Portal de Periódicos da CAPES

Myrica gale L.

2000; Wiley; Volume: 88; Issue: 6 Linguagem: Inglês

10.1046/j.1365-2745.2000.00522.x

1365-2745

Keith R. Skene, Janet I. Sprent, John A. Raven, Lindsey Herdman,

Flowering Plant Growth and Cultivation

Deciduous, much branched shrub, 60–150 (to 250) cm in height. Main stems brown, twigs reddish-brown with scattered shiny yellowish glands, becoming dark, ascending at an angle. Young twigs downy. Roots nodulated with cluster roots; suckering habit. Buds small, ovoid, obtuse, reddish-brown, with several scales (Fl. Br. Isl.). Leaves alternate. Leaf laminae 2–6 cm long, petiole 3–5 mm long. Leaves oblanceolate or oblong-obovate in outline; subsessile. Base of blade narrowly wedge-shaped, tapering into the stalk. Margin coarsely toothed in upper third, entire below and undulate. Adaxial surface dull dark green with fine hairs. Abaxial side pale green with fine hairs on a prominent midrib. Conspicuous, scattered, shining yellowish resin glands on both surfaces. Glandular trichomes can be sessile or stalked. Stomata hypostomatous, anomocytic, venation camptodromous (leaf venation in which secondary veins bend forward and anastomose before reaching the margin of the leaf). Wood and leaves fragrant when bruised. Flowers are borne on the bare wood of the previous year's growth and appear before the leaves. Myrica gale is, for the most part, dioecious, but, within a single population, monoecious and hermaphrodite flowers may also occur (Davey & Gibson 1917), along with a skewed sex ratio of stems (Lloyd 1981). The male catkins are unbranched, usually about 10 mm long with red-brown bracts, borne on leafless branches of the previous year's growth, usually in May or June. Bracteoles 0, stamens c. 4, anthers red. Female catkins smaller (6–7 mm), but thicker and closely set, with green bracts. The female flower is a gynoecium with 2 stigmas, 2 bracteoles, styles red and a single, basal unitegmic ovule. The ovary at pollination is superior, and very small. Following pollination, the bracts are raised up, owing to intercalary growth below the two transverse bracts, beneath the base of the gynoecium and around the locule. Thus, during fruit maturation, the ovary becomes inferior. Intercalary growth is significant in fruit wall formation. The female and male inflorescences are among the least branched of the Myricaceae, a trend that seems to correlate with an increasingly northern distribution. Fruit a nut, dry, flattened, gland-dotted, the adnate, accrescent bracteoles forming 2 wings, the exocarp secreting wax. Air-dry mass of nut (± standard error of the mean, n = 250) = 1.56 ± 0.75 mg. The genus Myrica includes some 60 species, among which M. gale is taxonomically isolated and has been placed as the sole representative of the subgenus Gale, sensuChevalier (1901). Sundberg (1985) concluded that evidence does not support raising M. gale to separate generic status. Recent molecular evidence places Myrica in a rosid clade with a predisposition towards nodule formation. This clade includes, inter alia, all actinorhizal plants and all legumes (Soltis et al. 1995). One variety, Myrica gale L. var. tomentosa C. D.C., is recognized. Native, in bogs, wet heaths and fens. A distinctive member of the British flora, particularly in spring, with its striking red-brown buds, red catkins and flowers. Its role as a nitrogen fixer has a significant impact on the communities of which it is a part. Myrica gale is found throughout the British Isles, with the exception of the Channel Isles and the Shetland Isles (Fig. 1). It is absent, however, from a number of counties, particularly in south-east Scotland, the Midlands, north-east England and south-east England. In north-west and north Scotland, where it is widespread, it extends from adjacent, acid peat substrata into marshes and sea lochs, showing some tolerance to salinity. It is a lowland plant with the highest occurrence in oceanic regions, ascending to 370 m in the Lake District, England, 430 m near Capel Curig, Wales, 460 m on Brandon, Kerry, Ireland, and 520 m in the Forest of Drumochter, Scotland (Alt. Range Br. Pl.). The distribution of Myrica gale in the British Isles. (○) Pre-1950; (●) 1950 onwards. Each dot represents at least one record in a 10-km square of the