Microsite availability and establishment of native species on degraded and reclaimed sites


  • Asrun Elmarsdottir,

    Corresponding author
    1. Colorado State University, Rangeland Ecosystem Science Department, Fort Collins, CO, USA and
      Present address and correspondence: A. Elmarsdottir, The Icelandic Institute of Natural History, Hlemmur 3, Box 5320, 125 Reykjavik, Iceland (fax +354 5900595; e-mail asrun@ni.is).
    Search for more papers by this author
  • Asa L. Aradottir,

    1. Icelandic Soil Conservation Service, Gunnarsholt, 815 Hella, Iceland
    Search for more papers by this author
  • M. J. Trlica

    1. Colorado State University, Rangeland Ecosystem Science Department, Fort Collins, CO, USA and
    Search for more papers by this author

Present address and correspondence: A. Elmarsdottir, The Icelandic Institute of Natural History, Hlemmur 3, Box 5320, 125 Reykjavik, Iceland (fax +354 5900595; e-mail asrun@ni.is).


  • 1Restoration of native plant communities on previously disturbed land requires the identification and modification of environmental factors that impede or restrict ecosystem succession or development. One of the key factors in successful reclamation is to ensure colonization and persistence of native species within an area. Therefore, the identification of microsite types that favour colonization by native species should improve our ability to successfully reclaim degraded areas.
  • 2The objectives of this study were to (i) identify and describe microsite types that were conducive to the establishment of five native plant species (Euphrasia frigida, Galium normanii, Thymus praecox, Cerastium spp. and Rumex acetosa) on degraded and reclaimed sites in Iceland; and (ii) to determine the effects of reclamation activities that included the application of manure and inorganic fertilizer on the distribution of microsite types and seedling establishment. Reclamation sites of different ages were examined on an eroded area where the target community was a grass or shrub heath.
  • 3The spatial distribution of plant establishment was not random but varied among microsite types and study sites. More seedlings than expected were found in microsite types characterized by small rocks and biological soil crusts. These may have entrapped seed and provided more soil stability and moisture than other microsite types. The cover of these microsite types varied among the study sites but the pattern of seedling establishment among the five native forbs was similar.
  • 4Synthesis and applications. Practical reclamation methods are improved by identifying those factors that promote successful seedling establishment by native species. Application of fertilizer without additional seeding proved to be a simple reclamation approach that increased the availability of one favourable microsite type, enhanced the establishment of native species and subsequently allowed vegetation cover to expand. Seed of native plants may be difficult to obtain commercially, but this approach can be used on degraded land where propagules are available but recovery is slowed by soil instability or nutrient limitations.


Both abiotic and biotic factors are critical at various stages in the development of plant communities, and reclamation strategies should be based on our knowledge of the effects of these factors on plant establishment (Munn, Harrington & McGirr 1987; Smith, Redente & Hooper 1987; Tucker, Berg & Gentz 1987; Pyke & Archer 1990). In the early stages of plant community succession, vegetation development proceeds slowly on poorly developed soils often characterized by deficiencies in macronutrients and poor physical features (Bradshaw 1983; Tsuyuzaki, Titus & del Moral 1997). Harper (1977), among others, has suggested that seed dispersal, seedling germination and establishment are precarious stages in the life cycle of plants. Therefore, the environment immediately surrounding a seedling will be important in determining the dynamics of plant populations on a local scale, as well as the composition of plant communities. Harper, Williams & Sagar (1965) proposed the term ‘safe site’ to describe the specific conditions that allow the seed of a particular species to germinate and emerge successfully from the soil. The importance of such sites and the abiotic factors that characterize them has been emphasized (Urbanska 1995; Tsuyuzaki et al. 1997), but they have rarely been studied during reclamation where the aim is to sustain selected plant species. Knowledge of the colonization patterns of native species in relation to microsite types (soil surface characteristics) that are conducive to seedling establishment is an important consideration in reclamation. Such knowledge may contribute to the design of effective reclamation strategies and be useful in assessing the importance of human intervention to ensure successful rehabilitation of disturbed areas where the availability of propagules is not a limiting factor.

