Formation of nutrient‐poor soil patches in a semi‐arid woodland by the European rabbit (Oryctolagus cuniculus L.)

Abstract European rabbits (Oryctolagus cuniculus L.) occur over large areas of eastern Australia, where they disturb significant quantities of soil while constructing warrens. Although the effects of rabbits on vegetation are generally well understood, little is known about their effects on soil physical or chemical properties. We studied the effect of rabbit warrens on soil chemistry in a semi‐arid woodland supporting a high density of rabbit warrens. Within one large warren, we examined nutrient concentrations within three microsites (mound, inter‐mound and an intermediate disturbed area) representing a gradient of increasing rabbit disturbance, and supported this with a study of soil nutrients with depth within 1‐m‐deep trenches through five warrens. Landscape‐level changes in surface chemistry were also examined at an additional 23 warrens. At both patch and landscape scales, pH, electrical conductivity, soluble and exchangeable Ca²⁺ and K⁺, and total Al and Ca generally increased with increasing rabbit disturbance, while total C and S (LECO), total P and S (ICP), and soluble Na⁺ declined, and total N remained unchanged. Although chemical changes with depth were generally ill defined, surface soils tended to be more similar in their composition than deeper soils. Overall, our results reinforced the view that rabbits have a negative effect on surface soils in semi‐arid woodlands, and suggested that restoration of the original woodland vegetation may be hampered by changes in soil biogeochemistry associated with the warrens.     

黄土高原典型草原群落结构和土壤水分对划破的响应

划破是草原改良的基础措施之一,划破强度是划破措施的关键环节,划破对草原健康持续管理有重要意义。目前的研究主要集中在划破对植物群落结构和生产力的影响上,然而草原植物群落与土壤水分对划破强度的响应尚不清楚。在黄土高原典型草原开展不同程度的草地划破试验,探究不同划破强度（27.4%、46.3%和61.9%）对草地植物群落物种多样性、生物量和土壤水分的影响。结果表明：3个划破强度下划破带物种丰富度显著低于未划破带1-3种/m²,划破带和未划破带群落相似性分别低于整区23.85%-119.23%和44.43%-84.55%。地上生物量随物种丰富度的增加而增大,且地上生物量与Simpson指数和Shannon Weiner指数显著负相关。3个划破强度下未划破带地下生物量和总生物量分别高于划破带88.2%-134.6%、52.4%-67.8%、2.5%-16.6%和103.9%-152.9%、59.3%-75.8%、9.1%-22.6%。植物群落物种丰富度和地上生物量随划破强度的增加呈"驼峰"型曲线变化,当划破强度分别为43.7%-55.3%和43.8%-45.7%时,植物群落物种丰富度和生物量均最高。本试验阐明了划破对典型草原植物群落特征和土壤水分的作用机制,研究结果为采用划破措施实现草地培育和草原修复提供了科学依据,对保护草地生物多样性和提高生产力具有重要意义。     

Habitat use for warren building by European rabbits (Oryctolagus cuniculus) in relation to landscape structure in a sand dune system

Several conservation efforts are being made to recover European rabbit populations (Oryctolagus cuniculus) on the Iberian Peninsula. Some of them focus on burrow management; others involve building different types of warren. A few studies have examined site selection for warren building, and these studies have considered only warren placement within sites and not the broader area surrounding these locations. The objective of this study was to evaluate how landscape pattern determines habitat selection by rabbits for warren building at different spatial scales. Landscape, home range scale, and microhabitat were the spatial scales used in this study. Warrens were not uniformly distributed over the study area but, rather were concentrated in areas with a high abundance and cover of Retama monosperma and high vegetation cover. Rabbits preferred digging warrens in areas with low fragmentation and where patches are few, large, and contiguous. Based on our results, we suggest that a study of landscape structure should be carried out before design habitat management, recovery or translocation programs. Such studies will need to take into account the physiognomy and size, shape, and continuity of patches in fragmented landscapes. Rabbit conservation programs must address areas that provide not only the maximum potential rate of intake, but also good soil and vegetation cover conditions for warren building and suitable surrounding areas.     

