - <I <I <I <1 <I <I <1 <I <1 40.0 73.3 6.7 40.0 40.0 26.7 46.7 6.7 6.7 - - 1.9 9.1 6.7 93.3 - - <I 13.3 - - <1 <1 <I <1 <1 <I <I <I <I <1 <1 <1 <1 <1 <I <I <1 - 33.3 50.0 72.2 27.8 33.3 11.1 5.6 5.6 - - 13.3 20.0 20.0 - 6.7 6.7 6.7 6.7 6.7 “Source of nomenclature: Grolle (1983) and Corley et al. (1981). bPre-liming data (1987) omitted. ‘Pre-1993 record. d1993 record. eBoth 1993 and pre-1993 record. Table 3. Mean ( 5 years) Spearman-ranked correlation coefficients between percentage cover values for dominant bryophyte species and physico-chemical variables (12 sites). To To To To To To Bedrock Boulder Cobble Pebble Gravel Mud PH Conductivity A1u mini u m Calcium Nardia (To) Scapania (To) 0.3705 -0.1557 - 0.1974 - 0.5392 - 0.0440 - 0.3679 - 0.7197** 0.2175 0.6255* -0.1974 0.1596 0.0520 0.0030 0.1993 - 0.2784 - 0.1664 - 0.3443 0.2098 0.4738 0.0280 Total bryophytes (To) 0.3435 - 0.0709 -0.1244 - 0.3636 0.0480 - 0.3346 -0.1868 0.2858 0.0749 0.0080 Transformations prior to correlation: arcsin. square root (percentage values); otherwise Log,, (x). Mean correlation coefficients derived by averaging z-transformation of individual year’s coefficients (Sokal and Rohlf, 1981). *Significant at 5 % , **significant at 1%. 301 LIMING AND AQUATIC BRYOPHYTES between group differences were limited to pH (one-way ANOVA, Fl,5= 15.08, p=O.OOl). All physicochemical data are given in Appendix 1 . Statistical comparisons of bryophyte cover between years and groups (limed versus acid) were limited by data availability to N. compressa, S. undulata, and total bryophyte cover (Table 4). Over the 7 years of the study, Nardia cover varied significantly between streams and years principally due to greater values in acid streams (Figure 1). On exclusion of the data for the three 'limed' streams pre-liming (1987), or all data for 1987, the significant between-year effect disappeared. This indicated that the differences in Nardia cover between limed and acid streams, and between years, were due to changes in treatment streams that accompanied liming. Despite a near-absence pre-liming (0.01 (70cover) in one limed stream, time-series analysis confirmed marked reductions in Nardia cover in the limed group: it was originally very common in two streams (41%-77% in 1987) where it fell to 7% and zero respectively (1988). Like Nardia, S. undulata cover over 7 years was significantly greater in acid streams than in limed. However, unlike Nardia there were no year to year effects related to liming, nor any evidence that this species declined post-liming (Figure 2). In other words, post-treatment differences for Scapania between limed and acid streams were probably unrelated to liming since they also occurred prior to treatment. Nardia made an important contribution to bryophyte cover in experimental streams (up to 98% of the total) and, as a result, total bryophyte cover varied significantly between years and streams (Figure 3). On removal of Nardia from the total cover values, no significant differences between years or streams arose, suggesting that this species was responsible for these patterns. DISCUSSION The 15 second-order streams at Llyn Brianne supported 29 bryophyte species in total, although most were scarce. This is a small number of species compared with other sites in Wales (e.g. Slater eta/., 1987), reflecting the concentration of our efforts in or near the stream wetted perimeter. It also reflects the exposed nature of many of the sites in our study. The most prolific species was the liverwort Nardia compressa, whose presence increased with stream acidity. Experimental evidence highlighted a clear decline in its abundance as Table 4. Percentage cover differences between bryophytes in acid streams (n = 3) and limed streams (n = 3) (GLM analysis). Data Total To cover by all bryophytes All years Excluding 1987 Excluding 1987 limed sitesa Vo cover by Nardia compressa All years Excluding 1987 Excluding 1987 limed sitesa 070 cover by Scapania undulata All years Excluding 1987 Excluding 1987 limed sitesa "Excludingpre-liming data from limed sites only. Effect Stream Year Stream Year Stream Year Stream Year Stream Year Stream Year Stream Year Stream Year Stream Year df F P 5 8.27 4.28 1 1.56 2.69 11.98 2.62 16.97 3.67 30.36 1.24 30.19 1.67 4.04 0.71 7.67 1.08 4.55 0.73 0.001 0.003 6 5 5 5 6 5 6 5 5 5 6 5 6 5 5 5 6 O.OO0 0.044 O.OO0 0.039 0.000 0.008 O.Oo0 0.318 O.Oo0 0.166 0.006 0.643 O.Oo0 0.395 0.004 0.626 302 S. M. WILKINSON A N D S. J . ORMEROD ! 