High Shrub Density Research Articles

贺兰山牦牛冬春季的生境选择

The yak(Bos grunniens),sometimes called boat of the plateau,is an endemic species of the Qinghai-Tibet Plateau,famous for its resistance to cold and importance as a form of transport.The total number of adults may be close to 10,000 because the total population was estimated to be around 15,000 in 1995.Nevertheless,the species was listed as Vulnerable in the World Conservation Union's Red List of Threatened Animals because it is inferred that it has declined by over 30% over the last 30 years.Poaching,including commercial poaching for meat,has been the most serious threat to wild yaks.The species was listed on CITES Appendix I.Wild yaks have been protected in China since 1962,and are currently listed as a Class I protected animal,which means that they are fully protected by government law.In China,wild yaks are found in a number of large nature reserves,including the Arjin Shan,Chang Tang,Kekexili,Sanjiangyuan,and Yanchiwan Nature Reserves,although none of these reserves provide complete protection from habitat loss or occasional poaching.The yak population in the Helan Mountains is a reintroduced population from Qinghai-Tibet Plateau,and is only distributed in the Halawu Valleys and the surrounding areas,in the Helan Mountain Reserve in Inner Mongolia.Because their habitat selection had not been previously studied,we conducted surveys from December 2009 to January 2010 and from April to May 2010 to study yak habitat selection in winter and spring.Twenty line transects were established along seven valleys.We identified a total number of 103 used plots in winter and 129 in spring.To provide comparison plots for the analysis of habitat selections,we surveyed 188 randomly placed plots in winter and 207 randomly placed plots in spring by systematically placing transects in areas with no obvious evidence of yak use.Data on 18 topographic and biological variables were collected in each plot.In winter,yaks preferred montane coniferous forest,lower altitude(2000 m),habitat dominated by Ulmus glaucescens,more gentle slopes(10°),sites closer to the foot of the mountain,areas distant from water resource(1200 m),moderate distance from human disturbance(2000—4000 m) and abundant hiding cover(70%).In spring,yak preferred subalpine shrubland and meadow,higher altitude(3000 m),lower tree density(1 tree/100 m2),lower tree height(3 m),greater distances from trees(3 m),higher shrub density(40 trees/100 m2),short distance to shrub(1 m) and higher vegetation coverage(70%).Habitat selection of the yak showed significant differences in vegetation type,landform feature,dominant tree species,shrub type,position on mountain(e.g.top or base),slope direction,distance from human disturbance and distance from water resource between winter and spring.Results of principal components analysis indicated that the first principal component axis accounted for 21.10% of the total variance in habitat use during winter,with the highest correlation coefficient for distance to nearest tree,dominant tree species,tree height,tree density.In spring,the first principal components axis accounted for 31.247% of the variance,with highest correlation coefficient for vegetation type,altitude,landform feature,shrub density.Compared to other migrating yak populations distributed in Qinghai-Tibet Plateau,there were significant differences in yak habitat selection between different seasons in the Helan Mountains.Yaks are acclimatized to the particular geography and climate in the Helan Mountains.

Do deer and shrubs override canopy gap size effects on growth and survival of yellow birch, northern red oak, eastern white pine, and eastern hemlock seedlings?

Innovative forestry practices that use natural disturbance and stand developmental processes as models to increase forest complexity are now being considered as a way to conserve biodiversity while managing for a range of objectives. We evaluated the influence of harvest-created gap size (6, 10, 20, 30, and 46m diameter gaps and uncut references) over 12 growing seasons on planted tree seedling growth and survival for four tree species that tend to experience poor recruitment in both managed and unmanaged northern hardwood forests in eastern North America. We expected, based on silvics, that the three mid-tolerant species (yellow birch [Betula alleghaniensis], red oak [Quercus rubra], and white pine [Pinus strobus]) would perform best in intermediate-sized gaps, and the one shade tolerant species (hemlock [Tsuga canadensis]) would perform best in small gaps. However, all four of the species grew taller with increasing gap size, while survival was highest in intermediate gap sizes. Although gap size had statistically significant effects on growth and survival, the magnitude of the effects were modest. With the exception of a small portion of white pine individuals (35% of survivors were >150cm tall), trees were short (<1m) and few survived (<30%) 12years after planting. Evidence from deer exclosures and individual gaps with high shrub (Rubus idaeus) densities suggest that browsing and shrub competition resulted in poor tree growth and survival, and may have constrained the magnitude of many potential tree seedling responses to gap size. Our study highlights the management challenges of using gap size as a tool to influence future forest composition in forests with overly abundant deer and pervasive shrub layers and underscores the importance of silvicultural prescriptions that include measures for reducing these impacts.