National Grid. Mapped by Mrs J. M. Croft, Biological Records Centre, Institute of Terrestrial Ecology, mainly from records collected by members of the Botanical Society of the British Isles. The Myricaceae are distributed world-wide on all the major land masses, with the exception of Australia, New Zealand and Antarctica (MacDonald 1989). Myrica gale is widely spread in the Northern Hemisphere, but there are large gaps in its distribution. It is scattered throughout northern North America as far south as Virginia, and in Asia to the Far East. In Europe (Fig. 2), M. gale has a decidedly atlantic and north-western distribution (Holm & Holm 1991), ranging from north-west Spain to central Germany and north-west Russia, including the coastal zone of the Baltic Sea (Behre 1999). It also grows around the edges of inland lakes in Finland (Svoboda et al. 1998). It inhabits riverbanks and freshwater ponds in temperate regions of North America (Berliner & Torrey 1989). Myrica gale L. var. tomentosa C. D.C. is distributed in Northern Japan, Sakhalin and Eastern Siberia. Preston & Hill (1997) cite M. gale as Suboceanic Boreo-temperate. The European range of Myrica gale based on Dist. Br. Fl. Myrica gale is often associated with oceanic climates, where rainfall can approach 2000 mm per year with at least 200 wet days in any given year. Ellenberg (1988) classifies M. gale as a light loving plant, rarely found where there is less than 40% relative light. However, it has also been reported as often growing in shade (e.g. Svoboda et al. 1998). It is a wet site indicator. It grows in bogs and also on their marginal slopes. Myrica gale grows in damp acidic soil, but also on fen peat. It is abundant on swamps, wet heath, bogs and wet upland (but not extreme upland) moors throughout the British Isles. It typically occurs in low level blanket bogs, or on steeper slopes, where the peat is shallow (50–80 cm) and of pH 3.8–6.1. In the east-central Highlands, it appears to become more exacting in soil requirements inland, being found where the soil is irrigated by mineral-rich water (Pl. Comm. Scot.). Nitrogen fixation allows M. gale to grow on soils with low available nitrogen. Newbould & Gorham (1956) reported that M. gale in the New Forest was found in soil water of (± standard error of the mean) pH 5.18 ± 0.15 (n = 28) and specific conductivity of 114 ± 4.5 µS cm−1 (equivalent to 0.665 kg m−3 salts) (n = 28). Myrica gale occurs as scattered clumps in the Menyanthes trifoliata subcommunity of the Cladium mariscus swamp and sedge beds (NVC code S2) (Rodwell 1995), and also occasionally in the Cladium mariscus subcommunity. In the Phragmites australis–Peucedanum palustre tall herb fen (S24), M. gale forms a distinctive, open vegetation in the M. gale subcommunity. From here, it often invades Schoenus nigricans subcommunities upon the cessation of mowing, for example in Broadland, East Anglia. This leads to the formation of the Myrica subcommunity of the Peucedano-Phragmitetum, followed by a Myrica scrub. This may eventually progress to a Betula–Molinia woodland (Wheeler 1980). Myrica gale occurs occasionally in the Cladium mariscus subcommunity of the Phragmites australis–Eupatorium cannabinum tall herb fen (S25) and occasionally in the Sphagnum auriculatum (M1) bog pool community (Rodwell 1991b). It occurs in some stands of the Sphagnum cuspidatum/recurvum bog pool community (M2), being restricted to the Rhynchospora alba subcommunity. Myrica gale is found, with local abundance, in some stands in the Carex rostrata–Sphagnum squarrosum mire community (M5), where increased occurrence is accompanied by greater abundance of Carex lasiocarpa (Spence in Veg. Scot., 1964, pp. 306–425; Birse 1980). In the Carex echinata–Sphagnum recurvum/auriculatum mire (M6), M. gale is common, along with Erica tetralix, particularly in the Sphagnum auriculatum variant of the Juncus acutiflorus subcommunity. In the Carex dioica (C. viridula ssp. oedocarpa)–Pinguicula vulgaris mire (M10), M. gale can be locally prominent in the Schoenus nigricans variant of the Carex demissa–Juncus bulbosus/kochii subcommunity. In the Schoenus nigricans–Juncus subnodulosus mire (M13), some stands of the Festuca rubra–Juncus acutiflorus subcommunity have a patchy canopy of M. gale, whereas in the Briza