Harvest mouse image courtesy of Reg Mckenna CC BY -SA 2.0 generic license. Originally posted to Flickr.
Although predominantly associated with agricultural habitats (Love et al., 2000), the harvest mouse is also frequently found in reed beds, and in undisturbed areas of rough grassland such as road verges (Harris, 1979b; Dickman, 1986). The species is now more commonly found in boundary features such as hedgerows, field margins and ditches, than within the cropped areas of fields (Harris et al., 1995; Moore et al., 2003). Summer foraging and nesting largely take place in the stem-zone of tall vegetation, whereas boundary features and shorter vegetation are used after harvest.
It is probable that the harvest mouse has been adversely affected by changes to agricultural practices, such as the switch to shorter-stemmed cereal varieties (Harris, 1979b), and the transition to winter cereals that are cut before the breeding season (Harris, 1979a). However, it is difficult to quantify the scale of any impacts, not only because of a lack of baseline data, but also because there are large seasonal and annual fluctuations in population size.
Status -native .
IUCN Red List (GB: NT; England: [LC]; Scotland: n/a; Wales: [VU]; Global: LC). • This species has not been assessed for Article 17 of the EU Habitats Directive.
Image courtesy of Chris Barber. CC BY-SA 2.0 Generic license originally posted to Flickr.
Harris et al. (1995) estimated the population size of harvest mice to be 1,425,000, with 1,415,000 in England and 10,000 in Wales. These estimates, however, were not adjusted to take into account the smaller distribution of harvest mice compared to wood mice. Rather, the population size in Britain was calculated by dividing the population size of wood mice by the ratio of wood mice to harvest mice, and proportions of this value were assigned to countries post hoc. This is likely to have overestimated the total population size, as wood mice are present in a larger area of Britain than harvest mice. Reassessment of the data from Harris et al. (1995) using the method presented here (i.e. (wood mouse population}. size/current wood mouse distribution area/ratio of wood mice to harvest mice) * area of harvest mouse distribution)), suggests a total population size in Britain of 793,000, although this method assumes that the species’ range has remained constant since 1995.
The reassessed population size from 1995 falls within the confidence limits of the current estimate. Further surveys are therefore suggested to improve the precision of population size estimates and to allow for an assessment of trends.
Information about harvest mouse density is very sparse, and no other reports of temporal trends are currently available.
Habitat declining. Range -stable. Population decreasing
Gaps in the species’ distribution in England are likely to represent areas lacking survey effort, rather than true absences. Surveys are therefore recommended in these areas.
Very few population density estimates exist for the harvest mouse, particularly for the prebreeding period. Harris et al. (1995) calculated population size on the basis of a ratio of 26.6 wood mice to one harvest mouse. Owing to a lack of pre-breeding population density estimates for harvest mice, the same approach was used in the current review. Additional data are now available that allow this ratio used by Harris et al. (1995) to be updated (Bellamy et al., 2000; Love et al., 2000; Moore et al., 2003; Woods et al., 2003; Askew et al., 2007; see Clapham, 2011).
Wood mouse population sizes per country (and their upper and lower 95% confidence limits) were divided by the geographical range size (extent of occupancy) to produce density estimates (ha-1) for the entire area over which the harvest mouse could potentially be found. These values were converted into harvest mouse densities by dividing by the mean ratio of wood mice:harvest mice. These density estimates were then multiplied by the geographical range size (extent of occupancy) to produce the population estimates.
Eleven papers and one government report were identified by the literature search. One paper contained pre-breeding estimates of population density (Clapham, 2011), and five reported post-breeding estimates and distribution data, or examined the relationship between habitat characteristics and harvest mouse presence. Five publications contained the ratio of wood mice to harvest mice; they are summarised by Clapham (2011) along with the ratios included by Harris et al. (1995.
Harvest mouse population estimates are extremely difficult to make with any level of certainty. Only one pre-breeding population density estimate has been published since 1995 (Clapham, 2011), with very few estimates prior to this. Populations are thought to have a clumped distribution (Harris, 1979b), with large seasonal fluctuations (Gosling and Baker, 2008). Evidence of harvest mouse presence (nests) is easily overlooked in surveys, and the species is difficult to trap in spring and summer as it is rarely found at ground level at that time of year (Harris et al., 1995). Consequently, any direct estimate of population size would be subject to considerable error.