Contribution of the Western Australian wheatbelt termite,Drepanotermes tamminensis (Hill), to the soil nutrient budget

The role of soil modification by the mound-building termite,Drepanotermes tamminensis (Hill), was studied during 1991 in the Durokoppin Nature Reserve, Western Australia. Soil chemical parameters were quantified for ‘soils’ in nests and for surrounding soil in both a Wandoo (Eucalyptus capillosa) woodland and a Casuarina (Allocasuarina campestris) shrubland plot. All ‘soils’ in nests were more acidic than the surrounding soil within each study plot. Generally, nutrient levels in the nested soils were higher than the un-nested soil within each study plot and were also higher in the woodland than in the shrubland plot. Depending on the nuttient concerned, the nested soil contained between 0.3 and 21.9% of the total nutrient load per hectare within each study plot. The quantities of nutrients per hectare in termite mounds were higher in the woodland than in the shrubland plot. It is concluded that mounds of this species of termite form a significant bank of nutrients, although time for release of such nutrients depends on the degree of erosion and on the longevity of mounds.     

Soil Ripping and Herbicides Enhance Tree and Shrub Restoration on Stripmines

I evaluated responses by 16 native woody species to differential soil compaction and density of ground cover. The trees and shrubs studied represent sites in southern Illinois that commonly have restrictions to root growth from soil or drainage conditions. The study site was a restored surface coal mine in southern Illinois with a rooting medium compacted by grading and a dense ground cover of pasture species. Soil compaction was alleviated in half the study area before tree planting by mechanically ripping the soil to a depth of 1.2 m. Roots of half the trees and shrubs were dipped in a Terra® slurry before planting, and the ground cover around all planting spots was afterwards sprayed with herbicide. In year 2 after planting the ground cover in half of the unripped and half of the ripped area was further controlled by repeated application of herbicides. Ripping significantly increased height growth of all trees combined and all species individually in each year of the study. Second-year control of ground cover increased height growth of all trees combined and of seven species individually. Some species were damaged by herbicides. Terra® had little evident effect on species performance. Animal damage reduced early survival and growth, especially of Acer (maple) and Cornus (dogwood) species, and later growth of Quercus rubra (red oak). Removal of ground cover with herbicides tended to increase deer browse. Soil ripping, herbicide application, and choosing tree species unattractive to deer can be recommended to increase success in planting trees for forest restoration.     

Environmental Heterogeneity, Animal Disturbances, Microsite Characteristics, and Seedling Establishment in a Quercus havardii Community

In restoration experiments it is imperative to consider the study of mechanisms of how species are maintained and preserved in a system. This paper reports on the results of a field experiment examining the growth and survival of seedlings of Schizachyrium scoparium, a dominant perennial bunchgrass member of the Quercus havardii (sand shinnery oak) communities of semiarid western Texas, on mounds of displaced soil produced by Sylvilagus auduboni (rabbit). The central question posed is: does environmental heterogeneity created by small mammals influence seedling survival and growth? The specific questions addressed are: (1) Does seedling survival, growth, and nutrient uptake vary when grown on mounds, off-mound soils, and artificially created mounds?; (2) What is the influence of the microbial and litter components of mound soils on seedling survival?; and (3) In communities where animal disturbances create environmental heterogeneity and may impact seedling establishment, is it possible to artificially create mounds that could serve a similar function? Results show that characteristics of mound soils increase seedling survival, shoot and root biomass, root length, number of tillers, mycorrhizal infection, and nutrient uptake more in plants grown on mounds than off mounds. Both the microbial and litter components of mound soils are essential components of this effect. Artificial mounds generated from soils associated with the herbaceous community were more similar to intact rabbit mounds than artificial mounds generated from soils associated with the oaks. The results indicate that rabbits produce rich patches (both nutrient and microbial) favorable to the growth of seedlings of the dominant bunchgrass, and point to the potential importance of rabbit disturbances in shaping the dynamics of this plant community. Thus, rabbit-generated disturbances produce environmental heterogeneity in the sand shinnery oak community, similar to that produced by harvester ants in this community as shown by an earlier study. Small animal activity that results in soil displacement, and influences soil characteristics, may indirectly contribute to the persistence of certain plant species within a community. Thus, disturbances may well operate in semiarid communities to produce nutrient and microbe rich microsites which may function to maintain diversity.     