87 88 89 90 91 92 93 Year Figure 1. Mean percentage cover (with SE) of Nardiu compressa in three acid streams ( A ) and three limed streams (0)over 7 years. Liming occurred in autumn 1987 or spring 1988 (see text). a response to catchment liming in three streams. No other taxa increased in abundance to replace it as the dominant large plant. At the same location, year to year changes in the abundance of Scapania undufufa, although evident, were not necessarily a response to liming. No other species was present in sufficient abundance to allow any assessment of change due to liming, although the bryophyte community in and around the base-poor streams indicates that there is scope for such effects to occur. Causal mechanisms Assuming the decline in Nardia was a real response to liming, what mechanisms might be responsible? The inability of aquatic bryophytes to photosynthesize hydrogen carbonate ions directly is well known, and means that they depend on dissolved, free C 0 2 as a source of photosynthetic carbon (Gessner, cited from Hynes, 1970). Since C 0 2 gas diffuses through water lo5 times slower than air, the availability of free C 0 2 to aquatic plants is markedly reduced (Hynes, 1970). This problem is exacerbated in calcareous streams, where HCO; is the dominant form of inorganic carbon, resulting in loss of C 0 2 from solution. Changes 87 88 89 90 91 92 93 Year Figure 2. Mean percentage cover (with SE) of Scupania undulutu in three acid streams ( A ) and three limed streams (0)over 7 years. 303 LIMING AND AQUATIC BRYOPHYTES a7 aa 89 90 91 Year Figure 3. Mean percentage cover (with SE) of all bryophytes in three acid streams 92 (A ) 93 and three limed streams ( 0 )over 7 years. such as this in the HCO; /COz equilibrium have important consequences for those bryophyte species present (Bain and Proctor, 1980). In this study, the addition of lime to acidic streams is liable to have caused a shift towards HCO; as the dominant form of inorganic carbon, mimicking more closely the situation in calcareous streams. Chemical conditions conducive to this altered equilibrium have persisted for at least 6 years following liming at Llyn Brianne (Rundle et ul., in press) during which time the most prolific species, Nurdiu compressu, has been all but eliminated. Although we have no information on the sensitivity of this species to altered C 0 2 concentrations, reductions in its abundance following liming are consistent with its scarcity in circumneutral streams (Hill, 1988), and consistent also with effects by the above mechanism. However, the effects of liming on stream PCOz are far from clearly known (C. Neal, personal communication), and other chemical effects on Nurdiu cannot yet be ruled out. Ecological significance Whereas aquatic bryophytes are likely to support a range of epiphytic organisms, including diatoms, their value as habitat for invertebrates has been more fully documented (Eglund, 1991; Suren and Winterbourn, 1991, 1992; Steinman and Boston, 1993). Within bryophytic mats, water velocity is reduced, and large quantities of periphyton and detritus can accumulate; two corrolaries are the provision of shelter and food for associated organisms. Greatly enhanced invertebrate densities occur probably as a result (Suren, 1991), although some invertebrates also utilize bryophytes directly as a food source (Suren and Winterbourn, 1992). At Llyn Brianne, however, no gross change in invertebrate abundance occurred as a result of liming (Rundle et ul., in press) and so the consequences of reduced Nurdiu cover appeared unimportant to such organisms at the stream level. However, the effects of liming on relationships between Nurdiu and invertebrates at the micro-habitat level have not yet been investigated. 