Our estimate is based on the mean ratio of wood mice to harvest mice. Yet the primary habitats for wood mice do not necessarily correspond with those for harvest mice (the former being highly dependent on woodland and the latter on long grass). Our estimation of harvest mouse numbers makes the assumption that the ratios shown in Table 7.10a are representative across the geographical range. No account is taken of the differing areas of each habitat. Therefore, rarer habitats may be over-represented in the mean ratio. Harvest mice also occur in some habitats, such as fenland, for which no ratios are available. The ratios in many habitats are likely to be unreliable because data are sparse (e.g., for road verges); even so, information for all habitats has been weighted equally
• IUCN Red List (GB: LC; England: [LC]; Scotland: [LC]; Wales: [LC]; Global: LC). • This species has not been assessed for Article 17 of the EU Habitats Directive.
The wood mouse is highly adaptable and is found in most habitats, including woodland, arable land, rough grassland, heather, blanket bog, sand dunes, urban areas and hedgerows (Kotzageorgis and Mason, 1997; Marsh and Harris, 2000; Flowerdew and Ellwood, 2001; Tattersall et al., 2001). Population densities in woodland vary with successional stage: mid-level regeneration of 5- to 10-year-old vegetation supports a higher density of wood mice than either ungrazed fields or 10-year-old regeneration (Marsh and Harris, 2000).
Hedgerows, including those distant from woodlands, are an important habitat for the wood mouse, and can support resident populations (Gelling et al., 2007). Population densities in hedgerows are high after arable crops are harvested, and also when grass swards are short because of grazing or cutting (Tew and Macdonald, 1993; Garratt et al., 2012). At these times, the wood mouse, like other small mammals, makes preferential use of boundary features rather than in-field areas (Tattersall et al., 2001). Similarly, hedgerows provide cover and food sources during autumn and winter (Kotzageorgis and Mason, 1997; Liu et al., 2013).
Population estimates from Harris et al. (1995) were derived from densities reported in the literature and by experts. The habitat classes were equivalent to those used in the current review. Population size was estimated as 38,000,000 in total, with 19,500,000 in England, 15,000,000 in Scotland and 3,500,000 in Wales. These figures are all within our confidence limits, so there is no evidence of a significant change in population size since 1995.
Nationally, there are changes between the two reviews in the estimated availability of key habitats (arable land, broadleaved woodland, coniferous woodland and improved grassland), generated by a combination of true change and methodological differences, irrespective of any range change (see Sections 2.3 and 32.3 for further details). The adjusting of results to reflect more probable temporal changes in the composition of the British landscape — using differences between the 1990 and 2007 Countryside Surveys (Carey et al., 2008) — produces a population size that differs from the original estimate by only 5%, and that falls within the confidence limits of the original. It is therefore concluded that methodological differences have no material impact on the comparisons between the two time periods.
Population,range and habitat all stable.
Wood mouse populations fluctuate seasonally, with peaks in autumn and early winter and troughs in spring and summer (Flowerdew, 1985; Harris et al., 1995). Density estimates were therefore derived only from studies which took place between March and June. As wood mice in arable land primarily use field margins, with incorporation of crop fields into home ranges only before the harvest (Tattersall et al., 2001), we assume that the in-field population will be included within the estimate for hedgerows. Arable land is therefore excluded from the analysis.
No recent density estimates were available for fen, marsh and swamp habitat. Had the estimates from Harris et al. (1995) for this habitat been included in the present review, confidence intervals would not have been calculable as none were provided in the original paper. In addition, the exclusion of fen, marsh and swamp altered the population estimate by <100,000 (<1%). This habitat class was therefore excluded.
Eighteen relevant papers were identified by the literature search. Of these, five provided an estimate of pre-breeding population density, four provided a relative measure of abundance (captures per trap night), and one contained percentage occupancy for hedgerows (Gelling et al., 2007). The remainder provided post-breeding density estimates, explored the relationship of habitat variables to abundance but gave no effect size, or provided descriptive data only.
No percentage occupancy data were available for most habitats; the population size is therefore overestimated. The population estimate is largely derived from broadleaved (28%) and coniferous (21%) woodland , where population density estimates are supported by five (n=311) and four (n=49) references, respectively. Given the abundance of evidence relative to most other species in this review, sensitivity analyses were not conducted for these habitats.
Improved grassland forms 49% of the habitat within the geographical range of the wood mouse, yet because of low population density (0.1ha-1), it contributes just 2% of the estimated population. The density used for improved grassland is the median value from estimates provided by three expert opinions: given the extent of this habitat, field validation of the values would considerably improve the precision of the population estimate. A reliability assessment is shown in Table 7.11c. For density estimates based on expert opinion, a conservative score of 1 has been applied to the ‘location of study sites’ section
Image courtesy of James Lindsey at Ecology of Commanster CC BY-SA 3.0 license.