Effect of three mound-building ant species on the formation of soil seed bank in mountain grassland

The effect of three ant species (Lasius flavus, Formica spp., Tetramorium caespitum) on soil seed bank formation was studied in temperate mountain grassland. Seed removal experiments, analysis of soil seed content and seed survival experiments were carried out to evaluate the influence of ground ants on the seed fate. In the seed removal experiment seeds of 16 species, including 5 species with elaiosome-bearing seeds (myrmecochores), were exposed and their removal followed for 39 h. On average, ants removed 63.8% of myrmecochorous seeds and 10.9% of seeds without adaptation to ant dispersal. Analysis of soil seed content revealed that myrmecochores, in spite of expectations that they would accumulate in nests of seed dispersing ants, were most abundant in the soil of control plots. Evidence on seed relocation to the ant nests was obtained from a comparison of mounds of seed dispersing and seed non-dispersing ant species, as more seeds were found in the mounds of Formica spp. and Tetramorium caespitum (seed dispersers) in comparison with the mounds of Lasius favus (non-disperser).The soil seed bank of the compared microhabitats (control plots and mounds of 3 ant species) differed in their species composition, seed abundance and vertical distribution. The most distinct qualitative differences were between seed flora of control plots and mounds of Tetramorium caespitum. Control plots had approximately 30,000 propagules per m², which was double the number of seeds found in the ant mounds. In control plots, abundance and diversity of seeds steeply declined with depth; this trend was not observed in the mounds probably due to bioturbation. In the seed survival experiment, more seeds (2 out of 3 species) survived in control plots, which may also contribute to the higher seed abundance in this microhabitat.This study showed that seed relocation by ants does not contribute significantly to seed bank build-up at this study site. Ants may, however, increase the regeneration success of myrmecochores, mainly by dispersal for distance and placement in a larger spectrum of microsites, in contrast to species not adapted for myrmecochory.     

Effects of ants on plant diversity in semi-natural grasslands

Ants (L. niger and L. flavus) build conspicuous mounds that are covered with vegetation. The aim of this study was to investigate whether the vegetation on ant mounds in semi-natural grasslands differed from that around the mounds. Another aim was to investigate whether the changes in the vegetation on ant mounds were influenced by grazing management or by habitat characteristics, semi-dry versus moist. Here, the total number of plant species and total plant cover were lower on ant mounds than in patches off-mound. The plant cover of perennials that form rosettes was twice as high on mounds inhabited by L. niger than on those inhabited by L. flavus. Only a few plant species were restricted to either ant mounds or adjacent field and the effects of ants on the plant diversity in semi-natural grasslands seemed to be low. Grazing management did not affect the differences in the vegetation on ant mounds and in equal-sized patches off-mound, whereas habitat characteristics affected ant-induced changes in vegetation cover of some plant species.     

Response of riparian plant communities to distance from surface water in the Okavango Delta,Botswana

The objective of this study was to determine the influence of distance from surface water on riparian woodland communities in the Okavango Delta. Vegetation sampling was conducted in seven sites within the Okavango Delta in 20 m × 10 m belted plots placed perpendicular to the river bank. The plots were placed at 0–10 m, 10–20 m, 20–30 m, 30–40 m and 40–50 m distance classes increasing away from the river bank. Tree height, basal area, species richness, canopy cover and diversity were determined for each distance class. Indicator species analysis was used to determine the characteristic species at each distance class. Single‐factor ANOVA and Tukey post hoc analysis were used to compare species diversity, mean tree height, cover and basal area between distance classes. Correlation between distance from surface water and vegetation parameters was sought using Spearman regression analysis. All parameters except for species richness varied significantly (P < 0.05) along distance from surface water. Distance from surface water was positively correlated all vegetation parameters except for mean species richness/plot. These results show that distance from surface water influences riparian plant community composition and distribution in the Okavango Delta. This implies that riparian plant species can be indicators of long‐term hydrologic conditions in the Delta.     

Successional trends and biomass of mosses on windthrow mounds in the temperate rainforests of Southeast Alaska

