Journal of Applied Ecology

In this paper we provide a theoretical explanation for the observed dependence of kappa on prevalence, and introduce into ecology an alternative measure of accuracy, the true skill statistic (TSS), which corrects for this dependence while still keeping all the advantages of kappa. We also compare the responses of kappa and TSS to prevalence using empirical data, by modelling distribution patterns of 128 species of woody plant in Israel.

The theoretical analysis shows that kappa responds in a unimodal fashion to variation in prevalence and that the level of prevalence that maximizes kappa depends on the ratio between sensitivity (the proportion of correctly predicted presences) and specificity (the proportion of correctly predicted absences). In contrast, TSS is independent of prevalence.

When the two measures of accuracy were compared using empirical data, kappa showed a unimodal response to prevalence, in agreement with the theoretical analysis. TSS showed a decreasing linear response to prevalence, a result we interpret as reflecting true ecological phenomena rather than a statistical artefact. This interpretation is supported by the fact that a similar pattern was found for the area under the ROC curve, a measure known to be independent of prevalence.

Synthesis and applications. Our results provide theoretical and empirical evidence that kappa, one of the most widely used measures of model performance in ecology, has serious limitations that make it unsuitable for such applications. The alternative we suggest, TSS, compensates for the shortcomings of kappa while keeping all of its advantages. We therefore recommend the TSS as a simple and intuitive measure for the performance of species distribution models when predictions are expressed as presence–absence maps.

Income growth is a primary driver of globalization and a clear association exists between Gross Domestic Product and the richness of alien floras and faunas for many regions of the world. In many cases, the exposure of these economies to trade is highlighted by the significant role of merchandise imports in biological invasions, especially for island ecosystems.

Post‐1950, technical and logistic improvements have accelerated the ease with which commodities are transported across the globe and hindered the traceability of goods and the ease of intercepting pests. New sea, land and air links in international trade and human transport have established novel pathways for the spread of alien species. Increasingly, the science advances underpinning invasive species management must move at the speed of commerce.

Increasing transport networks and demand for commodities have led to pathway risk assessments becoming the frontline in the prevention of biological invasions. The diverse routes of introduction arising from contaminant, stowaway, corridor and unaided pathways, in both aquatic and terrestrial biomes are complex. Nevertheless, common features enable comparable approaches to risk assessment. By bringing together spatial data on climate suitability, habitat availability and points of entry, as well a demographic models that include species dispersal (both natural and human‐mediated) and measures of propagule pressure, it is possible to generate risk maps highlighting potential invasion hotspots that can inform prevention strategies.

Synthesis and applications. To date, most attempts to model pathways have focused on describing the likelihood of invader establishment. Few have modelled explicit management strategies such as optimal detection and inspection strategies and assessments of the effectiveness of different management measures. A future focus in these areas will ensure research informs response.

There have been widespread declines in the populations of many groups of organisms associated with farmland in Britain and north‐west Europe. The declines have been particularly marked amongst habitat specialists; many of the taxa still common on farmland are habitat generalists.

Farming practices have become increasingly intensive in the post‐war period, with a dramatic reduction in landscape diversity. Since 1945, there has been a 65% decline in the number of farms, a 77% decline in farm labour and an almost fourfold increase in yield. Farms have become more specialized; the greatly increased use of machinery has made operations quicker and more efficient, but has resulted in the removal of 50% of the hedgerow stock. Autumn sowing of crops has become predominant, with winter stubbles now far less prevalent. The number and extent of chemical applications has increased greatly, but the net amount applied, and their persistence, has decreased in recent years.

Intensification has had a wide range of impacts on biodiversity, but data for many taxa are too scarce to permit a detailed assessment of the factors involved. Reduction in habitat diversity was important in the 1950s and 1960s; reduction in habitat quality is probably more important now.

As a case study, the declines in populations of seed‐eating birds populations were assessed in relation to changing agricultural management. Generally, the declines were likely to be caused by a reduced food supply in the non‐breeding season, although other factors may be important for particular species.

Agriculture will face a number of challenges in the medium term. While research into the mechanisms underlying species and habitat associations, and their interaction with scale, will be critical in under‐pinning management, consideration of farmer attitudes and socio‐economic factors is likely to be as important. Biodiversity may benefit from integrated farming techniques but these need to incorporate environmental objectives explicitly, rather than as a fringe benefit.

Seventy studies that quantify the ecological effects of harvesting NTFP from plant species were reviewed, with the aims of assessing the current state of knowledge and drawing lessons that can provide guidelines for management as well as better directing future ecological research in this area.

The case studies illustrated that NTFP harvest can affect ecological processes at many levels, from individual and population to community and ecosystem. However, the majority of research was focused at a population level and on a limited subset of plant parts that are harvested.

Tolerance to harvest varies according to life history and the part of plant that is harvested. Moreover, the effects of harvest for any one species are mediated by variation in environmental conditions over space and time, and by human management practices.

In order to withstand heavy harvest, specific management practices in addition to gathering are necessary for many NTFP species. Management practices can be carried out at different spatial scales and some are highly effective in fostering population persistence.

Synthesis and applications. Substantial advances have been made towards identifying the ecological impacts of NTFP harvest. However, there is a need for longer‐term studies that focus on multiple ecological levels (ranging from genes to ecosystems), that assess the mechanisms underlying impacts and that validate current models. Researchers and forest managers need to work with local harvesters in designing and evaluating management practices that can mitigate the negative effects of harvest.

In Britain, the management of grassland has changed substantially in the second half of the 20th century. A high proportion of lowland grassland is managed intensively. The major changes include a doubling in the use of inorganic nitrogen, a switch from hay to silage, and increased stocking densities, particularly of sheep. Structurally diverse and species‐rich swards have been largely replaced by relatively dense, fast‐growing and structurally uniform swards, dominated by competitive species.

Most of these changes have reduced the suitability of grassland as feeding and breeding habitat for birds.

The most important direct effects have been deterioration of the sward as nesting and wintering habitat, and loss of seed resources as food. Short uniform swards afford poor shelter and camouflage from predators, whereas increased mowing intensities and trampling by stock will destroy nests and young. Increased frequency of sward defoliation reduces flowering and seed set, and hence food availability for seed‐eating birds.

The indirect effects of intensification of management on birds relate largely to changes in the abundance and availability of invertebrate prey. The effects of management vary with its type, timing and intensity, and with invertebrate ecology and phenology, but, in general, the abundance and diversity of invertebrates declines with reductions in sward diversity and structural complexity.

Low input livestock systems are likely to be central to any future management strategies designed to maintain and restore the ecological diversity of semi‐natural lowland grasslands. Low additions of organic fertilizer benefit some invertebrate prey species, and moderate levels of grazing encourage sward heterogeneity.

There is now a need to improve understanding of how grassland management affects bird population dynamics. Particularly important areas of research include: (i) the interaction between changes in food abundance, due to changes in fertilizer inputs, and food accessibility, due to changes in sward structure; (ii) the interaction between predation rates and management‐related changes in habitat; and (iii) the impact of alternative anti‐helminithic treatments for livestock on invertebrates and birds.