Habitat correlates of the red panda in the temperate forests of Bhutan.

Anthropogenic activities and associated global climate change are threatening the biodiversity in the Himalayas against a backdrop of poor knowledge of the region's threatened species. The red panda (Ailurus fulgens) is a threatened mammal confined to the eastern Himalayas, and because of Bhutan's central location in the distributional range of red pandas, its forests are integral to the long-term viability of wild populations. Detailed habitat requirements of the red panda are largely speculative, and there is virtually no ecological information available on this species in Bhutan. Between 2007 and 2009, we established 615 presence/absence plots in a systematic sampling of resident habitat types within Jigme Dorji and Thrumshingla National Parks, Bhutan, to investigate broad and fine-scale red panda habitat associations. Additional locality records of red pandas were obtained from interviewing 664 park residents. Red pandas were generally confined to cool broadleaf and conifer forests from 2,110–4,389 m above sea level (asl), with the majority of records between 2,400–3,700 m asl on south and east-facing slopes. At a finer scale, multivariate analysis revealed that red pandas were strongly associated with old growth Bhutan Fir (Abies densa) forest dominated by a dense cover of Yushania and Arundanaria bamboo with a high density of fallen logs and tree stumps at ground level; a high density of trees, dead snags, and rhododendron shrubs in the mid-storey; and locations that were close to water. Because Bhutan's temperate forests that encompass prime red panda habitat are also integral to human subsistence and socio-economic development, there exists an inadvertent conflict between the needs of people and red pandas. As such, careful sustainable management of Bhutan's temperate forests is necessary if a balance is to be met between the socioeconomic needs of people and the conservation goals for red pandas.

Nest-Site Selection and Nesting Success of Grey-Backed Thrushes in Northeast China

We identified microhabitat features affecting nest-site selection and examined nest-site characteristics associated with success for the Grey-backed Thrush (Turdus hortulorum) in the Dagang Forestry Farm, Jilin Province, northeast China. We collected data from 79 nests from April to August 2008. Twenty-nine nests (36.7%) were successful, 47 (59.5%) failed due to predation, and the rest were either destroyed by storms or abandoned. The overall daily survival rate (DSR) was 0.9563 ± 0.0072. Nest attempts beginning late in the breeding season were more likely to be depredated. Averaged DSR of laying and late nestling (days 7–12 after hatching) periods were higher than those of incubation and brooding (days 1–6 after hatching). We compared habitat variables between nest and random sites and assessed the effects of nest-location and nest-patch characteristics on DSR. Grey-backed Thrushes selected nest sites with shorter ground cover and a high density of small trees and shrubs. DSR was positively related to distance from the nest to the main stem of the nest tree and negatively correlated with horizontal exposure. Further research should focus on identification of nest predators, implications of nest exposure and begging calls on nesting success, and breeding habitat requirements at different spatial and temporal scales of Grey-backed Thrush in fragmented landscapes of northeast China.

Factors affecting roe deer occurrence in a Mediterranean landscape, Northeastern Portugal

The European roe deer population in Portugal is on the southwestern edge of its distribution. Understanding limiting factors that act on these populations enlightens both local aspects concerning their conservation and wider scale aspects of the species bioclimatic envelope, which is crucial for being better able to predict the impacts of environmental change. Accordingly, a survey was conducted to explore roe deer distribution in a 75,000 ha area located in Trás-os-Montes region, a Mediterranean landscape in the northeast of Portugal. Pellet-group counts were used to examine how roe deer distribution was related to habitat structure and composition, landscape structure, and human disturbance. The analysis considered two spatial scales: habitat patch and the wider landscape. At the patch scale, roe deer distribution was positively associated with high density of shrubs and with increasing distance from roads. At the landscape scale, roe deer distribution was negatively associated with spatial heterogeneity, namely mean shape index. Our findings suggest that landscape structure, vegetation composition and distance to roads are all important factors influencing roe deer distribution, highlighting the importance of multi-scale approaches.