We investigated successional trends on windthrow mounds in two old-growth Tsuga heterophylla-Picea sitchensis forests in northern southeast Alaska to determine the influence of windthrow disturbance on the maintenance of plant diversity. We were particularly interested in assessing the value of mosses in detecting long-term effects of disturbance in temperate rainforests. Mosses established a dense carpet on windthrow mounds within the first few decades after the disturbance. No consistent changes were noted in total moss and vascular plant cover, moss biomass, or species diversity between young mounds (±50 yrs), intermediate mounds (±150 yrs) or old mounds (> 200 yrs), or between mounds and the undisturbed forest floor, despite consistent differences in soils development.Classification and ordination of the vegetation data did not show a consistent relationship between soil surface age or soil depth and overall species composition on the two sites. Young mounds were the most compositionally distinctive, primarily due to moss species. Pogonatum alpinum var. sylvaticum, P. contortum and Polytrichum formosum were generally confined to young mounds with unstable substrata, while Dicranum majus and Sphagnum girgensohnii were associated with old soil surfaces and deep organic soils. Vascular plant species with affinities for riparian or deep shade habitats (Tiarella trifoliata, Coptis asplenifolia and Dryopteris expansa) showed a general preference for the forest floor. Gymnocarpium dryopteris was the only vascular plant with a significant association with young mounds.Mosses comprised approximately 25% of understory plant biomass and as much as 50% of understory productivity. In cool temperate forests, the inclusion of mosses in vegetation analysis may provide valuable insights into the nature of vegetation patterns over subtle environmental gradients. The distinctiveness of the temperate rainforest type and the unique ecological effects of windthrow disturbance in this type are also suggested by this study.Abbreviations HR = Heintzleman Ridge (study site) - OP = Outer Point (study site)     

Burrowing by translocated boodie (Bettongia lesueur) populations alters soils but has limited effects on vegetation

Digging and burrowing mammals modify soil resources, creating shelter for other animals and influencing vegetation and soil biota. The use of conservation translocations to reinstate the ecosystem functions of digging and burrowing mammals is becoming more common. However, in an increasingly altered world, the roles of translocated populations, and their importance for other species, may be different. Boodies (Bettongia lesueur), a commonly translocated species in Australia, construct extensive warrens, but how their warrens affect soil properties and vegetation communities is unknown. We investigated soil properties, vegetation communities, and novel ecosystem elements (specifically non‐native flora and fauna) on boodie warrens at three translocation sites widely distributed across the species’ former range. We found that soil moisture and most soil nutrients were higher, and soil compaction was lower, on warrens in all sites and habitat types. In contrast, there were few substantial changes to vegetation species richness, cover, composition, or productivity. In one habitat type, the cover of shrubs less than 1 m tall was greater on warrens than control plots. At the two sites where non‐native plants were present, their cover was greater, and they were more commonly found on boodie warrens compared to control plots. Fourteen species of native mammals and reptiles were recorded using the warrens, but, where they occurred, the scat of the non‐native rabbit (Oryctolagus cuniculus) was also more abundant on the warrens. Together, our results suggest that translocated boodie populations may be benefiting both native and non‐native flora and fauna. Translocated boodies, through the construction of their warrens, substantially alter the sites where they are released, but this does not always reflect their historic ecosystem roles.     

The distribution and density of rabbit warrens on the Southern Tablelands of New South Wales

Rabbit warrens are distributed in a clumped manner in three areas totalling 1696 ha in the Southern Tablelands of New South Wales. While the discontinuous distribution may be partly related to original patterns of colonization, the present distribution is closely related to agricultural practices and the rabbits’biological requirements. The highest densities of warrens are in the agriculturally less valuable areas. The number of burrows in use in a warren is related to the number of rabbits living in the warren, although at the time of the study, large warrens contained fewer rabbits than expected.     

Impact of Formica exsecta Nyl. on seed bank and vegetation patterns in a subalpine grassland ecosystem

The mound building ant Formica exsecta Nyl. is widely distributed in grassland ecosystems of the Central European Alps. We studied the impact of these ants on seed bank and vegetation patterns in a 11 ha subalpine grassland, where we counted over 700 active ant mounds. The mounds showed a distinct spatial distribution with most of them being located in tall‐grass, which was rarely visited by ungulates (red deer; Cervus elaphus L.). Heavily grazed short‐grass, in contrast, seemed to be completely avoided by ants as only few mounds were found in this vegetation type. The species composition of the ant mound and grassland seed banks was quite similar, i.e. from 15 common plant species 12 were found in both seed bank types. We found the same proportions of myrmecochorous seeds in ant mound and grassland soil samples. In contrast, the number of seeds was 15 times higher in mound compared with the grassland soil samples. Also, the vegetation growing on ant mounds significantly differed from the vegetation outside the mounds: graminoids dominated on ant mounds, herbaceous and myrmecochorous species in the grassland vegetation. We found significant continuous changes in vegetation composition on gradients from the ant mound centre to 1 m away from the mound edge. Overall, F. exsecta was found to have a considerable impact on seed bank and vegetation patterns in the grassland ecosystem studied. These insects not only altered grassland characteristics in the close surrounding of their mounds, but also seem to affect the entire ecosystem including, for example, the spatial use of the grassland by red deer.     

