Ulex gallii Planch. and Ulex minor Roth

Authors

  • K. E. Stokes,

    Corresponding author
    1. Centre for Ecology, Evolution and Conservation, School of Environmental Sciences, University of East Anglia, Norwich, NR4 7TJ, UK,
    Search for more papers by this author
  • J. M. Bullock,

    1. Centre for Ecology, Evolution and Conservation, School of Environmental Sciences, University of East Anglia, Norwich, NR4 7TJ, UK,
    2. Centre for Ecology and Hydrology, CEH Dorset, Winfrith Technology Centre, Dorchester, Dorset, DT2 8ZD, UK, and
    Search for more papers by this author
  • A. R. Watkinson

    1. Centre for Ecology, Evolution and Conservation, School of Environmental Sciences, University of East Anglia, Norwich, NR4 7TJ, UK,
    2. Tyndall Centre for Climate Change Research, School of Environmental Sciences, University of East Anglia, Norwich, NR4 7TJ, UK
    Search for more papers by this author

  • *

    Abbreviated references are used for many standard works: see Journal of Ecology (1975), 63, 335–344. Nomenclature of plants follows Flora Europaea and Stace (1997) for British species.

K.E. Stokes, Queen's University Belfast, School of Biology & Biochemistry, Medical Biology Centre, 97 Lisburn Road, Belfast, BT9 7BL, Northern Ireland (e-mail k.stokes@qub.ac.uk).

Ulex gallii Planch. (Fabaceae). Western gorse. Densely spiny spreading shrub, ranging in height from 10 cm to 200 cm. Main branches usually ascending with abundant brown hairs, spines 1–2.5 cm, faintly furrowed or striate. Pedicels 3–5 mm, appressed hairy, bracteoles usually wider than the pedicels. Calyx slightly shorter than the corolla, appressed and hairy. Teeth of the lower calyx lip parallel to convergent, calyx 9–13(−15) mm, standard (12–)13–18(−22) mm. Wings strongly curved and usually longer than the keel. Bracteoles 0.5–0.8 mm long × 0.6–0.8 mm wide. Flowers 10–13 mm long, occurring in July–September, fruits (8–)9–13(−14) mm, dehiscing in spring. Seed dry mass 6.7 (SD ± 0.98) mg.

Ulex minor Roth (Fabaceae). Dwarf gorse. Ranges in height from 5 cm to 150 cm. Branches usually procumbent with abundant brown hairs. Spines varying 1–2 cm, faintly furrowed or striate. Pedicels 3–5 mm, appressed and hairy, bracteoles usually narrower than the pedicels. Calyx almost as long as the corolla; hairs on the calyx appressed and sparse. Teeth of the lower calyx lip divergent, calyx 5–9.5(−10) mm, standard 7–12(−13) mm. Wings straight and usually shorter than the keel. Bracteoles 0.6–0.8 mm long × 0.4–0.6 mm wide. Flowers 8–10 mm long, occurring in July to September, fruits 6–8.5 mm, dehiscing in spring. Seed dry mass 5.9 (SD ± 1.48) mg.

The two species show considerable overlap in many characters: including plant height, spine length, flower colour and pod size. The species are most easily distinguished by the lengths of the flower parts, which provide a reasonably sharp discontinuity (Proctor 1965). However both U. gallii and U. minor are also highly plastic and show considerable variation in distinguishing characters (Stace 1997; Kirchner & Bullock 1999).

Ulex gallii and U. minor are perennial dwarf shrubs characteristic of acid, nutrient-poor soils and are found most commonly in Britain on wet, dry and humid heaths. Both species can also occur in acid grasslands and in the drier parts of acid mire communities. In Europe as a whole they are found most commonly in heath vegetation but also occur in acid grasslands and dunes.

I. Geographical and altitudinal distribution

Three species in the genus Ulex are found in the British Isles. While U. europaeus is present throughout the British Isles, the smaller U. gallii and U. minor are confined to mid and southern regions. The geographical ranges of the two smaller Ulex species were recently described by Bullock et al. (2000). Both species are native to Western Europe and co-occur in Great Britain, France and Spain. Ulex gallii is also found in Ireland and U. minor in Portugal. Ulex minor is introduced in South America (Fl. Br. Isl. 1987), New Zealand (Webb et al. 1995) and the Azores (Tutin et al. 1968).

In Britain the two Ulex species show a parapatric range distribution in that the ranges are separate but slightly overlapping. Ulex minor is virtually confined to the south-east of England, whereas U. gallii occurs mostly in the west and north-west of England, Wales and the extreme south-west of Scotland. The distributions can be divided by a line running from Dorset on the English south coast, to the head of the Humber Estuary (Fig. 1a,b). However, there are a few localities where both U. minor and U. gallii are found within the range of the other species. Ulex gallii is found in several locations in south-east England, most notably in Kent in the extreme south-east and in large numbers on the East Anglian coast. Ulex minor has some records near the south-west Scottish border. In addition to these disjunct locations, there are several examples of co-occurrence around the Dorset–Humber line described as splitting the ranges. These are most notable in Dorset.

Figure 1.

The distribution of (a) Ulex gallii and (b) U. minor in the British Isles. Each dot represents at least one record in a 10-km square of the National Grid. Native: (○) pre-1950, (•) 1950 onwards; introduced: (×) pre-1950, (+) 1950 onwards. Mapped by Henry Arnold, Biological Records Centre, mainly from records made by members of the Botanical Society of the British Isles.

Heathland surveys that included information on the distribution of Ulex species in Dorset were carried out in 1978, 1987 (Webb & Haskins 1980; Webb 1990) and 1995 (Rose et al. 2000). All of the surveys showed clear separation of the distributions of the two Ulex species resulting in significant negative association between them (Bullock et al. 2000). Although none of the Ulex-dominated zones contains only one species, there are very few areas where the two species actually grow together on the same heath.

The sharp vicarism between the two species is also found in north-west France (Des Abbayes & Corillion 1949; Corillion 1950, 1959). Ulex minor is confined to the Atlantic climatic region of western France but is absent from the fertile, low-lying region of the mid-west coast, from western Brittany, from the coast around Cherbourg in Normandy and from the extreme south-west at the foot of the Pyrenees (Fig. 2). These last three gaps in the distribution are all occupied by U. gallii which shows an extreme westerly distribution. Around these three foci of the U. gallii range there are areas of overlap in the distributions of the two species roughly 20–60 km in width.

Figure 2.

The distribution of Ulex gallii and U. minor (presence/absence) on a 20 × 20 km grid in France (Dupont 1990).

The distributions of the Ulex species on the Iberian Peninsula are not mapped. However, Dupont (1962, 1990) states that U. minor is present in western Portugal, in Galicia and Asturia to the west and north of the Cantabrian range in north-west Spain. There are also a few records in the south-west in Andalucia and Extremadura. Ulex gallii is present in the extreme north-west of Spain in the north of Galicia and Asturia.

The altitudinal range of U. gallii in the British Isles extends from sea level to an upper limit of 1400 ft (425 m) on Exmoor in England, to 2000 ft (610 m) in Drygan in Wales and 2200 ft (670 m) on the Reeks in Ireland (Alt. range Br. Pl.). Ulex minor has an altitudinal range extending from sea level in England, Wales and Scotland to a maximum of 800 ft (240 m). The maximum recorded elevation for U. gallii is 5570 ft (1700 m) in northern Spain (Loidi et al. 1997).

II. Habitat

(a) climatic and topographical limitations

Ulex gallii has a westerly range and is associated with an oceanic climate typical of those areas bordering on the Atlantic (Rodwell 1991). Preston & Hill (1997) consider that U. gallii is an element of the Oceanic Temperate flora, while U. minor is a member of the Oceanic Southern-temperate flora. The replacement of U. gallii by U. minor appears to be influenced by a shift to a more continental type of climate in southern and eastern England. It has been postulated that the different ranges of the two species may be the result of climatic effects, with U. gallii being favoured by the wetter and warmer conditions in the west and U. minor by the drier, colder continental conditions in the east (Proctor 1965).

Climate envelopes have been constructed for each of the Ulex species (Wright, in Bullock et al. 2000) using data on the presence/absence of the species in relation to maximum July temperature, minimum January temperature and annual total precipitation. The climatic data were obtained from the Climatic Research Unit at the University of East Anglia and provide a 100 (10 × 10) km2 baseline climatology for Great Britain over the period 1961–90 (Hulme et al. 1995). At this scale, U. minor is found in areas with a warmer July maximum and also in areas that are marginally drier (Bullock et al. 2000).