media–Pinguicula vulgaris subcommunity, M. gale is represented only by scattered bushes. Myrica gale forms a low bushy canopy in some stands of the Schoenus nigricans–Narthecium ossifragum mire community (M14). In the Scirpus cespitosus–Erica tetralix wet heath community (M15), M. gale occurs frequently in the Carex panicea subcommunity, often forming an extensive, low canopy. Under denser canopies, associates tend to thin out, presumably owing to competition for light. In the Cladonia spp. subcommunity, M. gale is very scarce. Its elimination from the Carex panicea subcommunity may have led to the derivation of other Scirpus–Erica wet heaths from the Carex panicea subcommunity. The presence of Myrica in the Scirpus cespitosus–Erica tetralix wet heath community is linked closely to the presence of moving water. Myrica gale occurs occasionally in the Erica tetralix–Sphagnum compactum wet heath (M16), with local abundance in the Succisa pratensis–Carex panicea subcommunity. In Scirpus cespitosus–Eriophorum vaginatum blanket mire communities (M17), M. gale is largely confined to the wetter Drosera rotundifolia–Sphagnum spp. subcommunity, although it can be entirely absent from it, as on Rhum (Pl. Comm. Scot.), and, even when present, is patchy. Where it is abundant, it can spread into neighbouring Scirpus–Erica wet heath communities, masking floristic differences between smaller associates. It occurs occasionally in the wetter Sphagnum magellanicum–Andromeda polifolia subcommunity, increasing in soligenous zones, though remaining patchy. In the Narthecium ossifragum–Sphagnum papillosum valley mire (M21), M. gale is restricted, but when present, is very abundant, spreading to neighbouring communities, particularly Junco-Molinion grasslands. It is slightly more abundant in the Rhynchospora alba–Sphagnum auriculatum subcommunity than in the Vaccinium oxycoccos–Sphagnum auriculatum subcommunity. In the Molinia caerulea–Potentilla erecta mire (M25), M. gale is uncommon in the Angelica sylvestris subcommunity, but attains local prominence in the Anthoxanthum odoratum subcommunity. It has been suggested that here, the additional shade offered by M. gale, along with the luxuriant cover of Molinia in response to N enrichment from N-fixation by M. gale nodules, lead to the associated vegetation being species-poor (Rodwell 1991b). Myrica gale and Molinia caerulea vary in their occurrence in the Salix cinerea–Betula pubescens–Phragmites australis woodland community (W2) (Rodwell 1991a), being more abundant in the Sphagnum spp. subcommunity than in the Alnus–Filipendula subcommunity. This may echo inherited differences in preceding fen vegetation. Myrica gale and Salix repens form a patchy lower tier below the canopy. In the Betula pubescens–Molinia caerulea woodland community (W4), M. gale occurs occasionally as a fairly thick cover of leggy bushes. It also occurs in the Dryopteris dilatata–Rubus fruticosus subcommunity and the Sphagnum spp. subcommunity. Myrica gale replaces Molinia caerulea as the dominant species in Molinia–Myrica communities in the east central Highlands (Pl. Comm. Scot.). In mainland Europe, M. gale occurs in raised bog hummocks in the north-west (Class Oxycocco-Sphagnetea, Order Sphagnetalia magellanici, Alliance Sphagnion magellanici (Ellenberg 1988)). In the raised bogs south of the Lower Elbe, M. gale is found in association with Calluna vulgaris, Erica tetralix, Eriophorum angustifolium, Eriophorum vaginatum, Vaccinium oxycoccos, Carex limosa, Sphagnum cuspidatum and S. pulchrum (Ellenberg 1988). The marginal slopes of raised bogs in the north-west of central Europe give rise to a swamp wood-like community with Myrica gale, Calluna vulgaris, Erica tetralix and Empetrum nigrum. Myrica gale is also found in the mire willow scrub (Class Alnetea glutinosae, Order Salicetalia auritae, Alliance Frangulo-Salicion auritae (Salicion cinerea)), associated with Frangula alnus, Salix aurita, S. repens ssp. rosmarinifolia and S. repens ssp. repens. On the Polish (Baltic) coast, M. gale plays a surprisingly large part in the sand dune succession, particularly at the bog margin (Ellenberg 1988). Schaminée et al. (1995) record M. gale as occurring in the following associations in the Netherlands: Pallavicinio-Sphagnetum (9Aa2); Carici curtae-Agrostietum