Rabbits and the failure of regeneration in Australian arid zone Acacia

Three experiments are reported concerning the effect of rabbit grazing (Oryctolagus cuniculus L.) upon the recruitment of Acacia seedlings to populations in the South Australian arid zone. In western myall woodland (Acacia papyrocaqsa Benth.) under prevailing rabbit and sheep densities, seedlings exposed to grazing by these two herbivores or to rabbits alone were severely pruned, whereas totally protected seedlings grew unchecked. Seedlings of four Acacia species; A. papyrocarpa, A. oswaldii (F. Muell.), A. kempeana (F. Muell.) and A. burkittii (F. Muell. ex Benth.) were transplanted into four 50 m × 50 m rabbit-proof enclosures. Six rabbits were introduced into each enclosure and within 24 h half of the total seedling population had been grazed. This was at a seedling dry weight ratio of 1|150 000 of the total fodder on offer. In one of the enclosures no seedlings were eaten and there is evidence to suggest that a dense patch of grass had a buffering effect, reducing grazing pressure. Small shoots cut from old Acacia and transferred to the ground throughout 1000 ha of western myall woodland were grazed rapidly near rabbit warrens and progressively less rapidly with increased distance from warrens. The experiments demonstrated that even with the lowered post-myxomatosis population densities, rabbit grazing pressure would significantly affect recruitment in arid zone Acacia populations in the absence of stock.     

Role of termites in the restoration of soils and plant richness on bowé in West Africa

Bowé (hardened ferricrete soils formed by erosion, drought or deforestation) are often associated with termite mounds, but little is known about these mounds and their role in the restoration of soils and plant biodiversity on bowé. This study examined termite mounds on bowé and their effects on soil depth and plant richness. Sixty-four sampling plots were laid out randomly on bowé sites with mounds and on adjacent bowé sites without mounds. The height and circumference of each mound were measured. Species inventories were made and soil depth measured in each plot. Linear mixed effects and generalised mixed effects models with Poisson error distribution were used to assess the variation in soil depth and plant species richness in mound and nonmound microsites. Two types of mounds (small vs. large) associated with different termite species were observed on bowé, with the small mounds being most common. Plots with either large or small mounds had deeper soils and higher plant richness than the adjacent plots without mounds. Conservation of termite mounds is important for restoring soils and plant richness on bowé, and termite mounds should be taken into consideration in biodiversity and soil management strategies for bowé.     

Influence of vegetation and soil types on the wheatbelt termite,Drepanotermes tamminensis (Hill), in the Western Australian wheatbelt

A survey of the distribution and density of mounds of the harvester termite,Drepanotermes tamminensis (Hill), was carried out in the Durokoppin Nature Reserve, Western Australia in 1990. Vegetation and, to a lesser extent, soil type, appear to be important factors in determining density and distribution of termite mounds within the Reserve. A more detailed study of mounds in Wandoo (Eucalyptus capillosa) woodland and Casuarina (Allocasuarina campestris) shrubland indicated that the total number and size of mounds were significantly higher in the woodland than in the shrubland. The total wet weight biomass ofD. tamminensis was calculated as 3.74 gm⁻² (37.4 kg ha⁻¹) in the woodland and 1.69 gm⁻² (16.9 kg ha⁻¹) in the shrubland. Thus, of the two favored habitats, Wandoo woodland appears to be more optimal for this termite species than the Casuarina shrubland.     

Responses of naïve and experienced European rabbits to predator odour

The response of prey species to predator scent has been investigated in many mammalian species; however, there is little information about the responses of European wild rabbits at the population level. Therefore, we conducted a simple experiment to investigate the behavioural response of a rabbit population to native predator cues in the wild. We compared the response to the scent of a predator (red fox) in a wild rabbit population bred in semi-natural conditions and naïve to terrestrial predators with the response of a population in a similar environment where terrestrial predators were present. The response to predators was based on rabbit abundance, inferred from pellet counts and measured by the defecation rate per day (DRD). Our results indicate that rabbits responded to the odour of fox faeces in the treatment warrens, resulting in a lower DRD. The main anti-predator behaviour observed was spatial avoidance (warren abandonment), which seemed to be more accentuated for rabbits who had not previously had contact with foxes in the plot where terrestrial predators were excluded. In both the fenced and the unfenced plot, the differences in the effect of the predator odour between the control and treatment warrens disappeared after cessation of treatment, suggesting a flexible and adaptive behaviour of rabbits to predator cues.