The yellow-necked mouse is found primarily in mature broadleaved woodlands (Marsh et al. 2008), particularly ancient coppiced woodlands, where they favour older established compartments, rather than recent coppice (Gurnell et al., 1992; Capizzi and Luiselli, 1996). Hedgerows also provide an important habitat for the species in Britain, with telemetry studies showing that individuals can reside solely within a linear hedgerow habitat (Montgomery, 1978). The availability and diversity of tree seeds is an important predictor of yellow-necked mouse density (Marsh et al., 2001). Britain is at the western edge of the species’ European range, possibly owing to the impact of low summer temperatures on tree seed abundance. There is some potential for misidentification of this species with the wood mouse, particularly at the edges of its geographical range where abundance may be relatively low.
Conservation status- • IUCN Red List (GB: LC; England: [LC]; Scotland: n/a; Wales: [LC]; Global: LC). • This species has not been assessed for Article 17 of the EU Habitats Directive.
Harris et al. (1995) estimated the total population size for Britain as 750,000, with 662,500 in England and 87,500 in Wales. The British estimate was based on an estimate of 450,000 for ancient woodlands, with the remaining 300,000 added to allow for yellow-necked mice in other habitats. The current review uses broadleaved woodland (adjusted for 55% occupancy), rather than ancient woodland.
Nationally, there are changes between the two reviews in the estimated availability of key habitats (arable land, broadleaved woodland, coniferous woodland and improved grassland), generated by a combination of true change and methodological differences, irrespective of any range chang. The adjusting of results to reflect more probable temporal changes in the composition of the British landscape — using differences between the 1990 and 2007 Countryside Surveys (Carey et al., 2008) — produces a population size that differs from the original estimate by only 5%, and that falls within the confidence limits of the original. It is therefore concluded that methodological differences have no material impact on the comparisons between the two time periods
Trends for the future. Population and range -potential increase. Habitat stable.
The percentage of occupied sites for broadleaved woodland was taken from Marsh et al. (2001) where 80 of 146 sites (55%) within the species’ range were occupied. Only non-zero population density estimates were included to avoid accounting for occupancy twice. The total area of broadleaved woodland within the species’ distribution was multiplied by the percentage of occupied sites. The occupied area of broadleaved woodland was then used for all subsequent calculations of population size. In the absence of occupancy data for coniferous woodlands and urban areas, the same value (55%) was applied to these habitats. For hedgerows, percentage occupancy was taken from Gelling et al. (2007), where 180 hedgerows on 12 dairy farms in 4 geographical areas within the species’ range were surveyed. Occupancy (75%) was provided for one of these areas, but in the other areas very few yellow-necked mice were captured. To provide a more representative value for the species throughout its range, an average value of percentage occupancy was taken across all four areas, where the number of sites surveyed across all areas was assumed to be equal, and percentage occupancy at the remaining three sites was assumed to be roughly zero. As two of the sites were located towards the edge of the species’ range where densities might be expected to be lower, the resulting value may be a slight underestimate.
Six papers contained useful information for yellow-necked mice. Three of these reported prebreeding population density (Montgomery, 1980; Kotzageorgis and Mason, 1997; Marsh and Harris, 2000), one gave occupancy for hedgerows (Gelling et al., 2007), and one gave occupancy for broadleaved woodland (Marsh et al., 2001). The remaining papers reported only post-breeding density estimates (Marsh et al., 2001; Moro and Gadal, 2007) applied for broadleaved woodland are based on 25 separate estimates from two papers, although the confidence limits are wide (0.57-11ha-1), reflecting highly variable densities across the geographical range. Further surveys, specifically designed to include a representative sample, would improve confidence in the estimate.
Hedgerows contribute 21% of the estimated population, and there are 200,000km available within the species’ range (Table 7.12a, Figure 7.12b). The abundance estimate for hedgerows is derived from 6 estimates from one paper, with the resulting median density estimate being substantially higher than that in broadleaved woodland. Re-calculation of population size following stepwise removal and replacement of the individual density estimates for hedgerows did not result in a significant change in population size. Reliability scores are shown in Table 7.12c. For the purposes of this assessment, we consider the population density estimates of experts to be representative of a restricted area of the species’ range.