Ruffed Grouse Bonasa umbellus habitat selection in a spatially complex forest: evidence for spatial constraints on patch selection

Drumming display sites of male Ruffed Grouse Bonasa umbellus serve as activity centres and occur in a variety of forest types and age classes. We used a case‐control study design and model selection to evaluate a set of predictions about habitat constraints placed upon Grouse given the habitat where their breeding season activity centres were located. We located 110 individual Grouse activity centres near Cloquet, Minnesota, between 2002 and 2005, and 40 activity centres at two other sites in northern Minnesota in 2005. Our most parsimonious model indicated that Ruffed Grouse used logs more than other potential drumming structures (e.g. stumps, dirt mounds or roots) and sites with a higher density of shrubs as compared with unused sites. Predicted values from this model correlated with observed values from independently sampled areas. Structures used by Grouse for their drumming display were characterized by a greater density of aspen stems than unused sites when in young aspen and mature pine, but not when in other forest types. The patterns of habitat selection that we observed supported predictions (1) that the differences in habitats at larger spatial scales (e.g. forest stands or breeding territories) may impose certain constraints on specific sites selected for their drumming display, and (2) that some of these constraints may vary by forest type.

Notes on Asiatic black bears denning habits in the Misaka Mountains, central Japan

We investigated the denning ecology of 19 (12 males and 7 females) radiocollared Asiatic black bears (Ursus thibetanus) from 2001 to 2004 in the Misaka Mountains, central Japan. We approached 14 dens of 5 females on foot. Thirteen dens were located in cavities that formed between the roots of trees growing on the upper edges of landslides or steep inclines, and 1 den was located in a cavity under a rocky ledge. We were unable to access 39 more dens for which we had approximate locations due to a combination of distance from roads, steep terrain, and high shrub densities. All 53 dens we approached or approximately located were on steep slopes. There were few large trees with large cavities in the study area, which may explain why we failed to find any dens in such cavities.

Optimizing Nest Survival and Female Survival: Consequences of Nest Site Selection for Canada Geese

AbstractWe examined the relationship between attributes of nest sites used by Canada Geese (Branta canadensis) in the Copper River Delta, Alaska, and patterns in nest and female survival. We aimed to determine whether nest site attributes related to nest and female survival differed and whether nest site attributes related to nest survival changed within and among years. Nest site attributes that we examined included vegetation at and surrounding the nest, as well as associations with other nesting birds. Optimal nest site characteristics were different depending on whether nest survival or female survival was examined. Prior to 25 May, the odds of daily survival for nests in tall shrubs and on islands were 2.92 and 2.26 times greater, respectively, than for nests in short shrub sites. Bald Eagles (Halieaeetus leucocephalus) are the major predator during the early breeding season and their behavior was likely important in determining this pattern. After 25 May, when eagle predation is limited due to the availability of alternative prey, no differences in nest survival among the nest site types were found. In addition, nest survival was positively related to the density of other Canada Goose nests near the nest site. Although the number of detected mortalities for females was relatively low, a clear pattern was found, with mortality three times more likely at nest sites dominated by high shrub density within 50 m than at open sites dominated by low shrub density. The negative relationship of nest concealment and adult survival is consistent with that found in other studies of ground-nesting birds. Physical barriers that limited access to nest sites by predators and sites that allowed for early detection of predators were important characteristics of nest site quality for Canada Geese and nest site quality shifted within seasons, likely as a result of shifting predator-prey interactions.

Evaluation of competitive effects of green alder, willow and other tall shrubs on white spruce and lodgepole pine in Northern Alberta

In boreal forests of western Canada, lodgepole pine ( Pinus contorta Dougl. ex. Loud.) and white spruce ( Picea glauca (Moench) Voss) often grow together with numerous tall shrubs such as green alder ( Alnus crispa (Ait.) Pursh) and little-tree willow (Salix spp .). In an area south of Grande Prairie, Alberta, Canada, we examined the effects of shrubs, herbs and other trees on nutrient and light availability and growth of white spruce and lodgepole pine. For white spruce the best competition measure (tested against volume increments of the past 3 years) was visually estimated % ground cover times the height of the competitor (VCHT) with light (DIFN) ranking in third place. For lodgepole pine, DIFN was the best competition measure for predicting volume increment and the best competition index was again VCHT. Taller conifers had a stronger competitive effect than tall shrubs, with their effect on white spruce being larger than that on lodgepole pine. Comparison of the effects of alder and willow indicates that both species are having negative effects on growth of pine and spruce on these sites with green alder having a stronger influence on spruce than pine due to the smaller size of the spruce. Herbaceous cover was having more effect on white spruce than lodgepole pine and may be related to both their height and differences in rooting depth. The results did not show any significant relationship ( p > 0.05) between nitrogen availability and cover of alder and willow and between conifer foliar nutrient concentrations and cover of alder and willow. Our study suggests that the current Alberta regeneration standards may be overestimating the effect that competing vegetation has on conifer growth since significant growth reduction occurs only at high shrub densities.