304 S . M. WILKINSON AND S . J . ORMEROD Nature conservation importance In this study, many of the species recorded can be described as common, at least regionally in Wales (Nardia compressa, Scapania undulata, Sphagnum cuspidatum; Hill, 1988), although some of them are restricted to acidic habitats. In the case of Nardia, effects by limited liming would be unlikely to have major implications for gross range and abundance, but would nevertheless represent a local change in a representative community. Other, less frequent species in the aquatic community at Llyn Brianne (e.g. Racomitrium aquaticum) were too sparse to allow any assessments of effects by liming; this feature may repeatedly prevent the detection of changes amongst such rare taxa generally unless future work involved mapping in fixed quadrats in which they were known to occur. In fact, one lesson of hindsight from this study was that only such a method would have detected change among these rarer taxa. Given that many bryophyte species have restricted range, and often highly specific habitat requirements, such a targeted methodology might be profitably employed in future studies where the aim is to detect change due to environmental perturbation. The wider effects of most of the strategies aimed at reducing surface water acidification are as yet unclear, or are known solely from modelling studies (Ormerod et a/., 1990). This is unfortunate, precluding a rigorous environmental cost-benefit appraisal of any individual option (Hildrew and Ormerod, in press). In the case of liming, effects on aquatic invertebrate assemblages have been minimal (Rundle et a/.,in press), and the water quality conditions created by liming do not represent those which occurred prior to acidification (Ormerod et al., 1990). Fish species richness tends not to increase (Degerman and Appelberg, 1992), but there have sometimes been benefits to salmonid density (Weatherley, 1988 cf Diamond et a/., 1992) and the quality of water supplies. As a result, both water undertakings and conservation agencies have prescribed liming to protect fish populations that are valuable in economic and conservation terms (e.g. Farmer, 1992; see EU Directive 92/43/EEC). At the same time, some conservationists have drawn attention to the possible adverse consequences of lime additions to terrestrial ecosystems which are naturally oligotrophic, base-poor, and contain resources of intrinsic conservation value (Woodin and Skiba, 1990; Farmer, 1992). The response of an important and widespread bryophyte in upland acid streams, Nardia compressa, must now be considered as a further environmental cost in any future appraisals of lime additions. Appendix 1. Physico-chemical data for the 12-month post-liming period. Unless otherwise stated, all values are in mg L-I, except pH and conductivity (pS cm-I). Stream L1 L2 L3 L4* L5 L6 L7 L8 c1 c2* c3 c4 c5* uc4 G1 *Limed. Bedrock Boulder Cobble Pebble Gravel Mud Dissolved (Vo) (Yo) (Vo) (Yo) (Yo) (Yo) pH Conductivity Calcium Aluminium organic carbon 26.0 33.8 23.2 3.2 4.0 14.0 6.8 5.5 8.2 8.7 15.1 0.0 0.9 0.0 31.7 4.8 6.6 5.3 0.5 2.5 12.3 9.8 8.4 5.4 0.5 0.5 0.4 0.6 16.1 1.3 28.3 27.8 39.3 40.8 31.9 40.1 41.7 52.5 46.1 42.0 34.2 52.1 22.2 38.7 29.3 19.8 20.8 29.2 46.4 37.8 24.2 27.8 30.0 26.8 36.5 39.9 44.2 31.3 36.5 36.1 5.2 9.5 3.0 7.2 9.6 9.4 13.2 3.6 10.7 2.3 4.6 1.4 18.1 7.6 1.6 0.0 0.0 0.0 1.6 14.2 0.0 0.0 0.0 0.0 0.0 0.0 1.9 5.0 1.0 0.0 4.71 4.78 5.05 6.43 5.84 6.93 7.13 5.27 5.12 5.35 5.26 5.30 6.57 5.74 6.31 48.0 51.8 55.4 64.0 36.0 48.3 60.5 36.4 34.5 46.1 31.9 36.4 45.7 30.3 41.2 1.40 1.59 2.36 6.41 1.46 3.27 5.39 3.20 0.96 3.06 1.05 1.13 4.35 1.17 2.03 0.470 0.645 0.357 0.074 0.056 0.059 0.038 0.204 0.095 0.155 0.140 0.154 0.052 0.074 0.046 0.88 1.26 1.26 0.98 0.45 1.12 1.30 3.60 3.52 3.48 3.62 4.91 3.45 5.17 3.48 LIMING AND AQUATIC BRYOPHYTES 305 ACKNOWLEDGEMENTS Thanks are due to G. P. Rutt, E. C. Lloyd, H. Thomas, M. Jenkins and Drs K. Davies, N. S. Weatherley, S. D. Rundle and N. Stringer for assistance with fieldwork and bryophyte taxonomy. Financial assistance was provided by the Welsh Office, the National Rivers Authority and the Department of the Environment. This study formed part of the Llyn Brianne Project. REFERENCES Bain, J. T. and Proctor, M. C. F. 1980. ‘The requirement of aquatic bryophytes for free CO, as an inorganic carbon source: some experimental evidence’, The New Phytologist, 86, 393-400. Corley, M. F. V., Crundwell, A. C., Dull, R., Hill, M. 0. and Smith, A. J. E. 1981. ‘Mosses of Europe and the Azores’, Journal of Bryology, 11, 609-689. Degerman, E. and Appelberg, M. 1992. ‘The response of stream-dwelling fish to liming’, Environmental Pollution, 78. 