Within Dorset, Bullock et al. (2000) have examined whether the distribution of U. gallii and U. minor relates to climate at a local scale, by testing for differences in rainfall between the different Ulex-dominated zones within Dorset using annual and monthly rainfall data for the period 1955–94, compiled from 12 weather stations distributed over the Dorset heaths. Variation in rainfall did not provide a clear explanation for zonation of the Ulex species at the smaller scale in Dorset.

Transplant experiments that aimed to set seed of each Ulex species within the range of its counterpart found that both U. gallii and U. minor can germinate and establish to seedling stage within the climatic zone outside of their current ranges (Stokes 2002). Longer term survival of transplanted U. gallii seedlings has been recorded from transplant experiments performed in 1961 at Ashdown Forest in East Sussex. Ulex gallii seedlings transplanted into plots close to Wych Cross (TQ 413 321) survived a period of 42 years to be present in 2003 (David Streeter, personal communication), suggesting no climatic barrier to growth of U. gallii from the seedling to adult stage beyond its current range.

The two species are generally found on flat ground and moderate slopes but rarely at steeper inclinations. Both species can be present close to sea level but U. minor is more commonly found at lower elevations than U. gallii, which can extend up to levels of approximately 500 m on the moorland fringes of Dartmoor, Bodmin Moor and Exmoor (Rodwell 1991).

(b) substratum

Communities containing these Ulex species develop over a range of base-poor and oligotrophic soils. Profiles tend to be highly acidic; superficial pH beneath these lowland heath communities is between 3.5 and 4.5 and the soils are generally very impoverished, features reflected in the calcifuge nature of the vegetation and the poor representation of mesophytic plants (Rodwell 1991).

Ulex minor is predominantly present on the most free draining of the series of acid and impoverished soils, typically those podzols which show no tendency, or only a slight one, towards surface gleying in winter. Such soils develop from pervious sandy or pebbly parent material, which typically give rise to some kind of podzolic profile, either a classic humo-ferric podzol, or a palaeo-argillic podzol where there is more material of the finer fractions in the profile. These soils can be distinctly droughty in the drier months (Jarvis et al. 1984). However, U. minor is also present further west, for example in the Central Wealden heaths, where a higher silt fraction results in a shift to seasonally waterlogged gley-podzols.

Ulex gallii is similarly capable of growing in a range of soil moisture conditions and is present growing in moist profiles where soil conditions are often maintained by some impedance to drainage in brown earths or podzolic profiles with an argillic B horizon or impervious iron-pan. However in more western parts of Britain, for example on the Devonshire Pebble-Bed Commons, U. gallii is also present on wetter stagnogleys and gleyed podzols developed over the gently dissected surface of the Triassic dip slope (Rodwell 1991). Some of these lowland gleyed soils show a shallow accumulation of mor humus beneath sub-shrub covers that have not been burned for some time. However with a shift to higher ground there is a strong tendency for profiles with impeded drainage to develop a humose top-soil. Around the fringes of Dartmoor, Bodmin Moor and Exmoor over ill-draining stretches of Devonian and Carboniferous shales and mudstones and granite, U. gallii is often found on stagnohumic gleys that form an intergrade between mineral stagnogleys and the thick ombrogenous peats mantling the highest and wettest ground. A consequence of the increased rainfall characteristic of the higher ground in north-west Somerset and the heartlands of Devon and Cornwall is that U. gallii can extend on to more free-draining soils that are kept moist as much by high precipitation as owing to any drainage impedance (Rodwell 1991).

Within southern England, Bullock et al. (2000) have examined whether the distribution of U. gallii and U. minor relates to edaphic factors at a local scale, by testing for differences between the different Ulex-dominated zones of Dorset heaths in organic matter content, extractable phosphorus, exchangeable phosphorus, phosphorus adsorption capacity and pH value. A statistically significant difference was found in that pH values were higher in U. gallii-dominated zones. Comparison of edaphic factors across a gradient in southern England running from the core of the U. gallii range in Devon, through the zone of range overlap in Dorset, to the core of the U. minor range in Surrey found a correlation between pH and the presence or absence of the Ulex species (Stokes 2002). Ulex minor did not occur at sites with a soil pH greater than 4.0 and U. gallii individuals were not observed within heaths where soil pH fell below 3.4 (Fig. 3). This trend in pH values from the south-west to the south-east may simply be a reflection of the geology sampled in different parts of southern England. Sampling a larger number of heath sites across this gradient with a greater diversity of underlying geology would reveal if a relationship exists between pH values and the abundance of either Ulex species.

Figure 3.

The range of soil pH values at sites of U. gallii and U. minor within Devon, Dorset and Surrey in southern England. The box represents the middle 50% of the data, lines extending from the box represent the upper and lower 25%. The mean is represented by the circular dot and the median by the horizontal bar. Outliers are represented by asterisks.

III. Communities

These gorses occur as sub-shrubs and dominants in a variety of plagioclimax communities occurring on oligotrophic soils within north-west Europe. Much of the lowland heath vegetation within Britain is relatively species poor and the different communities can be understood largely as permutations on a fairly limited range of dominants and associates representing biotically derived replacements of the more calcifuge of oak–birch and beech forests (Rodwell 1991). Nonetheless, within Britain the communities can fall into a series that can be related broadly to variations in regional climate, soils and treatments (Rodwell 1991).

In Britain, from east to west the first appearance of U. minor is within the H2 Calluna vulgarisUlex minor community (Rodwell 1991), where the bulk of the cover consists of Erica cinerea, Calluna vulgaris, Ulex minor and to a lesser extent Deschampsia flexuosa. Within the H2 community, U. minor shows a variety of different abundances; in some areas where grazing still occurs it seems to have been widely reduced (Rodwell 1991), whereas in other areas it can be prevalent to the extent of co-dominance with C. vulgaris. In the absence of regular burning, the ground layer of CallunaUlex minor heath can be very patchy and more or less limited to the cores of degenerate heather bushes, where the characteristic sequences of mosses and lichens can be seen, Dicranum scoparium and Hypnum jutlandicum being the most frequent bryophytes overall. In areas where the soil is prone to seasonal waterlogging and the climate is wetter, a more humid heath community prevails: the H3 Ulex minorAgrostis curtisii community, which is characteristic of the area around the Hampshire Basin. The major distinction between the H2 and H3 U. minor communities is created by the joint presence of Erica tetralix and E. cinerea in the latter, maintained by moister soils. In drier climates, E. cinerea is confined to free draining acid soils and E. tetralix to wet heaths on strongly impeded profiles (Rutter 1955; Bannister 1965, 1966; Gimingham 1972).

West of Poole Harbour, from Dorset down into the South-West Peninsula and in Wales, the H2 heath is replaced by the Calluna vulgaris–Ulex gallii heath (H8), within which C. vulgaris, E. cinerea and U. gallii are constants. Again the shift to the more oceanic conditions in the south-west results in the replacement of the H8 community by its more humid counterpart, the Ulex gallii–Agrostis curtisii community (H4), which is very similar to the Ulex minorAgrostis curtisii heath (H3) in its general floristics; the major difference being the replacement of one gorse by the other (Rodwell 1991). However, the boundary between the H8 and H4 U. gallii communities is more diffuse than that observed between the H2 and H3 U. minor communities within the more continental part of Britain. Owing to higher rainfall within the south-west, wet-heath plants can transgress ever more extensively onto the more free-draining acid soils, from which in drier climates they are excluded by susceptibility to drought (Rodwell 1991). Transitional types of heath vegetation thus become more interposed between drier and wetter communities in this part of Britain with Erica tetralix, Molinia caerulea and Agrostis curtisii playing an important role alongside Calluna, E. cinerea and one or the other of U. gallii and U. minor.

In June 2000, plant community data were recorded at 10 heath sites along an environmental gradient in southern England, running from the core of the U. gallii range in Devon, through the zone of range overlap in Dorset, to the core of the U. minor range in Surrey (Table 1). Of the heathlands utilized, two sites from the west of England contained U. gallii (Aylesbeare Common and Coalton Radleigh Common) and two sites from the east in Surrey contained U. minor (Thursley Common A and Thursley Common B). Within Dorset, in the region of range overlap, two sites contained only U. gallii (Canford Heath and Upton Heath), two sites contained only U. minor (Arne Heath and Hartland Moor) and two sites contained both species (Gore Heath and Stoborough Heath).