The first objective of this study was to identify and describe microsite types that were conducive to the establishment of five native plant species on eroded and reclaimed sites. The second objective was to determine the effects of reclamation activities on seedling establishment in relation to the distribution of microsite types. To address these objectives, sites of different ages since reclamation efforts began were examined on an eroded area where the target community was a grass or shrub heath.


study area

The research was conducted on an eroded area with a chronosequence of reclamation sites at the Holar farm, which is situated in southern Iceland (63°59′N, 19°57′W; c. 150 m a.s.l.). The climate is maritime, with cool summers and mild winters (Takhtajan 1986). The closest synoptic weather station, Hæll (64°04′N, 20°15′W; 121 m a.s.l.), is about 17 km from the study area. The mean temperature ranged from a low of −1·2 °C in January to a high of 11·7 °C in July 2000 and annual precipitation was 1150 mm (Icelandic Meteorological Office 2001).

Anthropogenic and natural disturbances have contributed to land degradation and soil erosion at the Holar farm and the surrounding area. The nearby Hekla volcano is very active. It has erupted 20 times in historical times, producing both tephra and andesitic lava flows (Gudmundsson 1996) and has occasionally strewn tephra over the study area. The history of the Holar probably resembles that of large areas in Iceland where wood gathering and heavy grazing has destroyed native birch woodlands (Thorsteinsson 1986). Loss of woodlands was often followed by soil erosion and barren land (Thorarinsson 1961; Arnalds et al. 2001). The resulting soils are poorly developed, low in nitrogen and carbon content (Elmarsdottir 2001) and the surface is sandy-gravel, as frost-heave lifts stones to the surface. Furthermore, the soils (Typic or Lithic Vitricryands) have low allophanic content and rather limited water holding capacity (Arnalds & Kimble 2001).

The study area was approximately 50 ha in size and the topography nearly level (some 1–5% slopes). Four study sites were selected within the area, one unreclaimed site (will be referred to as control) and three reclamation sites that had undergone 2, 5 or 11 years of reclamation effort. The reclamation objective was to increase vegetation cover on eroded areas and improve the land for grazing animals. Reclamation activities began with spreading a 1- to 2-cm thick manure layer (organic fertilizer), followed by application of about 150 kg ha−1 of inorganic NPK (20 : 12 : 8) during the second year. Inorganic fertilizer was then applied to each site every second or third year thereafter. Approximately 350 sheep graze the study area and vicinity during the early growing season (mid-June to mid-July), 60 during the summer months and 200 sheep during the autumn (late September to early November).

The vegetation at the control site consisted mainly of forbs (Thymus praecox Opiz, Silene acaulis (L.) Jacq., Rumex acetosa L., Galium normanii O. Dahl), and grasses (Festuca richardsonii Hooker, Agrostis stolonifera L.) (Elmarsdottir 2001). Racomitrium ericoides (Brid) Brid. was the most common moss, but others (e.g. Bryoerythrophyllum recurvirostrum (Hedw.) Chen, Pogonatum urnigerum (Hedw.) P. Beauv. and Schistidium papillosum Culm.) were also common. Lichen species present were Peltigera leucophlebia (Nyl.) Gyeln., Stereocaulon alpinum Laur., S. rivulorum Magn. and P. canina (L.) Willd. Nomenclature for vascular species, mosses and lichens follow Kristinsson (1998), Johannsson (1998) and Kristinsson (1997), respectively.

Cover of vascular species was 10% at the control site and at the reclamation sites that had undergone 2, 5 or 11 years of reclamation effort, the cover was 50%, 77% and 75%, respectively (Elmarsdottir 2001). Cover of moss was less than 10% on the control site and the 2-year-old site and was 55% on the oldest reclamation site, but cover of lichens was less than 3% at all sites. Plant species diversity was greatest on the 5-year-old site and least on the control site, and grasses were more abundant on the reclaimed sites (Elmarsdottir 2001). Above-ground biomass, measured in adjacent plots in 1999, was 0·3 and 8·4 t ha−1 for the control and 11-year-old reclamation sites, respectively (Aradottir et al. 2000).

sampling and data analysis

Within each of the four study sites, four plots (10 × 10 m) were chosen subjectively to represent vegetation that was as homogeneous as possible within each plot, but encompassed what appeared to be differences in vegetation composition and aspect at that site. Within each plot, 10 quadrats (0·5 × 0·5 m) were placed randomly. All data were collected in July and August 2000.

Cover of each microsite type (Table 1) was assessed by 25 equally distributed point measurements within each quadrat. Seedling density was recorded within a portion (0·5 × 0·3 m) of each quadrat. All seedlings from the current growing season were identified to species or genus, and the relevant microsite type was recorded.