149-155. Diamond, M., Hirst, D., Winder, L., Crawshaw, D. H. and Prigg, R. F. 1992. ‘The effect of liming agricultural land on the chemistry and biology of the River Esk, north-west England’, Environmental Pollution, 78, 179-185. Donald, A. P. and Stoner, A. S. 1992. ‘Acid waters in upland Wales: causes, effects and remedies’, Envir~nmental Pollution, 7 8 , 141-148. Eglund, G. 1991. ‘Effects of disturbance on stream moss and invertebrate community structure’, Journal of the North American Benthological Society, 10, 143-153. Farmer, A. M. 1992. ‘Catchment liming and nature conservation’, Land Use Policy, January 1992, 8-10. Grolle, R. 1983. ‘Hepatics of Europe including the Azores: an annotated list of species, with synonyms from the recent literature’, Journal of Bryology, 12, 403-459. Hildrew, A. G.and Ormerod, S. J. in press. ‘Surface water acidification: consequences and solutions’, in Harper, D., Ferguson, A. and Edwards, R. W. (Eds), The Ecological Basis of River Management, John Wiley, Chichester. Hill, M. 0. 1988. ‘A bryophyte flora of North Wales’, Journal of Bryology, 15, 377-491. Hodgetts, N. G. 1992. ‘Measures to protect bryophytes in Great Britain’, Biological Conservation, 52, 259-264. Hynes, H. B. N. 1970. The Ecology of Running Waters, Liverpool University Press, Liverpool. Mackenzie, S. M., Lee, J. A. and Wright, J. M. (1990). EcologicalImpact of Liming Blanket Bog. Report to the Nature Conservancy Council, Peterborough. Ormerod, S. J., Weatherley, N. S., Merret, W. J., Gee, A. S. and Whitehead, P. G. 1990. ‘Restoring acidified streams in upland Wales: a modelling comparison on the chemical and biological effects of liming and reduced sulphate deposition’, Environmental Pollution, 64, 67-85. Ormerod, S. J., Weatherley, N. S., Varallo, P. V. and Whitehead, P. 1988. ‘Preliminary empirical models of the historical and future impact of acidification on the ecology of Welsh streams’, Freshwater Biology, 20, 127-140. Reynolds, B. and Ormerod, S. J. 1993. A Review of the Impact of Current and Future Acid Deposition in Wales. Report to the Welsh Office and the Countryside Council for Wales, Institute of Terrestrial Ecology, Bangor. Rundle, S. D., Weatherley, N. S. and Ormerod, S. J. in press. ‘The effects of catchment liming on the chemistry and biology of upland Welsh streams: testing predictions from empirical models’, Freshwater Biology. Rutt, G. P., Weatherley, N. S. and Ormerod, S. J. 1989. ‘Microhabitat availability in Welsh moorland and forest streams as a determinant of macroinvertebrate distribution’, Freshwater Biology, 22, 247-261. Ryan, B. F., Joiner, B. L. and Ryan, T. A. 1985. MINITAB Handbook, 2nd edn, PWS-Kent Publishing Company, Boston. Slater, F. M., Curry, P. and Chadwell, C. 1987. ‘A practical approach to the evaluation of the conservation status of vegetation in river corridors in Wales’, Biological Conservation, 40, 53-60. Sokal, R. R. and Rohlf, F. J. 1981. Biometry, Freeman, San Francisco. Steinman, A. D. and Boston, H. L. 1993. ‘The ecological role of aquatic bryophytes in a woodland stream’, Journal of the North American Benthological Society, 12, 17-26. Stoner, J. H., Wade, K. R. and Gee, A. S. 1984. ‘The effects of acidification on the ecology of streams in the upper Tywi catchment in west Wales’, Environmental Pollution (Series A), 35, 125-157. Suren, A. M. 1991. ‘Assessment of artificial bryophytes for invertebrate sampling in two New Zealand alpine streams’, New Zealand Journal of Marine and Freshwater Research, 25, 101-112. Suren, A. M. and Winterbourn, M. J. 1991. ‘Consumption of aquatic bryophytes by alpine stream invertebrates in New Zealand’, New Zealand Journal of Marine and Freshwater Research, 25, 331-343. Suren, A. M. and Winterbourn, M. J. 1992. ‘The influence of periphyton, detritus and shelter on invertebrate colonisation of aquatic bryophytes’, Freshwater Biology, 27, 327-339. 306 S. M. WILKINSON AND S. J . ORMEROD Usher, M. B. and Thompson, D. B. A. Eds. 1988. Ecological Change in the Uplands. Blackwell Scientific Publications, Oxford. Weatherley, N. S. 1988. ‘Liming to mitigate acidification in freshwater ecosystems: a review of the biological consequences’, Water, Air and Soil Pollution, 39, 421-437. Weatherley, N. S. and Ormerod, S. J. 1987. ‘The impact of acidification on macroinvertebrate assemblages in Welsh streams: towards an empirical model’, Environmental Pollution, 46, 223-240. Woodin, S. J. and Skiba, U . 1990. ‘Liming fails the acid test’, New Scientist, 1707, 50-53.