Table 1.  Floristic lists from 10 selected sites in southern England at which Ulex minor or Ulex gallii occurs; lists are from thirty-six 2 × 2 m quadrats at each site. Details of soils at the listed sites are given in Fig. 3. Frequency and abundance values follow standard NVC procedures (Rodwell 1991)
SpeciesSite
12345678910
  1. Site 1: Aylesbeare Common, Devon, SY 055 905; Site 2: Coalton Radleigh Common, Devon, SY 055 880; Site 3: Canford Heath, Dorset, SZ 030 955;

  2. Site 4: Upton Heath, Dorset, SY 985 945; Site 5: Gore Heath, Dorset, SY 927 903; Site 6: Stoborough Heath, Dorset, SY 925 853;

  3. Site 7: Hartland Moor, Dorset, SY 955 856; Site 8: Arne Heath, Dorset, SY 973 882; Site 9: Thursley Common A, Surrey, SU 908 401;

  4. Site 10: Thursley Common B, Surrey, SU 900 407.

Calluna vulgaris V (4–10)IV (4–9)   V (4–10)  I (4–6)   V (4–10)   V (5–10)    V (5–10) V (4–10)V (4–10)   V (4–10)
Ulex gallii V (5–9)   V (4–9)   V (4–10)   V (4–10)   V (4–10)IV (4–10)    
Ulex minor    IV (4–7)IV (4–10) V (4–8) V (4–7)V (4–10)   V (4–9)
Erica cinereaIII (4–7)IV (4–8)III (4–8)IV (4–9)   V (4–10)   V (4–8)III (4–7) V (4–9)V (4–10)IV (4–10)
Molinia caeruleaIV (4–10)III (4–10)  II (4–10)   V (4–10)     I (1–4)   V (4–10)   I (1–4) II (4–7)  I (4–5)  I (4–5)
Erica tetralixIV (4–9)  I (4–6)  II (4–6)  I (7)  I (4)IV (4–8)   II (4)III (4–5)  
Festuca filiformis    I (4–10)III (4–10)   I (2–6)III (4–10)   II (2–7)  I (7–8) III (5–10)  
Pteridium aquilinum    I (4–5)  I (4)IV (1–7)  I (5–7)  I (4)     I (4)  
Ulex europaeus    I (10)  I (10)  I (4)    I (5–10)   I (4–8) II (4–7)   II (4–7)
Betula pendula   I (4)  I (4)  I (4)     I (1–4)   I (4)  I (4)
Hypnum jutlandicum    I (4)   I (4)    I (4)   I (4–5)  I (4–5)
Polygala serpyllifolia    I (2–4)  I (2–4)   I (2–4)      I (1–4)  
Pinus sylvestris    I (1–4)    I (1–4)    
Cladonia coccifera      I (2–4)    I (2–4)   
Potentilla erecta   I (1–4)        

Ordination of the plant community data was carried out using detrended correspondence analysis within the program canoco (ter Braak 1988). As the primary objective of the analysis was to ascertain whether the vegetation at the sites differed in any other way apart from the identity of the Ulex species, the ordination analysis was repeated with the two Ulex species combined into a single taxon. When U. gallii and U. minor were classified separately in the analysis a distinctive separation in community composition was observed between Devon and Surrey (Stokes 2002; Fig. 4). This separation is maintained even if U. gallii and U. minor are treated as a single taxon in the analysis (Stokes 2002). However the plant communities within Dorset were more variable than those in either Devon or Surrey and encompass the variation in both. It is suggested that this greater variation relates to a wider range of soil pH values within the Dorset area (Fig. 3).

Figure 4.

Ordination of plant community data from 10 sites across southern England using detrended correspondence analysis (DCA). A plot of the axis 1 and 2 scores is shown from a site ordination of floristic data from Dorset (○), Surrey (◆) and Devon (▪). Ulex gallii and U. minor were classified as separate species in the analysis.

IV. Response to biotic factors

(a) competitive ability

The Ulex species are fast-growing woody legumes which are capable of establishing rapidly from seeds or growing from stumps, particularly in low fertility soils (Gaynor & MacCarter 1981). Both species are relatively shade tolerant (Rodwell 1991) and therefore able to establish themselves and maintain an existence as part of the sub-shrub canopy.

Competition is postulated as a cause of the parapatric distribution of these two Ulex species (Proctor 1965; Bullock et al. 2000). Competition experiments were designed to identify the role that intraspecific and interspecific competition play in the dynamics of U. gallii and U. minor along an environmental gradient within southern England that includes the range boundary of the two species and pure populations of the two species at either end (Stokes 2002). Seed of each of the Ulex species, collected from sites within Dorset, was grown under natural light in a glasshouse for 4.5 months prior to being transplanted into replicate plots at heathland sites within Devon, Dorset and Surrey. Each plot was divided into subplots and seedlings were then planted within each subplot, of which one was planted with U. minor, one with U. gallii and one with a mixture of both. The seedlings were planted in a hexagonal arrangement in close enough proximity to compete. Within the mixed subplots individuals were planted so that each individual had two interspecific neighbours (Harper 1977). After a 12-month period each seedling was harvested with as much of the root system as was possible, the length of the roots and shoots measured and the seedlings dried to constant mass and weighed.

Data were segregated by species and analysed using an anova design to examine the effect of range location and planting combination (monoculture vs. mixture) on individual yield and root and shoot length. The evidence for interspecific competition between the two Ulex species is weak. Comparison of mean plant biomass for U. gallii individuals within Devon, grown in pure and mixed planting combinations, found U. gallii to suffer a reduction in biomass owing to the presence of U. minor (F = 10.26, d.f. = 1,35, P < 0.01). A replacement series analysis provided some evidence for competition between the two species, in that U. minor appeared slightly the better competitor within Devon, although this is outside its current range. However the analyses consistently identified the influence of range location upon the different measures of plant performance as significant, rather than the presence or absence of a competitor and there was no evidence that the competitive balance shifted across the range boundary in a consistent manner (Stokes 2002).

A neighbourhood analysis of competition was also conducted within ten 2 × 2 m quadrats at Gore Heath in Dorset where both Ulex species grow in proximity together (Stokes 2002). Within each quadrat the total number of Ulex individuals were counted and the following measurements recorded for each individual Ulex plant: stem diameter, number of branches, pod number per sample branch and distance to the nearest conspecific and heterospecific Ulex neighbour. The study revealed that the fecundity of U. gallii individuals increased with increasing distance from U. minor individuals (y = 17.31 + 23.23x, F1,25 = 5.143, P < 0.05), whereas there was no relationship with conspecifics. For U. minor there were no correlations between fecundity and distance to either conspecifics or heterospecifics. However the fecundity of U. minor individuals increased with increasing density of conspecifics (y = 0.963 + 0.099x, F1,37 = 7.34, P < 0.05). Thus there is some indication from these results that U. gallii suffers interspecific competition to a greater extent than U. minor.

(b) grazing, browsing and trampling

Grazing has undoubtedly contributed to the maintenance of the H2 community in the past by curtailing the invasion of trees. Webb (1986) has suggested that in the lowlands this factor has been of greater long-term importance than burning in preserving the cover of heath vegetation. Grazing management declined towards the end of the 20th century, apart from in the New Forest where the various heath communities provide the bulk of the enclosed land that is still exploited by the traditional mixture of cattle and ‘heath-cropper’ ponies, together with some deer.

Gorse can be grazed only when young and tender; livestock such as deer, cattle and ponies are probably capable of grazing young shoots. It has been postulated that grazing affects the proportions of different sub-shrubs and may contribute towards the scarcity of U. minor in some stands as the soft young shoots are very palatable (Rodwell 1991). Rabbits may graze upon young seedlings and the smaller U. minor shrubs but are restricted by height from grazing upon taller parts of adult U. gallii plants. The contribution of the palatable grasses to the post-burn succession in Calluna vulgarisUlex minor heath (H2) is generally less prominent than in the H4 Ulex minorAgrostis curtisii heath (Rodwell 1991).

Within a typical U. gallii community the subshrub canopy is of high cover, often with only a very sparse herbaceous component. However, less commonly the bushes are separated by systems of grassy runnels where the vegetation has been open to grazing and the structural contrast between these two components can be sharply accentuated (Rodwell 1991). When the bushes themselves are nibbled the canopy can be reduced in height, resulting in dense ‘hedgehogs’ of gorse with short but untidy bushes of heather scattered amongst them. Grazing can mediate every graduation between the extremes of dense heath on the one hand and continuous grassy sward on the other, and is a major factor controlling zonation (Rodwell 1991).