caninae (9Aa3); Eriophoro-Caricetum lasiocarpae (10Ab1). Oberdorfer (Pfl. Exk.) places M. gale as a scrub species (Salicion cinereae). Westhoff & Den Held (1969) mention M. gale under Parvocaricetae (as a differential species against Scheuchzerietea), and Ericetalis tetralicis, but also sees it as mainly a wet acid scrub species (Franguletea : Salicion cinereae), and in Betuletum pubescentis. While M. gale can grow to 250 cm in height (Simpson et al. 1996), grazing often reduces this to 50 cm. Goats reduce the height of M. gale to a level of less than half the height produced by sheep (Fisher et al. 1994). Hjältén (1992) showed that mountain hares did not differentially browse male or female plants of M. gale, in spite of a skewed sex ratio prevailing and the different digestibilities of the bark of the different sexes. Myrica gale has a number of fungal pathogens, and insect herbivores (see Section IX). The oil of M. gale appears to have adverse effects on insects when consumed (see Section IX). The substantial suckering of M. gale leads to an aggregated distribution. Such aggregations can result in associates dying out, probably because of shading. This is observed in the Carex panicea subcommunity of the Scirpus cespitosus–Erica tetralix wet heath community. A highly significant negative correlation exists between the cover of M. gale and the CO2 and H2S concentrations in the ground water (Webster 1962). Plant growth is greatest in wet, but well aerated soil (Sprent & Scott 1979). Sprent & Scott (1979) have suggested that this would support the argument that these plants may grow in 'boggy' sites because they are tolerant of, rather than being favoured by, these conditions. Ellenberg (1988) suggests that M. gale grows in damp soils because there is less gas space in the topsoil. He relates this to his observation that M. gale is frost sensitive in central Europe. The whole plant biomass is lower on drier sites (Schwintzer & Lancelle 1983). In a comparison of M. gale in two habitats, Pmax (net photosynthetic rate at saturating irradiance), leaf N content and root nodule development were greater in the wetter habitat (Maeda et al. 1999). Sprent & Scott (1979), working on five Scottish stands of M. gale, have reported above-ground biomass of 301–580 g m–2, mean stem height of 46–88 cm, 21–24% of the aerial biomass as foliage and 76–79% of aerial biomass as stems. The following measurements were made using 10 leaves from five pot-grown plants. Results are presented, plus or minus the standard error of the mean. Fully expanded leaves have no stomata on their adaxial surface and 30.6 ± 9.8 stomata mm−2 on their abaxial surface. There were 8.3 ± 1.8 non-glandular hairs mm−2 on the adaxial surface and 3.7 ± 0.8 non-glandular hairs mm−2 on the abaxial surface. There were no multicellular stalked glandular trichomes on the adaxial surface and 10.2 ± 3.2 such trichomes mm−2 on the abaxial surface. There were 1.25 ± 0.2 sessile secretory glands mm−2 on the adaxial surface and 0.52 ± 0.13 mm−2 on the abaxial surface. Myrica gale produces cluster roots (Berliner & Torrey 1989). These structures consist of dense clusters of determinate rootlets, that develop on lateral roots opposite the protoxylem poles (Skene 1998). Cluster roots occur in many other species of Myrica and Comptonia. In the field in Maine, USA, they are lacking in non-nodulated plants (Berliner & Torrey 1989). One of us (L.H.) has seen extensive cluster root development on M. gale in Galloway, south-west Scotland. Myrica gale, Hippophae rhamnoides and Alnus glutinosa are the only species of actinorhizal plants native to the UK. These form nitrogen-fixing root nodules with the filamentous actinomycete bacterial genus Frankia (Benson & Silvester 1993). All are favoured by damp habitats, but M. gale is the only one well adapted to truly flooded conditions. Figure 3 shows a young nodule collected in the field in Scotland. Arising from each lobe is a negatively geotropic (gravitropic) root. Such roots form in response to low oxygen concentrations (Silvester et al. 1988) and, in the field, they grow, as necessary, until they reach the surface of the water. Length of these roots varies from less than 10 mm in soils that are free draining to over 30 mm in stagnant waters (Sprent & Scott 