Table 1.  Description of the nine microsite types used to derive indicators for safe sites for seedlings in nonreclaimed and reclaimed plots at the Holar study area
Microsite typeDescription and characteristics
Biological soil crustsBiological soil crusts* and ≤ 1-cm thick mosses
MossesMosses > 1 cm thick
LichensAll lichens
GrassesAll grasses
ForbsAll forbs
SandSand ≤ 0·2 cm in diameter on the surface
Small rocksRocks 0·2–2 cm in diameter on the surface
Large rocksRocks > 2 cm in diameter on the surface
LitterCombination of litter, humus and dung

The data were analysed with sas programs and were treated both as continuous and categorical data (SAS 1999). It should be noted that true replication of sites does not exist (Hurlbert 1984) and anova with subsampling was used to compare data among sites where plots were nested within sites and quadrats were nested within plots. Data were analysed with anova by using the sas proc mixed procedure with site as a fixed effect and plots within a site as the error term. If a significant (P < 0·05) overall F-test was found, individual site means were compared using a pairwise t-test. The Type I error rate was 0·05 per comparison. Variance was examined for homogeneity with residual plots and, when necessary, data were transformed to ensure normal distribution of residuals. Cover data of microsite types and density of all seedlings were transformed using a square-root prior to anova analysis and the data were log (natural) transformed to test for differences in seedling density among sites for each species. All means and values in tables and graphs are presented using nontransformed data.

Chi-square goodness-of-fit statistics were used to determine if observed frequency of seedlings showed a departure from a random occurrence (expected) among microsite types (Conover 1980). The analyses were carried out separately for each reclamation site. The observed frequency was the frequency of seedlings found in each microsite type within a reclamation site. The expected frequency was calculated as the total frequency of seedlings within a reclamation site multiplied by the proportion (cover) of each microsite type within that site. Data for microsite types that had low expected frequencies were combined for analysis to ensure that no more than 20% of microsite types had an expected number < 5 (Conover 1980). As the seedling data was based on numbers within quadrats (cluster sampling), the analysis might fail to meet the assumption of independent distribution, which would increase Type I error rates and results might be overstated (Garson & Moser 1995). To counter this problem, we used a 99% significance level and a conservative test statistic, i.e. the deviation for each microsite type had to be equal to or greater than the critical chi-square value for the full analysis for each reclamation site.


microsite types

The most common microsite types at the study sites were biological soil crusts, small rocks, large rocks and sand. The cover of sand, small and large rocks decreased while the cover of biological soil crusts, mosses, grasses, and forbs increased with greater age of reclamation sites (Fig. 1). The cover of small rocks was 39% on the nonreclaimed control site, compared with 1% cover on the oldest reclamation site. In contrast, the cover of the microsite type characterized as biological soil crusts was less than 1% on the control site, but 48% on the oldest reclamation site.

Figure 1.

Cover (mean + 1 SE) of each microsite type within the sites with different ages of reclamation. Different letters above bars for each microsite type represent significant differences (P < 0·05) among age of reclamation.

seedling density

Seedling density was approximately 100% greater on the 2-year-old site than the control site, but on the 5- and 11-year-old sites, the seedling density was not significantly different from the control site (Fig. 2). Overall, 4053 seedlings were counted on all plots at the study area, representing 14 species or genera of forbs, grasses and sedges (Table 2). The number of species or genera of seedlings recorded was greatest on the control site (13), and smallest on the oldest reclamation site (9). It was difficult to detect grass and sedge seedlings in the reclaimed sites as the vegetation became dense. Therefore, seedling densities on the reclaimed sites may be underestimated.

Figure 2.

Seedling density (number of seedlings per m−2 + 1 SE) within each site of varying reclamation age. Different letters above bars represent significant differences (P < 0·05) among age of reclamation.