Disturbance of the soils around settlements and plantations, or along tracks, can lead to a spread of the UlexRubus scrub within the CallunaUlex minor heath (Rodwell 1991).

(c) fire

Within the last 50 years,grazing has been in decline as a heathland management tool and often the only factor preventing succession to woodland is fire (Webb & Haskins 1980). Controlled burning is conducted according to strict guidelines governing the season and intensity of burning and the techniques used (Gimingham 1992). Fires are generally used on drier heaths, dominated by Calluna vulgaris, where the aim is to burn off ageing growth of the dwarf shrubs to allow their regeneration, to kill invading scrub and, on the larger scale, to create a mosaic of heathland of different ages (Khoon & Gimingham 1984; Gimingham 1992).

After a burning event both U. gallii and U. minor can produce new shoots from the burnt stumps (Gloaguen 1993), although the probability of doing so is greater in U. gallii than U. minor (Stokes 2002). Fire is stimulatory to germination in both species. Hossaert-Palauqui (1980) studied heathlands in Brittany in northern France and found that elevated temperatures and higher proportions of bare ground after fires resulted in an increase in germination for U. minor. Allchin (1998) found that burning conducted in the absence of a paraffin fuel stimulated germination of U. gallii seeds within the humic layer at Aylesbeare Common in Devon.

Heathland shrubs can be aged by size, enabling estimation of areas of vegetation with a chronological sequence of time since disturbance. Demographic data collected within Dorset on Ulex individuals within patches of differing time since fire were used to construct patch-specific matrix population models for each Ulex species (Stokes 2002). Matrix multiplication methods were utilized to simulate the response of differing fire return intervals upon population growth rate of either species. Both Ulex species are predicted to decline under annual burning. However as the time between burns increases, the population growth rate increases monotonically for each species. The minimum fire return intervals permitting persistence are 3 years for U. gallii and 4 years for U. minor. Maximum population growth rates are achieved from a 16-year fire return interval for both species.

V. Response to environment

(a) gregariousness

Dispersal is limited within these species resulting in aggregated distribution of individuals. Analysis of the spatial distribution of individuals was conducted within ten 2 × 2 m quadrats at Gore Heath in Dorset, where both U. gallii and U. minor grow in proximity together (Stokes 2002). The distribution of individuals in the field indicated that U. minor plants were found closer to conspecifics than heterospecific U. gallii neighbours, whereas there was no significant difference for U. gallii (see IV (A)).

(b) performance in various habitats

Seedlings transplanted across an environmental gradient within southern England showed significant differences between performance as measured by root and shoot length at the different sites (see IV (A)). Seedlings of both species performed least well in Surrey where annual precipitation is lowest (Stokes 2002; Fig. 5).

Figure 5.

Mean shoot length and root length for (a) Ulex gallii seedlings and (b) U. minor seedlings along an environmental gradient within southern England. The error bars are ± 1 SE.

Both of the Ulex species reproduce freely throughout their respective ranges. Interestingly, analysis of seed production per plant measured at the core and margin of each species range did not indicate a decline in fecundity towards the range margin for either of the Ulex species (Stokes 2002; Fig. 6). Therefore, U. gallii and U. minor populations at the western and eastern limits of their respective ranges are not restricted owing to failure to set seed.

Figure 6.

Estimates of total seed production per m2 along an environmental gradient within southern England, spanning the range distributions of Ulex gallii and U. minor. The mixed site within Dorset refers to a heathland site containing both U. minor and U. gallii within the plant community. All other sites contain only one of the Ulex species. The error bars are ± 1 SE.

Within the zone of range overlap in Dorset, both U. gallii and U. minor showed significant allometric relationships between size (measured by stem diameter) and seed production. Within Dorset, seed production per plant was higher for both U. gallii and U. minor at sites containing both Ulex species (mixed heaths) as opposed to sites containing only one of the pair. Consequently estimation of total seed production per unit area was greatest within mixed Ulex sites in Dorset than elsewhere in the range (Fig. 6). Interestingly, the allometric relationships indicated that whilst larger U. gallii plants produce more seed in mixed Dorset heaths, in the case of U. minor it is the smaller plants that have greater seed production in mixed heaths than in heaths containing only one of the Ulex species. No relationship was observed between size and seed production per plant at the core of the U. gallii range in Devon, or at the core of the U. minor range in Surrey (see VIII (C)).

(c) effects of frost and drought, etc.

There is little evidence of mortality of adult plants due to frost. Within Britain at the core of the U. minor range in Surrey the mean January temperature is 0.6 °C but the mean number of days of grass frost during January is 19.4; therefore there are frequent frosts. Within Britain, U. gallii extends further northwards than U. minor, reaching the southern regions of Scotland, indicating a potentially higher ability to survive more prolonged periods of frost than U. minor. Experiments conducted by Millener (1952), subjecting 6-week-old U. gallii and U. minor seedlings to a prolonged period of low temperature (−10 °C for 16 h), resulted in 7% mortality of U. gallii seedlings and 29% mortality of U. minor seedlings. Severe winter frost can damage shoots of established U. gallii in Devon (M.C.F. Proctor, personal communication).

It is probable that lower temperatures stimulate germination, as there is a distinct peak in germination during October and November. However seedlings of both species appear to be highly susceptible to prolonged periods of drought, which is a frequent cause of mortality during juvenile development.

Ulex gallii has been shown to be sensitive to water stress observed through an increase in meiotic abnormalities (Misset 1992). Ulex gallii individuals occurring in coastal heathlands of Brittany, France, frequently exhibited irregularities during meiosis that led to partial or total male sterility. Microsporogenesis in individuals grown in an experimental garden located far from the coast appear to be normal and their fertility is also higher, as shown by counts made on seeds and pods. Comparison of plants growing in both places suggests that the low level of fertility may result from environmental factors such as drought or salinity (Misset 1992).

VI. Structure and physiology

(a) morphology

The framework of the adult plant is built essentially from accessory branches which typically begin to develop in the second season from buds lying between the primary branches and the leaves. Accessory branches bear subulate leaves and primary branches (usually spines), between these are further accessory branches which generally burst forth the following spring; thus the system of scaffold branches is built up (Fig. 7). Ulex minor is characterized by the early development of abundant primary branches – the vigour of the buds explains the frequent development of U. minor mats in the field (Millener 1952). In U. gallii, accessory branches are usually stout and reach a maximum of approximately 30 cm long, whereas in U. minor accessory branches can be slender, although again rarely over 30 cm long. Primary spines of U. gallii can be up to 4 cm long, fairly stout and re-curved, whereas in U. minor primary spines rarely exceed 2 cm long and are slender, needle-like and may be curved or straight (Millener 1952). Secondary spines of U. gallii rarely extend beyond the basal third of the primary spine. Secondary spines of U. minor are weak and sometimes extend to the middle of the primary spines. Tertiary spines, if developed, are feeble in both species (Millener 1952).

Figure 7.

Morphological structure of the shoot system of Ulex gallii and U. minor.

Information on the rooting system is sparse but both U. gallii and U. minor have long-lived deep tap roots and shallow adventitious roots, formed from the swollen bases of the stems (Millener 1952). The rooting system of U. gallii is described by Heath & Luckwill (1938): both tap and lateral roots are very stout and woody, 2.5–4.0 cm in diameter at the top and root hairs are present on finer rootlets. Laterals run for a metre or more a few centimetres below the surface, with a maximum lateral spread recorded at 183 cm and maximum depth of the tap root as 76 cm (Heath & Luckwill 1938). Young plants have a well developed tap and short, unbranched, horizontal laterals (Fig. 8). Later laterals elongate rapidly and become finely branched at their extremities. The main root system is supplemented by a fine mat of adventitious roots from the swollen bases of aerial stems (Heath & Luckwill 1938). In U. minor, finer, fibrous surface roots are often more prominent than deeper tap roots (Millener 1952). Ulex minor plants appear to be shallower rooted than U. gallii plants of an equivalent stem diameter; U. minor plants of 2 cm stem diameter possess tap roots extending approximately 40 cm beneath the soil in Dorset heaths, whereas the tap roots of equivalently sized U. gallii plants at the same location extend to approximately 1 m.

Figure 8.

