More-than-human histories and the failure of grand state schemes: sylviculture in the New Forest, England

As James Scott’s Seeing Like a State attests, forests played a central role in the rise of the modern state, specifically as test spaces for evolving methods of managing state resources at a distance, and as the location for grand state schemes. Together, such ambitions necessitated both the elimination of local understandings of forest management – to be replaced by centrally controlled scientific precision – and a narrowing of state vision. Forests thus began to be conflated with trees (and their timber) alone. All other aspects of the forest, both human and non-human, were ignored. Through the lens of the 18th and early 19th century New Forest in southern England, this paper examines the impact of government attempts to shift the focus of state forests from being remnant medieval hunting spaces to spaces of income generation through the creation of vast sylvicultural plantations. This state scheme not only reworked the relationship between the metropole and the provinces – something effected through systematic surveys and novel bureaucratic procedures – but also dramatically impacted upon the biophysical and cultural geographies of the forest. By equating forest space with trees alone, the British state failed to legislate for the actions of both local commoners and non-human others in resisting their schemes. Indeed, subsequent oppositions proved not only the tenacity of commoners in protecting their livelihoods but also the destructive power of non-human actants, specifically rabbits and mice. The paper concludes that grand state schemes necessarily fail due to their own internal illogic: the narrowing of state vision creates blind spots in which human and non-human lives assert their own visions.

The response of small mammals to natural and human-altered edges associated with Afromontane forests of South Africa

An increase in edge area reduces the effective size of habitat fragments and thus the area available for habitat-interior specialists. However, it is unclear how edge effects compare at different ecotones in the same system. We investigated the response of a small mammal community associated with Afromontane forests to edge effects at three different habitat transitions: natural forest to grassland (natural edge, structurally different vegetation types), natural forest to mature plantation (human-altered edge, structurally similar vegetation types) and natural forest to harvested plantation (human-altered edge, structurally different vegetation types). We predicted that edge effects should be less severe at natural ecotones and at similarly structured contiguous vegetation types than human-altered ecotones and differently structured contiguous vegetation types, respectively. We found that forest species seemed to avoid all habitat edges in our study area. Surprisingly, natural edges supported a less diverse small mammal community than human-altered forest edges. However, edge effects were observed deeper into native forests surrounded by mature alien plantations (and more so at harvested plantations) than into native forests surrounded by native grasslands. The net effect of mature plantations was therefore to reduce the functional size of the natural forest by creating a larger edge. We suggest that when plantations are established a buffer zone of natural vegetation be left between natural forests and newly established plantations to mitigate the negative effects of plantation forestry.

Habitat associations of the red squirrel Sciurus vulgaris and grey squirrel S-carolinensis in Northern Ireland

A survey of red and grey squirrel habitat associations in Northern Ireland was conducted between September 1994 and August 1995. Two hundred and sixty-one sites were visited and a list of habitat characteristics for each site was noted. Multiple discriminant function analysis of the habitat type was employed to group squirrel occurrence, while contingency analysis examined independence of habitat type and squirrel species presence. Habitat associations differed between the two species. One-way ANOVAs of habitat data suggested that sites occupied by red squirrels only were predominantly coniferous, at higher altitude and latitude and much larger in area than sites occupied by grey squirrels only, which were mostly deciduous. When both species were sympatric, sites were more likely to be coniferous and larger in area than sites occupied by either species. Grey squirrels were less frequent than expected in upland plantations and more frequent than expected in parkland and gardens; the opposite was true for red squirrels. The mean distance between sites with only red squirrels and the nearest site with grey squirrels was greater than the mean distance between sites with only grey squirrels and the nearest site with red squirrels. An approach to conserving the red squirrel in view of the continued expansion in the grey squirrel's distribution in Ireland is discussed.

INTRAPOPULATION VARIATION IN THE DIET OF THE WOOD MOUSE APODEMUS-SYLVATICUS

Variation in the diet of wood mice, Apodemus sylvaticus, was investigated in two contrasting habitats, deciduous and coniferous woodland, over 26 months and in 10 additional sites trapped during winter. Stomach contents were categorized as seed, fruit, green plant, root and animal material. Diet was evaluated using the percentage occurrence of each food type. Age and sex differences in diet occurred infrequently. Seed predominated throughout but was especially prevalent in autumn and winter. There was a peak in the incidence of animal material in the spring and early summer. Animal food was generally more frequent in mice caught in conifer plantations than in deciduous woodland during the longer-term study. Further, mice from the additional coniferous habitats had greater percentage occurrence of animal food than those from the additional deciduous sites. There was a negative, non-linear association between relative population size and diet in these winter samples. This suggests that spatial variation in numbers of A. sylvaticus is dictated by food availability and density is locally food limited.

Invasive alien species disrupt spatial and temporal ecology and threaten extinction in an insular, small mammal community

The impact of invasive bank vole (Myodes glareolus) and greater white-toothed shrew (Crocidura russula) on indigenous Irish small mammals, varies with season and habitat. We caught bank voles in deciduous woodland, young coniferous plantations and open habitats such as rank grass. The greater white-toothed shrew was absent from deciduous woods and plantations but did use open habitats with low level cover in addition to field margins. Numbers of both invasive species in field margins during summer were higher than in the previous spring. The indigenous wood mouse (Apodemus sylvaticus) and pygmy shrew (Sorex minutus), differed in degrees of negative response to invasive species. Wood mice with bank voles in hedgerows had reduced recruitment and lower peak abundance. This effect was less extreme where both invasive species were present. Wood mice numbers along field margins and open habitats were significantly depressed by the presence of the bank vole with no such effect in deciduous woodland or coniferous plantations. Summer recruitment in pygmy shrews was reduced in hedgerows with bank voles. Where greater white-toothed shrew was present, the pygmy shrew was entirely absent from field margins. Species replacement due to invasive small mammals is occurring in their major habitat i.e. field margins and open habitats where there is good ground cover. Pygmy shrew will probably disappear from these habitats throughout Ireland. Wood mice and possibly pygmy shrew may survive in deciduous woodland and conifer plantations. Mitigation of impacts of invasive species should include expansion of woodland in which native species can survive.

A decision support system for managing irrigation in agriculture

In this paper, an automatic Smart Irrigation Decision Support System, SIDSS, is proposed to manage irrigation in agriculture. Our system estimates the weekly irrigations needs of a plantation, on the basis of both soil measurements and climatic variables gathered by several autonomous nodes deployed in field. This enables a closed loop control scheme to adapt the decision support system to local perturbations and estimation errors. Two machine learning techniques, PLSR and ANFIS, are proposed as reasoning engine of our SIDSS. Our approach is validated on three commercial plantations of citrus trees located in the South-East of Spain. Performance is tested against decisions taken by a human expert.