Table 2.  Seedlings density (number of seedlings per m2 ± 1 SE) for each species, or group of species, averaged over all sites and within each site. Different superscript letters by means within a row represent significant differences (P < 0·05) among age of reclamation
Species Age of reclamation (years)
All sites averageControl2511
Alchemilla alpina L. 1   0 ± 0a   0 ± 0a 0·2 ± 0·2a 0·2 ± 0·2a
Arenaria norvegica Gunn. 2 1·5 ± 1·0b 4·0 ± 3·1b 1·3 ± 0·9b 0·2 ± 0·2a
Cardaminopsis petraea (L.) Hiit. 5 5·7 ± 4·1a 8·0 ± 3·0a 6·8 ± 7·3a   0 ± 0a
Cerastium spp.3813·7 ± 9·0a38·3 ± 21·0a83·3 ± 60·4a15·3 ± 5·7a
Euphrasia frigida2718·0 ± 10·8a67·0 ± 14·7b18·0 ± 15·0a 6·3 ± 4·1a
Galium normanii1915·8 ± 4·1a28·5 ± 21·4a24·5 ± 4·2a 6·0 ± 1·5a
Grasses and sedges 3 9·7 ± 1·4c 1·0 ± 0·9b   0 ± 0a   0 ± 0a
Minuartia spp. 1 1·2 ± 1·4a 0·5 ± 0·5a 0·5 ± 0·5a 0·2 ± 0·2a
Rumex acetosa3015·0 ± 7·7a26·0 ± 11·5a21·0 ± 7·0a59·7 ± 64·4a
Sedum acre L. 5 5·0 ± 2·7bc13·5 ± 3·7c 1·0 ± 1·2ab 0·2 ± 0·2a
Silene acaulis 3 2·7 ± 1·6b 7·0 ± 2·6b 1·3 ± 0·8b   0 ± 0a
Silene uniflora A. Roth 1 2·5 ± 2·8b   0 ± 0a   0 ± 0a   0 ± 0a
Stellaria media (L.) Vill. 1 0·2 ± 0·2a 0·8 ± 0·6a 0·2 ± 0·2a   0 ± 0a
Thymus praecox3633·8 ± 8·9b90·2 ± 16·0c11·2 ± 6·1ab 8·7 ± 7·6a

seedling preferences for microsite types

Cerestium spp., T. praecox, R. acetosa, Euphrasia frigida Pugsl., and G. normanii were the most common species in the study area and accounted for 22%, 21%, 17%, 16% and 11%, respectively, of the total seedling number (Table 2). These five groups of forbs were the only species considered separately in analyses of preferences for microsite types. They are all perennial, except for E. frigida (Grime, Hodgson & Hunt 1988), are less than 40 cm in height at maturity and are common in Iceland (Kristinsson 1998). The Cerastium spp. group was composed of two species, Cerastium alpinum L. and Cerastium fontanum Baumg.

The goodness-of-fit analysis for distribution of all seedlings in relation to microsite types showed that more seedlings were found adjacent to small rocks in the control (51%) and 2-year-old reclamation site (57%) than would be expected from random occurrences (Table 3). In contrast, fewer seedlings than would be expected from random occurrence were found in sand (16%) in the control site and in sand (24%), large rocks (9%), litter (2%) and forbs (< 1%) in the 2-year-old reclamation site. More seedlings were found in biological soil crust than expected from random occurrence in the 5- (69%) and 11-year-old (81%) reclamation sites, and fewer seedlings in the grass (13%), large rock (≤ 2%) and forb (< 1%) microsite types. There were also fewer seedlings than expected from random occurrence in the moss microsite type (3%) at the 11-year-old reclamation site. No seedlings were observed in the microsite type characterized by lichen. Analysis of individual species showed the same general pattern, but there were fewer incidents where seedling numbers deviated significantly from random distribution (Table 3).

Table 3.  Number of seedlings and results of Chi-square goodness-of-fit analysis (α = 0·01) for occurrences of seedlings in microsite types within each age of reclamation. ‘+’ indicates that significantly more seedlings were found in the microsite type than would be expected from random distribution. ‘–’ indicates that significantly fewer seedlings were found in the microsite type than would be expected from random distribution, and ns indicates that seedling number did not deviate significantly from random distribution. Empty cells indicate microsite types that had low expected numbers of seedlings; their data was combined in the ‘Other’ category
 Age of reclamation (years)
All seedlingsCerastium spp.Euphrasia frigidaGalium normaniiRumex acetosaThymus praecox
No. of seedlings748170910165808223050092108402108389517114736901561263582035416752
Microsite type
 Sandns nsnsns ns  nsns  nsns    
 Small rocks++ns ns+ns ns+ns ns+ns ns+ns ++ns 
 Large rocksnsnsnsnsnsns nsnsns nsnsnsnsns 
 Grasses ns + nsnsns nsnsns nsns nsnsns
 Forbsns ns  nsns  nsns  nsnsns   
 Biological soil crusts ns++  ++  nsns  +ns  ++  ++
 Mosses  ns  nsns   ns  nsns  ns   ns
 Lichens  ns  ns                 
 Litternsnsns nsns  ns   nsns  ns ns ns  
Othernsns nsnsns nsnsnsnsnsnsnsnsnsnsnsnsnsnsnsnsns
Degrees of freedom  5   7   8  7 3  6  8 4  3  6  5 2 3  6  7 3 3  6  6  6  3  6 4 3
Chi-square 93 987 475269 6 8527052 23277 23 918 56 5413 5229104183 383903525