Morphological structure of the root system: (a) 6-week-old seedlings of Ulex gallii and (b) 6-week-old seedlings of U. minor grown in John Innes Compost no. 2 July–August. Nodules are present. Each square equals 3 × 3 cm. (c) Rooting system of an adult U. gallii plant reproduced from Heath & Luckwill (1938). The system was exposed by digging a trench alongside the plant. The lower limit of the subpeat is shown by a dotted line, beneath which is a heavy loam containing numerous large stones. The inset shows the rooting systems of young plants before elongation of the laterals has set in. Each square equals 30 × 30 cm.

The Ulex species are leguminous and harbour symbiotic nitrogen-fixing bacteria (Rhizobium) in nodules on the roots. The Rhizobia of Ulex have been assigned to the slow-growing members of the cowpea miscellany, i.e. Bradyrhizobium (Allen & Allen 1981). Root nodules are apparent on plants after one year of growth. In the second and later years of growth older nodules are sloughed off the root system until in mature plants nodules are to be found only on minor fibrous roots. In water-logged or poorly aerated soils nodules are restricted to roots exploring the upper soil horizons.

Studies by Pate (1958) of U. gallii populations near Belfast in Northern Ireland report nodules at least 20 months old noted on several plants in their third year of growth. The life cycle of these biennial nodules was further studied by tagging individual nodules in pot-grown plants. It was found that nodules which developed on young roots in spring or early summer remained active in fixation until the following autumn when the lower halves turned green (Pate 1958). In those few nodules destined to resume growth, the nodule meristem was preserved and allowed a second season's increment of tissue to be made to the existing tissue. At the end of its second season the nodule had become an elongate structure, the lower half brown and shrivelled and the upper half composed of healthy tissue. There was no evidence that any nodules persisted into the spring of a third season. Successful over-wintering of nodules is observed to be a common event in Northern Ireland, suggesting that the life span of the nodule is not shaped by climatic influences. The anatomical and physiological constitution of nodules will allow for perennation if suitable climatic conditions persist throughout the life span (Pate 1958).

Allometric relationships, of the form Y =Axb, indicate differences in growth and development patterns between the two Ulex species (Stokes 2002). Data obtained in southern England for U. gallii plants within Devon and Dorset and U. minor plants within Surrey and Dorset indicated that for a given stem diameter a U. minor individual will produce a wider canopy than U. gallii. Ulex minor plants also develop longer branch lengths in relation to stem diameter of a branch than U. gallii plants. Comparison of the height of the two species within heathland environments indicates that U. galli is the taller plant, with a mean and standard error of 44.4 cm ± 0.85, whereas U. minor has a mean height of 32.1 cm ± 0.59. These relationships imply that U. minor is a low-lying straggler, exploiting the horizontal dimension, whilst U. gallii is a taller shrub, exploiting the vertical dimension to a greater degree. However both species are highly plastic with respect to height. Within different environments, U. gallii can form a shrub from 1.5 m to 2 m tall, as it does, for example, alongside the Hartside Pass road across the North Pennines where it descends into the Eden Valley. Ulex minor can attain heights of 70 cm on the fringes of plantation forestry within Surrey and Dorset.

The surface of the leaf blade is highly corrugated. Foliage surface features such as wax and trichomes vary greatly with age. Stomata (see also VIII (E)) occur along the insides of the corrugations and within the troughs of the leaf blade. Stomata are also present on the photosynthetic stems. Mean stomatal diameter (± SE) in U. minor is recorded as 17.47 µm ± 0.042, whereas for U. gallii (tetraploid) mean stomatal diameter was recorded as 21.69 µm ± 0.063, and for hexaploid U. gallii mean stomatal diameter was recorded as 23.95 µm ± 0.052 (Misset & Gourret 1996).

(b) mycorrhiza

Fine root samples of U. gallii were collected from a heathland site in the Peak District near Sheffield. Ulex minor samples were collected from Gore Heath in Dorset. Twenty separate subsamples of each species were examined for colonization by arbuscular mycorrhizal (AM) fungi. All that was found in the way of typical AM colonization were some vegetative hyphae and vesicles in 10% of the root length of each of two of the 20 subsamples of U. gallii. No AM colonization was observed in U. minor root samples (David Read, personal communication). Heath & Luckwill (1938) also report an absence of AM colonization in U. gallii growing in heathland communities. However this does not negate the possibility of mycotrophy in these species. Heathland environments are unlikely to support a vigorous AM inoculum. If the Ulex species were growing in species-diverse grassland one might expect a higher degree of AM colonization, especially if the soil pH was somewhat higher than that usually found in heathlands. These communities are typically prone to greater levels of colonization, and so would provide a higher inoculum potential. Whilst sampling a more diverse array of plant communities could reveal a greater extent of AM colonization in this genus, the Ulex species can clearly perform adequately in the absence of colonization.

(c) perennation: reproduction

Phanerophyte. Polycarpic. The normal method of propagation is by seed. A critical size threshold for flowering exists for both Ulex species. Ulex gallii plants with a stem diameter of 2 mm produce flowers, whereas U. minor plants attaining a stem diameter of 1 mm can produce flowers (Stokes 2002). All individuals with stem diameters larger than 4.5 mm and 3.5 mm flowered in the case of U. gallii and U. minor, respectively. Given the low threshold sizes of U. gallii and U. minor, it might be expected that both species would flower within 2 years. No diminution of reproductive output was observed with size. Ulex minor individuals measured from populations in Surrey and Dorset produce a greater number of flowers per plant than U. gallii individuals of equivalent stem diameter in Devon and Dorset (Stokes 2002).

(d) chromosomes

Chromosome numbers provide an unambiguous way of separating the two species. Ulex minor is diploid (2n = 32) (Alvarez Martinez et al. 1988; Castroviejo & Valdez-Bermejo 1990; Misset 1990; Fernandez Prieto et al. 1993; Misset & Gourret 1996; Bullock et al. 1998) and U. gallii has been shown to be either tetraploid (2n = 64) (Fernandez Prieto et al. 1993; Bullock et al. 1998) or hexaploid (2n = 96) (Misset 1990; Misset & Gourret 1996). Although several studies have reported both tetraploid and hexaploid U. gallii (Alvarez Martinez et al. 1988; Castroviejo & Valdez-Bermejo 1990; Fernandez Prieto et al. 1993), it is sufficient for taxonomic separation that the two species have consistently different ploidy levels (Kirchner & Bullock 1999).

(e) physiological data

Experiments conducted by O’Toole et al. (1991) in the glasshouse examined the effects of soluble sodium phosphate, insoluble ground rock phosphate and nitrogen on biomass production, nodulation and acetylene reduction by U. gallii seedlings transplanted into an acid, N- and P-deficient Old Red Sandstone forest soil. The equivalent of 450 g of air-dried soil was weighed into pots of 10 cm diameter. Ulex gallii seedlings of 1–2 years of age collected in County Wicklow were transplanted at one seedling per pot. Both P and N sources were supplied in differing combinations and at two different concentrations.

The growth of U. gallii seedlings in the impoverished soil was poor in the absence of applied P. Both P sources significantly and equally increased shoot biomass production and dry matter yields, whereas the application of N had no significant effect on shoot biomass production by either the sodium- or rock-phosphate fertilized plants. Root growth was preferentially increased by fertilizer P compared to shoot production and root : shoot ratios increased with increasing rates of P application, especially with sodium phosphate and in the absence of fertilizer N. The application of N tended to reduce the root : shoot ratios, especially in plants fertilized with sodium phosphate (O’Toole et al. 1991). Added N promoted only small, non-significant increases in shoot N uptake, suggesting that U. gallii growth occurred independently of the supplemental N (O’Toole et al. 1991).

The application of phosphate was essential for successful nodulation and acetylene reduction activity of U. gallii in this Old Red Sandstone soil. Overall, both the sodium phosphate and the rock phosphate were equally effective P sources. Maximum nodulation occurred at 6.75 mg P pot−1 with sodium phosphate but at 13.5 mg P pot−1 with rock phosphate. Increased acetylene reduction activity was observed at higher concentrations of P. Maximum acetylene reduction by intact roots was measured at 4.09 and 4.69 µmol C2H4 g−1 fresh weight nodule h−1 with sodium and rock phosphate, respectively. Applied N generally suppressed nodulation and severely restricted acetylene reduction activity. However the effects of N fertilizer on acetylene reduction activity were rate-dependent and apparently less inhibitory than in the case of nodulation (O’Toole et al. 1991).