1979). Silvester et al. (1988) showed that nodule root length at 2 kPa O2 reached 62 mm, while at 40 kPa O2, nodule root length was 6 mm. In cross section they show an extensive system of air spaces, enabling oxygen to reach the nitrogen-fixing tissue. Nodules are formed following entry of Frankia filaments through root hairs (Callaham et al. 1979). Photograph of a nodule of Myrica gale from Rannoch Moor, Scotland. Nodule lobes (nl) and rootlets (ro) arising from them are clearly seen. Scale bar = 2 mm. Myrica gale has been reported as forming ectomycorrhizas (Rose 1980; Harley & Smith 1983). However, no evidence of ectomycorrhizal infection was found in M. gale by Berliner & Torrey (1989). Harley & Harley (1987) report that M. gale had arbuscular mycorrhizas in better drained sites, but did not show any infection in wet soil conditions. Rose (1980) reported arbuscular mycorrhizal infection in M. gale seedlings collected from natural habitats, although it is unclear if spores found in the rhizosphere were actually associated with the roots concerned. Berliner & Torrey (1989) found no evidence of arbuscular mycorrhizas in M. gale. Myrica gale commonly spreads by suckers to produce thickets. These suckers (often referred to as rhizomes) become very woody and act as a major nutrient store over winter. It can be propagated by seeds, stem cuttings, root division or suckers transplanted in early autumn or spring. Seedlings are rare in the field (Poore 1956). The somatic chromosome number is reported as 48 (Chr. Atl.). The base number for the family is 8 (MacDonald 1989); thus M. gale is at least hexaploid. Myrica gale is a subarctic, hygrophilous plant species, although in wet organic soils it displays adaptive drought avoidance mechanisms typical of a xerophyte, including mid-day depression in stomatal conductance. This is particularly noticeable when atmospheric water vapour pressure deficit is high (> 2 kPa) (Blanken & Rouse 1996). It is regarded as an avoider of water stress under the terminology of Levitt (1980). However, Maeda et al. (1999) report that M. gale showed high transpiration (E) and no midday depression of Pmax even under high irradiance and large vapour pressure deficit between leaves and ambient air on a midsummer day. These traits of photosynthesis and water relations were associated with the dominance of this shrub in wetter sites such as stream sides and hollows (Maeda et al. 1999). Very few data exist on transpiration rates and stomatal conductance in the field. Blanken & Rouse (1996) reported values between 0.03 and 1.2 mol m−2 s−1 for stomatal conductance (g) with a midday stomatal depression, while transpiration, at a soil to leaf water potential of 1 MPa, was 8 mmol m−2 s−1. For M. gale var. tomentosa, maximum stomatal conductance (gmax) varied from 0.41 to 1.37 mmol m−2 s−1, and maximum transpiration rate (Tmax) 10.7 mmol m−2 s−1 (Takagi et al. 1998). Few data are available on photosynthesis in M. gale. In an oligotrophic mire in central Japan, Pmax of M. gale var. tomentosa was 15.2–16.5 µmol(CO2) m−2 s−1, higher than those of neighbouring Betula platyphylla var. japonica and Rhododendron japonicum, throughout the growing season (Maeda et al. 1999). Pmax was positively correlated with leaf N content among the three species. Although M. gale is a relatively broad-leaved plant with high N content, the decomposition rate of summer litter in a riparian situation was found to be surprisingly low, below that of other deciduous species and similar to that of the conifer Tsuga canadensis. Phytochemicals in both species may be responsible for this. Chironomid larvae, however, were attracted to the litter of M. gale (Maloney & Lamberti 1995). Watershed liming (used as a strategy of mitigation for the effects of acid deposition on lakes) led to higher levels of Mo, and lower levels of Al, B, Mn and Ni, in leaves of Myrica gale, with higher levels of Al, Cr and Mo, and lower levels of Cd, Mn and Se, in twigs (Rossell et al. 1994). Other records of elemental composition of M. gale in an ecological context include those of Small (1972), Langille & MacLean (1976), and Mackun et al. (1993). Cluster roots create their own phosphate-rich patch within the soil by releasing large amounts of organic acids into the rhizosphere, leading to