Landscape ecology of the burrowing bettong: fire and marsupial biocontrol of shrubs in semi-arid Australia

Prior to European settlement, medium-sized marsupials, especially bettongs (Bettongia spp.), were widely distributed across arid and semi-arid Australia. Most disappeared rapidly in the late 1800s in the earliest settled rangelands such as the West Darling region of western New South Wales following the spread of domestic herbivores, rabbit invasion, exotic predators and loss of habitat. Because the burrowing bettong (Bettongia lesueur) is the only fossorial macropod species, it left a clearly visible record of its past presence, distribution and habitat preferences in the form of substantial relict warrens, particularly in stony, ‘hard-red’ habitats. With the reduction in fire frequency because of excessive grazing pressures following European settlement in the 19th century, there was a rapid increase in the density of unpalatable native shrubs. We examine the hypothesis that periodic wildfires and browsing by bettongs were together able to regulate shrub densities in semi-arid rangelands in Australia. Information from various sources concerning the effects of fire, rainfall and browsing on the demography of shrubs was used to construct a model of shrub population dynamics. The model indicates the potential for two states for a given bettong density: first, a low shrub density maintained by a combination of periodic fire and bettong browsing; and second, a high shrub density in the absence of fire. These results have broad implications for pastoral and conservation management in Australian semi-arid rangelands.

Variability in plant fitness influences range expansion of Leptospermum scoparium

A reciprocal transplant experiment was conducted to uncouple genetic and environmental effects on plant fitness of a range‐expanding indigenous shrub, Leptospermum scoparium. Patterns of encroachment differed significantly between three regions of Eucalyptuscamaldulensis woodland in the Grampians National Park, south‐eastern Australia, with greater shrub densities found in the Victoria Valley region. Higher shrub densities in this region are likely due to greater fruit production/shrub, increased field germination and increased early seedling survival. The results of the reciprocal transplant experiment showed increased fitness appears to be due to phenotypic plasticity in response to favourable environmental conditions rather than genotypic variation. Field germination and seedling survival were influenced by region of planting and not seed origin. However, differences in shoot length and biomass of seedlings between the regions were detected after five months of growth in uniform conditions which may indicate genetic differentiation. The results, however, suggest range expansion is partly due to increased plant fitness which is likely a plastic response to favourable environmental conditions, especially soil moisture.

Habitat Structure Associated withBorrelia burgdorferiPrevalence in Small Mammals in New York State

The relationship between habitat structural composition, presence of Ixodes scapularis Say (I. dammini Spielman, Clifford, Piesman, and Corwin), and the prevalence of Borrelia burgdorferi infection in small mammal populations was studied at 12 4-ha study sites selected within two perpendicular transects spanning New York State. Species-adjusted small mammal infection rates (SARs) were calculated to enable comparison of B. burgdorferi infection rates among sites with differing small mammal species composition and were used as the outcome variable in a predictive model. Sites with high SARs were characterized by lower overstory tree canopy height and basal area, increased understory coverage, substantial understory shrub coverage, decreased presence of surface water or saturated soil, high dominance values for I. scapularis, and higher tick burdens on small mammals. These differences were statistically significant from sites with moderate or low SAR values. Understory foliage height profiles were created for each study site, and significant differences in vegetation structural composition between high SAR sites and those with low or moderate SAR were documented. High SAR sites had increased density of herbaceous foliage at 0 and 25 cm, and higher shrub density at 0, 25, and 50 cm measurements above the ground, associated with I. scapularis questing success, and lower densities of sapling trees at 25, 50, and 75 cm. The structural composition of understory vegetation may dictate vector density and predict B. burgdorferi infection rates in small mammals in New York State.