Maximum acetylene reduction activity in the U. gallii seedlings was converted to estimates of field N2-fixation rates based on the following assumptions: (1) a C2H2 : N2 ratio of 3 : 1; (2) a field plant density based on the 10-cm diameter pot used, and (3) maintenance of ‘active’ N2-fixation for 8 h day−1 over 200 days (O’Toole et al. 1991). The maximum potential U. gallii N2-fixation rate based on the above assumptions is 21 kg N ha−1 year−1, which is very similar to rates derived for field measured acetylene reduction activity in U. europaeus in Cornwall (Skeffington & Bradshaw 1980) and in Cytisus scoparius in Oregon (Helgerson et al. 1979).

Chemical analyses of U. gallii fresh photosynthetic shoots obtained at Aylesbeare Common in Devon are described by Hayati & Proctor (1990), expressed as percentage dry weight ± SD: Ca = 0.34 ± 0.07, Mg = 0.20 ± 0.08, K = 1.54 ± 0.18, Na = 0.26 ± 0.11, Fe = 0.02 ± 0.01, Mn = 0.018 ± 0.007, N = 2.47 ± 0.46 and P = 0.169 ± 0.081. The uptake of Mg and K by U. gallii is strongly negatively correlated with soil Ca, in spite of the strong positive association of these ions in the peats. A paradoxical consequence is that the tissue concentration of Mg in U. gallii is negatively correlated with the amount of extractable Mg in the soil (Hayati & Proctor 1990). Seedling mineral nutrient contents of leaf weight for U. gallii seedlings were recorded as 38.2 mg g−1 for N, 2.86 mg g−1 for P and 41.8 mg g−1 for K. Leaf Narea in g m−2 was recorded as 1.84 and leaf nitrogen weight ratio as 17.9 mg g−1 (Cornelissen et al. 1997).

Measurements by Cornelissen et al. (1996) of seedling attributes at final harvest after 3 weeks of growth found U. gallii to have a mean RGR of 0.077 ss (d−1) ± 0.0028, a leaf area ratio (LAR) of 9.73 mm2 mg−1± 0.27, a leaf weight fraction (LWF) of 0.550 ± 0.004, a specific leaf area (SLA) of 20.74 mm2 mg−1 ± 0.50 and a specific saturated leaf area (SSLA) of 2.6 mm2 mg−1± 0.04.

Stem traits that are indicators of hydraulic capacity (conduit diameter), or mechanical strength (stem density, SD) were investigated in U. gallii seedlings (Castro-Diez et al. 1998). Stem cross-sections were examined under a microscope and average diameter of the 10 widest conduits of the section (Dmax) was measured as a correlate of the theoretical hydraulic conductivity (t − Kh). For U. gallii seedlings Dmax= 16.9 µm and SD = 0.17 mg mm−3.

(f) biochemical data

HPAEC-PAD analyses of different plant organs revealed that soluble sugar content of U. minor flowers is greater than that of U. minor shoots and stems (Picaud et al. 2002). Concentrations of soluble carbohydrates differ throughout the growing season: at the end of July concentrations in mg g−1 (± SE) for dry weight of U. minor flowers were 58.3 ± 0.9 for glucose, 67 ± 6.7 for fructose and 3.5 ± 0.2 for sucrose, whereas at the end of September these concentrations had changed to 69.1 ± 0.5 for glucose, 77.4 ± 0.9 for fructose and 18.9 ± 0.7 for sucrose. Concentrations in the stems were lower: 5.6 ± 2.1 for glucose, 2.6 ± 0.4 for fructose and 8.0 ± 0.8 for sucrose and those in the shoots were lower still: 3.95 ± 0.8 for glucose, 3.0 ± 1.4 for fructose and 6.6 ± 0.8 for sucrose. In comparison, U. europaeus flowers lack sucrose and contain half the fructose and two-thirds of the glucose of U. minor flowers (Picaud et al. 2002).

A variety of flavonoids are present in the tribe Genisteae, for example chalcones occur in the Ulex species, mainly as yellow pigments in flowers. Isoliquiritigenin is a chalcone occurring in the flowers of U. gallii and U. europeaus (Harborne 1971). A survey of leaf flavonoids of 128 species of 23 genera in the Genisteae found the isoflavones, daidzein, genistein and 5-methylgenistein to be present in both U. gallii and U. minor (Harborne 1971). Terpenoids, such as carotenoids, identified in petal pigments of U. gallii are α-carotene, β-carotene, flavoxanthin and taraxanthin; in contrast U. europaeus lacks flavoxanthin (Harborne 1971).

Fisher et al. (1986a) isolated fungal endophytes from U. gallii and U. europaeus spines and stems to examine the hypothesis that endophytic fungi may be able to produce antibiotics in order to compete successfully with antagonists. Antibiotic activity was detected in 4 out of 25 isolates of endophytic fungi selected from a collection of 330 endophytic isolates (160 from U. europaeus and 170 from U. gallii) which were grown in shake culture in the laboratory.

Ulex gallii seeds contain phytohaemagglutinins (PHAs), specifically anti-H lectins capable of agglutinating specific human erythrocytes, a useful tool in blood subgroup diagnosis (Toms & Western 1971).

VII. Phenology

The two species are extremely similar in phenology. Germination takes place throughout the year with peaks in early autumn and spring. Slow but continued growth occurs throughout the winter. Growth accelerates in early spring when new shoots are seen on the plants. Inflorescences develop in July and plants continue to form flowers into September or even later. The two species show some differences in flowering phenology. Kirchner & Bullock (1999) examined flowering phenology on Dorset heaths in southern England. A sample of 135 plants was censused every 11 days and comparisons made in terms of the changes in the mean number of flowers per plant and the proportion of plants in each census which reached their peak flower number at that census. The first U. gallii flowers were seen on 13th July but U. minor started flowering later on 23rd July. While the U. gallii population reached maxima in both the mean flower number and the proportion of plants at peak flower production on the 18th August, these maxima were attained by the U. minor population on 9th September. Chi-squared tests showed that the relative distribution of flower numbers between the censuses differed significantly between the species (χ2 = 557, d.f. = 10, P < 0.001) and the average date of the peak in flower number per bush was later for U. minor (median = 28th August) than U. gallii (median = 18th August).

In Dorset (southern England) both species continue flowering through to mid-October, whereas in Brittany (northern France) U. minor is recorded as flowering from August to November and U. gallii from August to January (Anne Atlan, personal communication). Fertilized flowers soon begin to develop a swelling green pod which develops a purplish-black colour as the pods ripen in May. Seed dispersal occurs by explosive dehiscence throughout May and early June.

VIII. Floral and seed characters

(a) floral biology

Flowers are bilaterally symmetrical (zygomorphic). Flowers are yellow and occur singly or in clusters at the base of the phyllode axils, at the union of spines and shoots. As in other legume flowers, there are five petals: a large standard turning upwards, two lateral petals (the wings) and two lower petals joined at their lower margins into a keel. The calyx is also yellowish, tube-like but divided to the base into two lips; the upper lip has two teeth and the lower lip three teeth. The 10 stamens consist of the pollen-bearing anthers and the filaments fused about halfway down their length into a tube. The style sits inside the stamen tube and projects from the front.

Measurement of floral and standard calyx lengths in mixed U. gallii/U. minor and single species populations throughout Dorset found the average floral standard and calyx lengths to be significantly different (single-species populations, standard, t = 19.4, P < 0.001; calyx, t = 16.3, P < 0.001; mixed populations, standard, t = 28.6, P < 0.001; calyx, t = 25.8, P < 0.001, Bullock et al. 1998). Most U. gallii plants had longer standards and calyces than any U. minor plant, although there was an overlap in flower sizes, with both species having flowers with mean standard lengths of 11–12.5 mm and mean calyx lengths of 8–10 mm (Bullock et al. 1998). Millener (1952) records pollen grains to be 28 µm in length for U. gallii and 22 µm in length for U. minor.

Flowers are adapted to ensure that anthers and stigma contact the underside of bee pollinators. Although the flowers do not contain nectar, bees force entry into a fresh flower as though looking for nectar. The keel petals are forced apart releasing stamens and style, which are brought sharply onto the underside of the pollinator in an explosive pollen presentation mechanism, the force dusting the underside of the pollinator with pollen. Once ‘exploded’ the flower hangs limply and is seldom visited again. Both species are largely self incompatible. However when selfing is induced by hand pollination, a small proportion of seedpods are produced, indicating that neither species is exclusively self incompatible (C.M. Ormston, unpublished data).