the mobilization of phosphate from sparingly soluble iron phosphates; this phosphate can then be absorbed. The extent of cluster root formation in M. gale is inversely related to the amount of soil phosphate available and the amount of organic matter in the soil (Crocker & Schwintzer 1994). When grown in water culture, cluster root formation has also been found to be influenced by nitrogen source and aeration (Crocker & Schwintzer 1993). Seedlings supplied with urea produced substantially more cluster roots than those supplied with nitrate, while aeration increased the speed of cluster root production and the number produced per plant (Crocker & Schwintzer 1993). Cluster roots may thus provide M. gale with an alternative mechanism for enhancing phosphate uptake, as mycorrhizas are absent from M. gale on wet soils in Great Britain (Harley & Harley 1987). Cluster roots may also be important in the invasion of new sites, as seedlings would have to form only one symbiosis (with Frankia) in order to establish (Crocker & Schwintzer 1993). One of the dilemmas of all nodulated plants, including legumes, is to balance the need for oxygen to support production of ATP necessary to provide energy to break the N≡N bond without inactivating the O2-sensitive nitrogenase enzyme complex (Sprent & Sprent 1990). The normally submerged nature of Myrica nodules exacerbates the oxygen supply problem, and it is one of the few actinorhizal genera that produce a form of haemoglobin which facilitates O2 diffusion to the Frankia vesicles, the site of nitrogen fixation. This is the normal situation in legume nodules, which are characterized by the pink colouration of their central infected regions. In Myrica, this pink colour is masked by tannins and other compounds. Nodules of M. gale contain similar levels of total haem (127 nmol g−1 f.w.) to that of a typical legume nodule (Lupinus albus: 197 nmol g−1 f.w.) (Tjepkema & Asa 1987). In all countries where it is indigenous, M. gale has been found to be nodulated, suggesting that compatible Frankia strains are widespread. For further details of these and other aspects of the symbiosis, see Schwintzer & Tjepkema (1990). Myrica gale is promiscuous in terms of its endophyte specificity (Torrey 1990), being nodulated by Frankia strains from Alnus, Casuarina, Ceanothus, Colletia, Comptonia, Cowania, Elaeagnus and Myrica. Although it is known from glasshouse and controlled environment studies that the fixation of nitrogen by actinorhizal plants can equal that of legumes, how much is fixed in the field is much less clear. Because actinorhizal plants are usually perennial and woody, nitrogen fixation is very difficult to quantify on a field scale (Sprent & Parsons 2000). The situation is complicated in M. gale by the fact that reproduction is largely vegetative—seedlings, as mentioned earlier, are rare in the field (Poore 1956). As with all perennial nitrogen-fixing plants, the proportion of nitrogen fixed, as a fraction of plant N, reduces with age because an increasing amount is recycled within the plant. This was shown for sites in Perthshire by Sprent et al. (1978). Figure 4 shows the seasonal nitrogen balance. Early growth of leaves was fuelled by nitrogen from the previous year's stems and suckers (which alone accounted for 70% of the total plant mass). Nitrogenase activity was not detected until leaves were quite well developed and new nodule tissue had started to be produced; it continued until late autumn. It was concluded that the main role of the current season's nitrogen fixation was to replace N lost with fallen leaves. Nitrogen is probably also needed to support new growth during vegetative reproduction. However, although some nodules can usually be found in any field site in the UK, they are seldom seen on new rhizomatous material (J. Sprent, unpublished). Seasonal studies by Schwintzer et al. (1982) indicate that nitrogenase activity first occurred upon development of symbiotic vesicles, and ceased when vesicles disappeared in the autumn. Cumulative distribution of N in Myrica gale in a Myrica–Molinia community on relatively poorly aerated soil at a site in central Scotland in 1976. Note the major contribution from the suckers (rhizomes). At the end of the season, the total N was similar to that at the