     

Burrowing activities of kangaroo rats and patterns in plant species dominance at a shortgrass steppe-desert grassland ecotone

Abstract. Our objective was to evaluate the effects of burrowing activities by banner-tail kangaroo rats (Dipodomys spectabilis Merriam) on plant community structure and species dominance for two patch types at the ecotone between shortgrass steppe and desert grassland in New Mexico, USA. 10 mounds produced by kangaroo rats were selected in patches dominated by Bouteloua gracilis (the dominant in shortgrass steppe communities) and 10 mounds were selected in patches dominated by B. eriopoda (the dominant in Chihuahuan desert grasslands). Plant cover and density by species were sampled from three locations associated with each mound: the mound proper, the edge of the mound in the transition area, and the off-mound vegetation. Similar cover of B. eriopoda for the edges of mounds in both patch types indicates the ability of this species to respond to animal disturbances regardless of the amount of cover in the surrounding undisturbed vegetation. By contrast, cover of B. gracilis was low for all mounds and mound edges in patches dominated by this species. Much higher cover of B. eriopoda on mound edges compared to the undisturbed vegetation in B. gracilis-dominated patches indicates that kangaroo rats have important positive effects on this species. Lower cover of perennial grasses and higher cover of forbs, shrubs, and succulents on the edges of mounds in B. eriopoda-dominated patches compared to patches dominated by B. gracilis indicate the importance of surrounding vegetation to plant responses on disturbed areas. Our results show that kangaroo rats have important effects on both species dominance and composition for different patch types, and may provide a mechanism for small-scale dominance patterns at an ecotone; thus providing further support for their role as keystone species in desert grasslands.     

EFFECT OF THE NON-MYCORRHIZAL PIONEER PLANT SALSOLA KALI L. (CHENOPODIACEAE) ON VESICULAR-ARBUSCULAR MYCORRHIZAL (VAM) FUNGI

The effects of Salsola kali (Russian thistle) on chlamydospore numbers and viable propagules of the vesicular-arbuscular mycorrhizal fungus Glomus fasciculatum were monitored for 18 wk under laboratory conditions. When S. kali or Agropyron smithii (a mycorrhizal species) were grown for 6 wk, the number of chlamydospores in soil declined to the same number as in the control (non-planted) soil; the number of chlamydospores in the S. kali and control soils continued to decline over the 18 wk period. However, using a Zea mays bioassay, the per cent colonization of corn roots obtained from soil in which S. kali grew for 6, 12, or 18 wk was significantly greater than in the control soil. These results suggest that S. kali may stabilize mycorrhizal infectivity of disturbed soils for at least short periods of time.     

Woody plant encroachment reduces species richness of herb‐rich woodlands in southern Australia

Abstract Woody plants have been increasing in many woodland and savanna ecosystems owing to land use changes in recent decades. We examined the effects of encroachment by the indigenous shrub Leptospermum scoparium (Myrtaceae) on herb‐rich Eucalyptus camaldulensis woodlands in southern Australia. Species richness and compositional patterns were examined under the canopy of L. scoparium and in surrounding open areas to determine the species most susceptible to structural changes. Richness was significantly lower in areas of moderate to high L. scoparium cover (>15%), suggesting that a threshold shrub cover caused major change in this ecosystem. Shrubs were associated with a significant reduction in above‐ground biomass of the ground‐layer flora and a significant shift in community composition. The few species that were positively associated with high L. scoparium cover were also common in the woodland flora; no new species were recorded under the shrub canopy. Important environmental changes associated with L. scoparium cover were decreased light availability and increased litter cover, which were likely a consequence of encroachment. Leptospermum scoparium cover was also associated with greater surface soil moisture, which may be a consequence of increased shading under the shrub canopy or indicate favourable soil conditions for L. scoparium establishment. Reductions in species richness and abundance of the germinable seed bank were found in soil samples taken from under L. scoparium. With ongoing recruitment of L. scoparium and consequent increases in shrub cover, ground‐layer diversity in these species‐rich woodlands should continue to decline over time.