Kirchner & Bullock (1999) showed that the bee pollinator community of U. gallii and U. minor did not differ significantly and was characterized by a high degree of overlap in the assemblage, as quantified by the Proportionality Similarity Index, Ps = 0.83. Apis mellifera, Bombus terrestris, B. lucorum and B. humilis have all been observed visiting U. gallii and U. minor flowers. Bombus ruderarius, B. lapidarius and Andrena ovatula have also been observed visiting U. gallii flowers. Four hoverfly species (Syrphidae, Diptera) also visited flowers (Table 2) but further research is necessary to assess if these visits have significance for pollination.

Table 2.  Insect species observed visiting and potentially pollinating Ulex minor and U. gallii flowers. Total numbers of observations with their relative frequencies are listed for each species. Reproduced from Kirchner & Bullock (1999)
 Insect speciesUlex gallii numberUlex gallii frequencyUlex minor numberUlex minor frequency
BumblebeesBombus terrestris/lucorum280.364160.356
Bombus humilis 50.065 40.089
BeesAndrena ovatula 50.065 20.044
Apis mellifera 20.026 00.000
HoverfliesSphaerophoria scripta 90.117 30.067
Syritta pipiens210.273190.422
Eristalis sp. 30.039 10.022
Episyrphus sp. 40.052 00.000

(b) hybrids

Although both species flower from July to early autumn and are outcrossing, there is no chromosomal evidence of any hybrids (Bullock et al. 1998; Kirchner & Bullock 1999). Millener (1952) failed to produce hybrid seeds or pods in crossing experiments. Later work by C.M. Ormston (unpublished data) did successfully produce hybrid seed pods resulting from hand pollination experiments in the field, indicating that the absence of hybrids within natural populations may result from postzygotic mechanisms reducing the success of any seeds derived from hybrid crosses. Ulex gallii × U. europaeus hybrids are commonly described based on intermediate vegetative and floral characteristics (Benoit 1962; Gloaguen 1986; Misset & Fontenelle 1992; Stace 1997) but only Misset & Fontenelle (1992) give evidence which is based on differing isoenzyme systems of U. gallii and U. europaeus, showing that putative hybrids have elements of both isoenzyme systems. The flowering season of U. gallii × U. europaeus hybrids is described as extending over the seasons for both parent species (Millener 1952; Gloaguen 1986; Stace 1997). There have been no suggestions of U. minor × U. europaeus hybrids to date.

(c) seed production and dispersal

Pod samples from 12 U. gallii populations in Devon, Dorset and Somerset and 10 U. minor populations in Hertfordshire, Dorset and the Isle of Wight were collected by Proctor in 1958. A satisfactory separation of these two species can be obtained from pod length and breadth measurements, U. gallii having greater pod dimensions than U. minor. Pods of U. gallii are approximately 10 mm long (Millener 1952; Misset 1992). There are indications that U. gallii averages slightly more seeds per pod than U. minor but seed number and ovule number are highly variable in both species. Proctor (1965) records a mean seed number per pod of 1.8–2.9 for U. gallii and 1.4–2.8 for U. minor. In Dorset mean seed number per pod has been recorded as 1.70 for U. gallii (SE ± 0.04), as opposed to 1.59 (SE ± 0.03) for U. minor (Stokes 2002). López et al. (2000) record a mean seed number per pod of 2.77 (SD ± 1.14) for U. minor populations sampled in south-west Spain.

Mean ovule number per flower is recorded from the populations surveyed by Proctor (1965) as 4.1–5.8 for U. gallii and 4.4–5.9 for U. minor. Ulex gallii plants sampled in Asturias province, northern Spain, produced 5.80 (SD ± 0.69) ovules per flower and the seed to ovule ratio has been measured at 0.34 (SD ± 0.15) (Gutiérrez et al. 1996). The probability of an ovule producing a seed was dependent on its position in the ovary; ovules in a stylar position were favoured, the smallest seeds within the pod being in a stylar position more frequently (Gutiérrez et al. 1996). The fruit to flower ratio for a population of 450 U gallii individuals was 0.002 (Gutiérrez et al. 1996). Herrera (1999) sampled a population of 100 U minor individuals in Andalucia, southern Spain, recording mean ovule number per ovary as 6.2 (SE ± 0.1), mean seed number per fruit as 2.3 (SE ± 0.1) and the seed to ovule ratio as 0.37 (SE ± 2.0) (Herrera 1999). Again the position of ovules in the ovary influences the probability of seed set, with ovules in the stylar position being favoured (Herrera 1999). López et al. (2000) record 5.80 (SD ± 1.24) ovules per pod for U. minor and a seed to ovule ratio of 0.48 (SD ± 0.17). The fruit to flower ratio for U. minor has been recorded as 0.16 in southern Spain (Herrera 1987).

Seed mean dry weight obtained from samples collected in Dorset is recorded as 6.7 mg (SD ± 0.98) for U. gallii and 5.9 mg (SD ± 1.48) for U. minor (Stokes 2002). Seed weight can vary between populations; López et al. (2000) record seed dry weight for U. minor to be 3.2 mg (SD ± 0.01) and seed size to be 1.97 mm (SD ± 0.20) for populations in south-west Spain.

Seed production per plant is dependent on both plant size and geographical location (see V (B) and Fig. 6). Contrary to the situation observed at the range margin in Dorset, at the core of each species range no relationship was observed between size and seed production per plant (Stokes 2002). For U. gallii in Devon, seed number per pod varies with the size of a plant; larger plants produce a reduced number of seeds per pod compared to smaller plants, explaining the lack of size-dependent seed production per plant. In Surrey, however, there was no relationship between the number of seeds per pod and the size of a U. minor plant. The lack of an allometric relationship between seed output and plant size in this case presumably results from high variability in the allometric relationships. Further research is needed to investigate the reason for the decline in the number of seeds per pod with plant size observed in U. gallii populations in Devon; possible causes include higher ovule abortion. For U. gallii plants in Devon mean seed production per plant was 10.63 (SE ± 1.12), and for U. minor plants in Surrey mean seed production per plant was 27.51 (SE ± 1.21), a reflection of the greater number of flowers per plant produced by U. minor (see VI (C)).

For a given density of plants, it might be expected that U. minor populations in Dorset would produce more seed than U. gallii. However in reality, for a sample of 200 Ulex plants measured in Dorset (where seed production is dependent on size), total seed production of U. gallii plants was greater than that of U. minor (Stokes 2002). This is partly because the populations of U. gallii surveyed generally contained larger individuals than U. minor, but also because in Dorset U. gallii produces a greater number of seeds per pod than U. minor and no decline in seed number per pod is observed with size.

Dispersal occurs by explosive dehiscence; thus the majority of seed is dispersed within the vicinity of the parent plant. The mean ballistic dispersal distance of U. minor individuals is recorded as 18.4 cm (J.M. Bullock, personal communication). Ants remove on average 79.5% of the dehisced seed of U. minor individuals, thus increasing dispersal distance. Ballistic dispersal was measured at a maximum of 195 cm, whereas ant-dispersed seed was recorded as travelling up to 4 m from the parent plant (J.M. Bullock, personal communication). Ants observed carrying U. minor seeds in Dorset heathlands are Myrmica ruginodis, M. scabrinodis, Tetramorium cespitum and more rarely Formica fusca. In laboratory studies Myrmica ruginodis, M. scabrinodis and Tetramorium cespitum carry both U. minor and U. gallii seeds but Formica fusca, Lasius niger and L. platythorax do not carry either (J.M. Bullock, personal communication).

(d) seed germination and viability

There is strong evidence for population regulation in U. gallii and U. minor through the density-dependent control of seedling emergence. In order to quantify seedling emergence and survival at a range of densities, sowing experiments were conducted across the geographical ranges of U. gallii and U. minor (Stokes 2002). Seed collected from several sites in Dorset was sown at a depth of 0.5 cm at a range of densities into 25 × 25 cm plots located within degenerate vegetation at sites across the range. Seedling emergence declined at higher densities. The reduction in seedling emergence with an increase in sowing density from 3200 m−2 to 20 000 m−2 was found to be 48% for U. gallii and 31% for U. minor. Overall seedling emergence has been recorded as 28.5% higher in the larger seeded U. gallii.