the role of microsite type in species establishment

Data presented here indicated that the spatial distribution of plant establishment was not random and that establishment varied among the different study sites (Fig. 2; Table 2) and microsite types (Table 3). More seedlings were found growing adjacent to small rocks and in biological soil crusts than would be expected from a random occurrence of seedlings (Table 3). These two microsite types may have characteristics that encourage seedling establishment, e.g. roughness for seed entrapment, soil stability, moisture and nutrient availability.

Seeds that are deposited on exposed soils in windy environments are often blown over the soil surface until they encounter a barrier or depression that traps them (Magnusson 1994; Jumpponen et al. 1999). As seeds become trapped between the soil particles or adjacent to rocks, it allows partial seed burial and soil contact, and therefore a greater chance of successful seedling establishment.

Seed germination and establishment on volcanic material or in the arctic is often inhibited by drought (Bishop & Chapin 1989; Chapin & Bliss 1989). The presence of small rocks will provide shade and shelter, which will most likely improve establishment of seedlings. Biological soil crusts may offer similar advantages. The crust stabilizes the soil surface, retains surface moisture and increases nitrogen fixation from cyanobacteria, thereby encouraging seedling establishment (Bliss & Gold 1999; Dickson 2000; Warren 2001). Our results are in agreement with other studies in Iceland that found biological soil crusts to be important for colonization of Betula pubescens Ehrh. (Magnusson & Magnusson 1990; Aradottir & Arnalds 2001), Salix lanata L. and S. phylicifolia L. (Aradottir et al. 1999).

Other microsite types may inhibit successful seedling establishment. Studies have shown low seedling establishment on bare soil and fine-textured surfaces (Chambers, MacMahon & Haefner 1991; Winkel, Roundy & Cox 1991; Tsuyuzaki et al. 1997). For example, sand is unstable, has few barriers or depressions to trap seeds, is subject to erosion and frost–heave, and does not retain moisture (Arnalds & Kimble 2001). In the microsite types characterized by forbs and grasses, the vascular plants may compete with the seedlings for resources. However, studies have shown that neighbouring plants can have both positive and negative impacts on seedling establishment, but the impacts can be different at various life stages (Fowler 1988; Callaway & Walker 1997). These impacts co-occur in nature and therefore produce complex and variable effects that influence the community structure.

A site suitable for seed germination is not necessarily suitable for successful seedling establishment and growth (Barrett & Silander 1992; Schupp 1995; Urbanska 1995). Schupp (1995) advocates caution in the interpretation of results from field studies of safe sites when the focus is only on a single stage in plant development, as it may give misleading predictions. In the control and the 2-year-old reclamation site most seedlings were found in the small rock microsite type which is common on the sandy-gravel surface of many eroded areas in Iceland. This surface type is formed by frost–heaving of coarse fragments that are underlain by finer materials (Arnalds & Kimble 2001). The frost–heaving may be detrimental for overwinter survival of small seedlings, especially those that are not well anchored (Perfect, Miller & Burton 1988; Aradottir 1991).

differences among study sites

The reclamation treatments at the study sites included the application of manure and inorganic fertilizer to stabilize the surface and supply nutrients and organic matter to the soil (Elmarsdottir 2001). This enhanced seedling establishment and growth leading to an expansion of vegetation cover (Elmarsdottir 2001) and enhanced biomass production (Aradottir et al. 2000). Given the low nitrogen levels and the large size of degraded areas in Iceland (Arnalds & Kimble 2001) it is unlikely that such low fertilization application will cause eutrophication problems.