The house mouse lives commensally with humans: its movement patterns and current widespread distribution are attributed to this relationship (Searle et al., 2009). Although listed as native to Britain by IUCN, the best available evidence suggests the species arrived in western Europe in the Bronze age and is recorded in Britain by the Iron age (Yalden, 1999). The species is therefore not subject to many of the environmental factors that regulate the population sizes of most other small mammals. It is, however, sensitive to human activities, including the alteration of buildings and the deployment of rodenticide (Pocock et al., 2004). The decline in urban infestations in the 1970s is likely to be the result of increased rodenticide efficacy (Richards, 1989; Harris et al., 1995). Detailed analyses of the habitat preferences of the house mouse are, however, lacking.
Status. non native -naturalised.
IUCN Red List (GB: LC; England: n/a; Scotland: n/a; Wales: n/a; Global: LC). • This species has not been assessed for Article 17 of the EU Habitats Directive
Harris et al. (1995) reported a total population size of 5,192,000, comprising 4,535,000 in England, 657,000 in Scotland and 206,000 in Wales. Most of the data used for the current review were the same as those used by Harris et al. (1995), with the following exceptions: Harris et al. (1995) included density estimates for arable and pastoral habitats (based on Montgomery and Dowie (1993) and Rowe et al. (1983)), but these were not included in the current estimate; conversely, evidence on population density in farm buildings (Pocock et al., 2004) was available for inclusion in the current review. The house mouse primarily occupies urban and rural dwellings, so these differences are unlikely to affect the population size estimate significantly. Although the area of urban land in the current analysis differs by 45% compared to the area quoted in Harris et al. (1995), the population of house mice was calculated from the number of dwellings, rather than the total area, so this difference does not affect the conclusions. It was not possible to calculate confidence limits for the current population size estimates, but they differ from those in Harris et al. (1995) by less than 1%.
Population,habitat and range all stable.
Although house mice are found in habitats such as field margins and woodland, they are poor competitors with other rodents, particularly wood mice (Berry and Tricker, 1969; Tattersall et al., 1997). In one study in pastoral farmland in Ireland, although 10% of the rodents captured in field margins in summer were house mice, no house mice were captured in field margins in the spring (Montgomery and Dowie, 1993). No population density estimates were identified in the literature review for habitats other than buildings, so the population estimate is based on buildings only.
The number of farm holdings per country was taken from the Agriculture category in the UK 2015 key statistics dataset (Office for National Statistics). The number of dwellings per country was derived from the 2014-2015 dwelling stock reports from the English, Scottish and Welsh governments. The numbers of dwellings considered ‘urban’ and ‘rural’ were calculated using percentage of residences that are classed as urban and rural from the 2011 census analysis (Office for National Statistics, 2013). Rural residences were divided into farms (farm holdings per country) and other rural dwellings (hereafter ‘rural’) by subtracting the number of farms from the total rural dwellings.
For the population of house mice in farm buildings, it was assumed that, on average, each farm contains the same number of buildings as the study farms in Pocock et al. (2004). The total population was calculated as ‘number of occupied dwellings * population size per farm’. For urban and rural buildings, data based on surveys by Rennison and Drummond (1984) were taken from Harris et al. (1995); the number of mice per infestation (4.5) was multiplied by the number of occupied dwellings, where 3.8% of urban and 5.6% of rural buildings were reported as occupied. There is evidence that population size on farms does not vary by season (Pocock et al., 2004), so results from all seasons were included in the analysis.
Two papers were identified by the literature search, both of which contained population sizes in farm buildings. One paper (Quy et al., 2009) was excluded as populations were artificially maintained and so did not represent natural population size. House mouse populations exhibit boom-and-bust fluctuations, depending largely on resource availability and rodenticide use. It is therefore difficult to make precise population estimates (Harris et al., 1995). The assessment was based on house mouse density in buildings. The values used for farm buildings were taken from a single paper that studied two adjacent farms (Pocock et al., 2004), and the overall value per holding was multiplied by the number of farm dwellings. Adjustments for occupancy were made on the assumption that the proportion of farm buildings occupied by house mice was the same as for rural houses generally (for which some data were available). The figures are therefore likely to provide a reasonable estimate of the numbers of animals across farm buildings of all types. However, the extent to which these farms are typical of those found nationally is unclear. The house mouse uses habitats other than buildings, such as field margins and woodland. However, no density or occupancy estimates were available for these habitats, so the population size is underestimated.
There are pages on this site relating to the lifestyle,habits and breeding of the Wood mouse and the Harvest mouse. Click on the relevant content banners above and scroll down to read.