Seed bank size of U. gallii ranges between 126 m−2 and 707 seeds m−2 according to Thompson et al. (1997). Allchin (1998) found seed bank densities of U. gallii at Aylesbeare Common in Devon, southern England, to be < 500 m−2. Estimates within Dorset heaths range from 23 to 395 seeds m−2 for U. gallii and U. minor (Stokes 2002), values that are considerably lower than those for the related but much larger U. europaeus. Seed longevity in the seed bank was examined through burial experiments in Dorset heaths (Stokes 2002). Seed of each species was collected from plants at all sites in Dorset prior to burial. Seed was then retrieved from Dorset sites at 3-month intervals over 1 year and tested for viability using 2,3,5-tetrazolium chloride. For U. gallii 37.5% of the seed was viable 1 year after production, whereas for U. minor 30% of seed retained viability.

Gutiérrez et al. (1996) examined the effects of seed weight on the probability of germination and seedling weight in U. gallii. Seedling weight was dependent on seed weight but seed weight had no effect on the probability of germination.

(e) seedling morphology

Seeds have two cotyledons; stomatal frequencies on the upper surface of the cotyledon are recorded as 140 mm−2 for U. gallii and 150 mm−2 for U. minor (Millener 1952).

In both species the first pair of leaves is trifoliate (Fig. 9). All subsequent leaves are reduced to small scale-like appendages or modified into the characteristic spines or phyllodes. Stomatal frequencies on the leaves are recorded as 160 mm−2 for U. gallii and 300 mm−2 for U. minor (Millener 1952). As the Ulex seedlings develop and leaves become reduced from the compound, through flat–simple, to triangular or linear–spinous, stem ribbing becomes more and more pronounced, with consequent increase in the photosynthetic capacity of the stem (Millener 1952). Mean fresh weight of U. gallii seedlings 30 days after imbibition is in the range 40–80 mg (Gutiérrez et al. 1996).

Figure 9.

Seedling morphology of Ulex gallii and U. minor at (a) 1 week and (b) 3 weeks’ growth. Species cannot be distinguished by eye at either stage of development.

Figure 8 shows the root systems of 6-week-old seedlings. The root systems of U. gallii seedlings are deep and narrow with relatively short, unbranched roots. Ulex minor seedlings produce more spreading, branched roots with secondary roots. Millener (1952) records the mean depth of root penetration for 6-week-old seedlings to be 12.5 cm ± 0.71 for U. gallii and 6.8 cm ± 0.63 for U. minor. Mean shoot height at 6-weeks’ growth is recorded to be 2.8 cm ± 0.16 for U. gallii and 2.8 cm ± 0.23 for U. minor (Millener 1952).

Measurement of 1-year-old seedlings indicated that U. gallii produced larger shoots and roots than U. minor (Stokes 2002). However the differential between the root and shoot growth in the two species was such that U. minor had a larger root : shoot ratio at this stage of development.

IX. Herbivory and disease

(a) animal feeders and parasites

Invertebrate fauna associated with U. minor and U. gallii are given in Table 3. Ulex minor appears to host a greater number of phytophagous fauna than U. gallii. Pre-dispersal seed predation by the seed weevil Apion (Exapion) ulicis may reduce the seed rain of U. gallii and U. minor by 1–20% (Stokes 2002). Predation by A. ulicis was found to be density-independent. In mixed stands where the two Ulex species grow together U. minor was the preferred choice of host plant. Ulex minor pods are possibly inherently weaker and more susceptible to invasion than U. gallii pods, the greater rigidity of the stems and longer spine length of U. gallii indicating a higher investment in plant defence. Apion uliciperda also attacks U. gallii seeds (Gutiérrez et al. 1996).

Table 3.  Invertebrate species recorded as feeding on Ulex gallii and Ul minor
Host speciesOrderFamilySpeciesPhyto- phagous stageFeeding sourceReference
U. galliiLepidopteraOecophoridaeAgonopterix ulicetella StaintonLarvaeYoung foliageMichaelis et al. (1981)
U. minorLepidopteraOecophoridaeDepressaria umbellana StephensShootsZwofler et al. (1963)
U. gallii & U. minorLepidopteraLycaenidaeCallophrys rubi LinnaeusLarvaeHeath et al. (1984); Howarth et al. (1973); Allan et al. (1949); Noble et al. (1975)
U. minorLepidopteraPyralidaePempelia genistella DuponchelLarvaeMature foliageEmmet et al. (1979)
U. gallii & U. minorLepidopteraScythrididaeScythris gradipennis HaworthLarvaeMature foliageAgassiz et al. (1984); Emmet et al.(1979); Emmet et al. (1992)
U. minorHomopteraAphididaeAphis ulicis WalkerLarvae, adultsStroyan et al. (1984)
U. minorColeopteraApionidaeLepidapion pseudogallaecianum HoffmannSeedsEhret et al. (1990)
U. minorColeopteraApionidaeApion lemovicinum HoffmannEhret et al. (1990)
U. minorColeopteraApionidaeApion uliciperda PandelléSeedsEhret et al. (1990)
U. gallii & U. minorColeopteraApionidaeApion scutellare KirbyAdultsMines stemsHoffmann et al. (1958); Ehret et al. (1990); O’Donnell et al. (1986); Zwofler et al. (1963)
U. minorColeopteraApionidaeApion elongatulum DesbrochersSeedsHoffmann et al. (1958)
U. minorColeopteraApionidaeApion ulicis ForsterLarvae, adultsSeedsHoffmann et al. (1958); Ehret et al.(1990); O’Donnell et al. (1986); Zwofler et al. (1963)
U. minorColeopteraApionidaeProtopirapion atratulum GemarFlowersHoffmann et al. (1958); Morris et al. (1990); 1963O’Donnnell etal. (1986); Zwofler et al. (1963)
U. minorColeopteraCurculionidaePolydrusus confluens StephensAdultHoffmann et al. (1950); Kloet et al. (1977)
U. minorColeopteraCurculionidaeHypera trilineata MarshZwofler et al. (1963); Hoffmann et al. (1954)
U. minorColeopteraCurculionidaeHypera venusta FabriciusHoffmann et al. (1954)
U. minorColeopteraCurculionidaeSitonia regensteinensis HerbstHoffmann et al. (1958); Kloet et al.(1977); O’Donnell et al. (1986)
U. minorColeopteraCurculionidaeSitonia tibialis HerbstZwofler et al. (1963)
U. minorColeopteraCurculionidaeSitonia striatellus GyllenhalAdultHoffmann et al. (1950)
U. minorThysanopteraThripidaeOdontothrips ignobilis BagnallLarvae, adultsMound et al. (1976)
U. minorDipteraCecidomyiidaeAsphondylia ulicis VerrallFlower bud gallsZwofler et al. (1963)

In France, Chorthippus binotatus feeds exclusively on the leaves of U. minor during the nymphal stages (Picaud et al. 2002). The subspecies C. binotatus binotatus has been observed to shift its diet during the season, feeding on stems and shoots of U. minor in June to July and transferring to flowers of the same species in August, early September and October (Picaud et al. 2002).

(b) plant parasites and diseases

The Hyphomycete fungus Coniothyrium sphaerospermum Fuckel affects the spines of U. gallii (Ellis & Ellis 1985). The smut Thecaphora deformans Dur. & Mont. infects the flowers, fruits and seeds of U. minor (Brett 1966). Ellis & Ellis (1985) record that U. minor pods affected by T. deformans become deformed, with shrivelled seeds containing rusty or chocolate brown masses of spore balls. Microthyrium cytisi Fuckel var. ulicis-gallii J.P. Ellis, an ascomycete, is found on dead leaves and branches of U. gallii and U. minor from January to March (Ellis & Ellis 1985). Fisher et al. (1986b) isolated fungal endophytes of 14 different genera from U. gallii spines and stems. An increase in the frequency of fungal endophyte isolation was found with age. The median number of fungal endophyte isolations per fragment (spine or stem) of U. gallii was 2–3.

X. History

The Iberian peninsula is generally regarded as the evolutionary centre of Ulex species (Rothmaler 1941). In 1942, Rothmaler subdivided the genus in two sections, Sampaioa and Neowilkommia; the former covered the most meridional area, while the latter extended northward along the Atlantic border of Europe, to Denmark and the British Isles. Within Britain, U. gallii was first recorded in Dorset by Planchon (1849).

Acknowledgements

We thank David Roy and A.W. Sheppard for their assistance with the herbivory and disease section, Henry Arnold for providing the distribution maps, David Read for supplying information on mycorrhiza and finally Arthur Willis and the associate editors for their comments on the manuscript. Financial support from NERC is gratefully acknowledged.

Ancillary