beginning. Fixed N was estimated to be enough to replace that lost in fallen leaves. Arrows indicate when nitrogenase activity was first detected (mid-May) and when it ceased (late October). Figure redrawn from Sprent et al. (1978). Diurnal fluctuations have been noted in nitrogenase activity in glasshouse studies on Myrica, with a peak at midday (Wheeler 1969), but no record of natural fluctuations has been made. The seasonal pattern of nodule activity almost certainly varies with site. Factors likely to have major effects are temperature, soil pH and presence of combined nitrogen in the flooded soil water. Laboratory studies on temperate actinorhizal species suggest that it is necessary to grow plants at least at 15 °C in order to obtain satisfactory nodulation and N fixation (Schwintzer & Tjepkema 1990). However, in the study of Sprent et al. (1978) the soil temperature during the growing season varied between 8 °C and 12 °C. The site was exposed and at an elevation of 330 m. The only other field measurements appear to be those of Schwintzer et al. (1982), carried out in New England. There were only small differences in the pattern of growth and nodule activity between this study and the Scottish one. In particular, nodule activity began earlier and finished earlier in the US study, an observation that may be related to temperature. While Silvester et al. (1988) found that nodulated M. gale grows well in aeroponics with 5, 21 and 40 kPa O2, Schwintzer (1985) demonstrated that spring floods strongly restricted the growth of M. gale mainly because of nitrogen limitation. She reported that plants exposed to 67 days of flooding had substantially less leaf biomass, lower leaf and stem nitrogen concentrations and less total shoot nitrogen than unflooded plants (Schwintzer 1985). Myrica gale is a native of acid soils and thus it is not surprising that the optimum pH for nitrogen fixation was found to be 5.4 (Bond 1951), although activity at 4.2 and 6.3 was also high. Bond (1976) commented that soil water content might affect nodulation, with drier sites having reduced nodulation. Published data on effects of combined nitrogen are all based on glasshouse experiments and are difficult to interpret. Some suggest that low levels of N (1.4 µM) aid nodulation and others that this level may suppress nodulation (see papers cited by Schwintzer & Tjepkema 1990). These variations probably result from other, interacting factors and their relation with field conditions is unclear. Bark from twigs of male and female plants showed no significant difference in nitrogen concentration, tannins, phenolic glycosides and total phenolics, but were significantly different in terms of dry matter digestibility (Hjältén 1992). Leaves of M. gale yield 0.05–0.29% essential oils, while flower oil yield is 0.97% (on a fresh weight basis). The main components are shown in Table 1. Mathiesen et al. (1996) reported that the C-methylated dihydrochalcones, myrigalone (myr) A, B and C, extracted from the fruits of M. gale, uncoupled oxidative phosphorylation in rat liver mitochondria. Myrigalone A was most potent, displaying an effect twice that of 2,4-dinitrophenol. The uncoupling activity appears to be unrelated to antioxidative properties, as demonstrated for myrigalone A and B (Mathiesen et al. 1995). It has been suggested that myr B may help protect liver tissue from toxin-induced injury (Diep et al. 1991; Mathiesen & Osmundsen 1992) and may have a role in the prevention of atherosclerosis (Wang et al. 1995). Dihydrochalcones in M. gale are active as antimicrobials against both bacteria and fungi (Malterud & Faegri 1982, 1987). Chantrill et al. (1952) showed the suppression of Influenza A and a bacteriophage of Pseudomonas pyocyanea by an extract of M. gale. A new flavonol glycoside, kaempferol-3-(2,3-diacetoxy-4-p-coumaroyl) rhamnoside, was identified by Carlton et al. (1990). Non-volatile oils include triterpenoids, unusual chalcone derivatives and diarylheptanoids (Nagai et al. 1995; Morihara et al. 1997; Sakurai et al. 1997). Catkins open in April or May, nodule activity begins in late May or early June, and new shoots grow from buds on the previous year's stems and from new aerial shoots derived from the suckers. Leaves are larger on these sucker shoots, and exceed the area of