Non‐destructive dry matter estimation of Alhagi sparsifolia vegetation in a desert oasis of Northwest China

Abstract:Question: Can above‐ground biomass of naturally growing Alhagi sparsifolia shrubs be estimated non‐destructively?Location: Qira oasis (37° 01′N, 80° 48′E, 1365 ma.s.l.) at the southern fringe of the Taklamakan desert, Xinjiang, NW China.Methods: Two methods were compared to estimate above‐ground biomass (AGB) of Alhagi. At first shrub AGB was estimated by manual ground measurements (called ‘allometric approach’) of length, width and height of 50 individuals. Subsequently regression equations were established between calculated shrub canopy volume and shrub AGB (r2= 0.96). These equations were used to calculate AGB from manual ground measurements in 20 sample plots within the Alhagi field. Secondly, kite‐based colour aerial photography coupled with the use of a Geographic Information System (called ‘GIS approach’) was tested. First and second order polynomial regressions between AGB data of the 50 individual shrubs and their respective canopy area allowed to automatically calculate the AGB of all remaining shrubs covered by the photograph (r2= 0.92 to 0.96). The use of non‐linear AGB regression equations required an automatised separation of shrubs growing solitary or in clumps. Separation criteria were the size and shape of shrub canopies.Results: The allometric approach was more reliable but also more time‐consuming than the GIS‐based approach. The latter led to an overestimation of Alhagi dry matter in densely vegetated areas. However, this systematic error decreased with increasing size of the surveyed area. Future research in this field should focus on improvements of AGB estimates in areas of high shrub density.

Non-destructive dry matter estimation of Alhagi sparsifolia vegetation in a desert oasis of Northwest China

Question: Can above-ground biomass of naturally growing Alhagi sparsifolia shrubs be estimated non-destructively? Location: Qira oasis (37∞ 01' N, 80∞ 48' E, 1365 m a.s.l.) at the southern fringe of the Taklamakan desert, Xinjiang, NW China. Methods: Two methods were compared to estimate aboveground biomass (AGB) of Alhagi. At first shrub AGB was estimated by manual ground measurements (called ‘allometric approach’) of length, width and height of 50 individuals. Subsequently regression equations were established between calculated shrub canopy volume and shrub AGB (r 2 = 0.96). These equations were used to calculate AGB from manual ground measurements in 20 sample plots within the Alhagi field. Secondly, kite-based colour aerial photography coupled with the use of a Geographic Information System (called ‘GIS approach’) was tested. First and second order polynomial regressions between AGB data of the 50 individual shrubs and their respective canopy area allowed to automatically calculate the AGB of all remaining shrubs covered by the photograph (r 2 = 0.92 to 0.96). The use of non-linear AGB regression equations required an automatised separation of shrubs growing solitary or in clumps. Separation criteria were the size and shape of shrub canopies. Results: The allometric approach was more reliable but also more time-consuming than the GIS-based approach. The latter led to an overestimation of Alhagi dry matter in densely vegetated areas. However, this systematic error decreased with increasing size of the surveyed area. Future research in this field should focus on improvements of AGB estimates in areas of high shrub density.

Negative impacts of a vertebrate predator on insect pollinator visitation and seed output in Chuquiraga oppositifolia, a high Andean shrub.

Studies on plant-pollinator interactions have largely neglected the potential negative effects of the predators of pollinators on seed output, even though anti-predatory behaviour of pollinators may affect visitation patterns, pollen transfer, and therefore potentially, plant reproductive output. We tested the hypothesis that the presence of lizards and insectivorous birds, by reducing pollinator visitation, can have significant negative effects on seed output in the insect-pollinated, genetically self-incompatible lower alpine Andean shrub, Chuquiraga oppositifolia (Asteraceae). The lower alpine belt supports a high density of territorial Liolaemus (Tropiduridae) lizards and low shrubs interspersed among rocks of varying sizes, the latter inhabited by lizards and commonly used by flycatchers Muscisaxicola (Tyrannidae) as perching sites. In a 2x2 factorial predator-exclusion experiment, visitation rates of the most frequent pollinators of C. oppositifolia (the satyrid butterfly Cosmosatyrus chilensis and the syrphid fly Scaeva melanostoma), the duration of pollinator visits, and seed output, were 2-4 times greater when lizards were excluded, while birds had no effect. In a natural experiment, visits by S. melanostoma were 9 times shorter, and pollinator visitation rates of C. chilensis and S. melanostoma, and C. oppositifolia seed output were 2-3 times lower on shrubs growing adjacent to lizard-occupied rocks compared to those growing distant from rocks. Our results, verified for additional Andean sites, suggest that lizard predators can alter the behaviour of pollinators and elicit strong top-down indirect negative effects on seed output. Such effects may be especially important in high alpine plant communities, where pollinator activity can be low and erratic, and pollen limitation has been reported.