Most seedlings were found in the small rock microsite types at the control site and 2-year-old site, but greater numbers of seedlings were found in the biological soil crust microsite at the 5- and 11-year-old sites (Table 3). This reflects differences in microsite availability among the study sites (Fig. 1). The small rock microsite together with other nonvegetated microsites, sand and large rocks, was most abundant in the control and 2-year-old sites, but the biological soil crust had the greatest cover in the 5- and 11-year-old sites. Other studies have also shown that biological soil crust can be an important component of the vegetation cover on old reclamation sites in Iceland (Gretarsdottir 2002), and its formation can enhance plant colonization and advance succession after reclamation (Aradottir 1991; Aradottir et al. 1999).

Seedling density within the control site did not differ significantly from that of the 5- and 11-year-old sites (Fig. 2), even though total vascular plant cover was less than 10% on the control site compared with over 75% on the two oldest sites (Elmarsdottir 2001). Low cover of vascular plants on the control site, despite considerable seedling emergence, indicated that seedling survival and growth may have been restricted, e.g. by frost–heave, low soil stability, limited water holding capacity, low nutrient status, and the fact that soil temperature fluctuations were greater than on the oldest reclamation site (Elmarsdottir 2001).

The greater seedling establishment on the 2-year-old site compared with the control site indicated the benefits of the reclamation efforts (Fig. 2). It has been shown that organic mulch increases the number of naturally occurring seedlings (del Moral & Wood 1993), which might be a response to increased nutrient availability and a more stable soil surface as vegetation cover increased (Elmarsdottir 2001). Furthermore, increased local seed rain could be a positive factor as lack of seed has been shown to limit plant establishment in several reclamation projects (Urbanska & Fattorini 2000; Pywell et al. 2002).

On the oldest reclamation site where vascular plant cover was 75% seedling establishment might be limited by competition (Elmarsdottir 2001). Other studies have shown that mature vegetation inhibits seedling establishment because of competition for resources, including safe sites (Gross & Werner 1982; del Moral & Wood 1986; Chambers, MacMahon & Brown 1990).

No effort was made to exclude grazing animals from the study plots. However, grazing may have reduced the availability of seeds and microsites through herbivory and trampling (Oesterheld & Sala 1990; Archer & Pyke 1991; Walker 1999), and affected the rate and patterns of vegetation succession at the reclamation sites (Archer & Pyke 1991; Whisenant 1999).

conclusions and implications for reclamation

The overall goal of reclamation treatments is to promote revegetation and soil development, and to enhance ecological function, which is the key to ecosystem resilience and repair (Bradshaw 1997; Whisenant 1999). However, short-term reclamation to halt degradation has to be balanced with long-term recovery goals for the ecosystem (Holl 2002). Reclamation treatments that result in a diverse plant community are more likely to support and foster ecosystem resilience (Myers 1996) than reclamation resulting in a community dominated by one or a few seeded species.

Results from this study indicated that seeding an area is not always necessary. Seedling density within the control site demonstrated that seeds could germinate fairly easily and most of the species found on the older reclamation sites were already present on the control site (Elmarsdottir 2001). However, the difference in vegetation cover between the control site and the older sites suggest that seedling survival and plant growth was restricted at the control site, in common with many eroded areas that have remained sparsely vegetated for decades or centuries. The fertilizer and manure application probably provided an opportunity for successful seedling establishment, enhanced growth of existing plant individuals and an expansion of vegetation cover. The biomass production increased and soil organic matter accumulated at the study sites (Aradottir et al. 2000; Elmarsdottir 2001). On sites where recovery is slowed by soil instability and nutrient limitations but not the availability of propagules, this simple method of adding fertilizer and manure enhances the colonization and growth of native species and is an alternative to more intensive methods based on seeding or planting.


This project was a part of a M.S. research project at Colorado State University. It was funded by the Icelandic Research Fund for Graduate Students, the Icelandic Research Council Research Programme in Environmental Science and Technology, the Agricultural Productivity Fund in Iceland, Fulbright Association and the National Power Company. We want to acknowledge the farmers at Holar, Kristjan Gislason and Gudrun Audur Haraldsdottir, who allowed us to conduct this study on their land. Field support was provided by Lilja Karlsdottir and Sigmar Metusalemsson. Dr Phillip L. Chapman and Richard J. McNally assisted with statistical and SAS programming. R. H. Marrs and an anonymous referee provided helpful comments on a previous version of this manuscript. All other assistance is also gratefully acknowledged.