The western screech-owl and habitat alteration in Baja California: a gradient from urban and rural landscapes to natural habitat

We studied the western screech-owl (Otus kennicottii) in the desert of the southern Baja California peninsula to determine its status and habitat selection and whether it benefits from moderate human-caused habitat changes. Based on the response to tape-recorded call broadcasts, western screech-owls were more abundant in undisturbed vegetation than in human-altered habitat. In man-made environments, more owls responded in rural than in urban areas; indeed, they were practically absent in urban areas. In natural areas, a total of 1.6 owls/km was estimated in the nonbreeding season and 2.7 owls/km in the breeding season. The most important habitat features selected in natural areas were greater shrub cover in the strata containing shrubs 1–2 and 2–3 m high and greater canopy cover of trees &gt;5 m high. In rural areas adjacent to cultivated areas, owls were present in areas with the greatest cover of shrubs 1–3 m high and a higher density of shrubs. The western screech-owl also occurs at lower densities in rural areas than in natural areas. Towns appear to be avoided in the arid desert of Baja California Sur. Western screech-owls may be still able to persist in rural areas that occupy an intermediate position on the gradient from a functional natural system (desert vegetation) to a completely altered ecological system, urbanized areas.

Vegetation Patterns as Affected by Aspect and Elevation in Small Desert Mountains

Studies of vegetation patterns on small desert mountains have rarely been conduct- ed. The purpose of this study was to evaluate the influence of aspect and elevation on vegetation of Mt. Summerford, a small desert mountain characteristic of the Basin and Range Province of southwestern United States and northern Mexico. Vegetation was randomly sampled on 4 aspects and 3 elevations (0-80 m, 80-160 m, and 160-240 m) of the mountain. South (S) and west (W) aspects had lower grass cover and higher shrub density than east (E) and north (N) aspects. Grass cover decreased as elevation increased on E and N exposures. In contrast, grass cover increased as elevation increased on S and W exposures. A high density of succulent species was found on lower elevations of the S and W aspects. The most frequent species on the mountain were Bouteloua curtipendula, Bouteloua eriopoda, Muhlenbergia porteri, Aristida ternipes, Aloysia wrightii, Ericameria lari- cifolia, Gutierrezia sarothrae, Fouquieria splendens, Dasylirion wheeleri, and Opuntia. Composition of these species, as a mosaic unit, was significantly different among the aspect-elevation units. Species richness was higher in S and W than in E and N aspects. Differences in vegetation patterns were attributed to differential water and nutrient availability on aspects and elevations. This study sug- gests that topographical change is responsible for vegetation patterns in mountains less than 500 m high.

Patterns of shrub-dwelling arthropod diversity across a desert shrubland–grassland ecotone: a test of island biogeographic theory

This study tested the applicability of Island Biogeographic Theory to arthropod assemblages on ‘islands’ of creosotebush (Larrea tridentata) across a desert shrubland-grassland ecotone in central New Mexico, U.S.A. We assumed that herbivorous arthropod taxa that specialize on creosotebush view these shrubs as ‘islands’ of suitable habitat in a ‘sea’ of grassland, and used the Island Biogeographic Theory to examine the observed plant and arthropod patterns. Specifically, we addressed predictions concerning the relationship among arthropod species richness and (1) shrub ‘island’ area, (2) distance between ‘island’ and ‘continent’, and (3) ‘island’ isolation. Using a pyrethrum insecticide, we defaunated 36 shrubs within three ecotonal areas (12 shrubs per area), defined as ‘continental’ (high shrub density), ‘archipelago’ (medium shrub density, close to the ‘continent’), and ‘oceanic’ (low shrub density, far from the ‘continent’). We assessed the initial arthropod distribution and then repeated the defaunation after two weeks to assess immigration patterns. The arthropods from both samples were catalogued and related to ‘island area’ (shrub volume), ‘island’ distance from the defined ‘continent’, and ‘island’ isolation (density). We found that in the initial sample, specialist herbivore species richness was (1) positively related to shrub volume, (2) negatively related to shrub distance from the ‘continent’, but (3) not significantly related to shrub density. Following the recolonization period, we found that specialist herbivore species richness was again significantly positively related to shrub volume, but was not significantly related to shrub distance from the ‘continent’ or to shrub density. We also examined the relationship between generalist predator arthropod (spiders and mantids) populations and shrub ‘island’ characteristics as well as specialist prey abundances on the shrubs. We found that predator abundances on individual shrubs were positively related to shrub volume prior to and following defaunation, negatively related to shrub density both initially and following recolonization, and positively related to the architectural complexity of host shrubs initially. We also found that predator population sizes on shrubs were positively correlated with abundances of specialist prey species. We concluded that our model conformed fully to the species richnes·island area and species richnes·island distance relationships of Island Biogeographic Theory, but only in part to the species richnes·island isolation relationship. We also found that species-specific dispersal ability was a critical factor with regard to the observed species distribution patterns.

Dynamics of shrub encroachment in an African savanna: relative influences of fire, herbivory, rainfall and density dependence

Summary Shrub encroachment has been widely observed in savanna regions. This study analysed the causes of shrub encroachment in the lowveld savanna of north‐eastern Swaziland, southern Africa, and highlighted management regimes that can be used to reduce or prevent it. The rates and dynamics of shrub encroachment were quantified for the period 1947–90 using aerial photographs, and for 1997 using a ground survey. Five similar areas with different land‐use histories were compared to investigate the relative importance of fire, herbivory, rainfall, soil type and shrub density in driving shrub dynamics. In the study area as a whole, shrub cover increased from a mean of 2% in 1947 to 31% in 1990. Dichrostachys cinerea accounted for most of the increase in cover, contributing 81% to total shrub cover during 1997. Shrub cover was strongly correlated with shrub density and weakly negatively correlated with tree cover. Shrub encroachment varied across land‐use fence lines. The key determinants of shrub dynamics were grazing, through its negative effect on fire frequency and an interaction between drought frequency and high shrub cover. Browsing pressure had a significant but minor impact on dynamics, while soil type had no significant effect. High grazing pressures through their effect on fire frequency were critical throughout the study period in promoting shrub encroachment, while the interaction between drought and high shrub cover produced declines in the later stages. Browsing had an impact on encroachment only in the early stages. Frequent fires, facilitated by low grazing pressures, were capable of preventing shrub encroachment. When coupled with drought, frequent fires could reduce high shrub densities. As cover and density were strongly correlated, it can be inferred from the negative correlation between change in cover (density) and initial cover (density) that the rate of shrub encroachment was cover (density)‐dependent and that there was a shrub equilibrium of 40% cover, approximating to 2400 plants ha−1. Shrub population growth was driven by events (fire, drought) as well as continuous agents (density dependence, mean browsing and grazing pressure). Bush encroachment can be reversed by a combination of management (frequent fires) and climatic events (drought). The implications for savanna management are discussed.

Spatio‐temporal colonization patterns of forest plant species in a mixed deciduous forest

Abstract. An integrated analysis of the colonization patterns of forest plant species was carried out in a 34‐ha, mixed deciduous forest in northern Belgium. First, we sought to describe the relationships between land use history and environmental conditions. Land use history and soil type were related and negative correlations between pH and secondary forest age were found. The density of the shrub layer increases with secondary forest age. Litter quantity and cover of Urtica dioica were mainly indirectly influenced by land use history. Litter starts accumulating at low pH values and high shrub density and Urtica dioica grows vigorously on nutrient enriched soils where much light can reach the ground. Next, the importance of these human‐altered environmental conditions for the colonization of forest plant species was assessed relative to the importance of dispersal limitation. Therefore, the distribution of 16 forest species was mapped and species‐specific spatio‐temporal isolation measures were calculated. The analysis revealed that the colonization patterns of the slowly colonizing species (i.e. ‘ancient forest plant species’) are best explained by a combination of spatio‐temporal isolation, soil type, pH and the (non‐)cover of Urtica dioica. By contrast, spatio‐temporal isolation was never a limiting factor for good colonizing forest species. Our results suggest that colonization of ‘ancient forest plant species’ is hampered by a combination of dispersal‐ and recruitment limitation and that the relative importance of both factors is species‐specific.