Living with yellow-necked mice
By Aidan Marsh and Stephen Harris

Few people will have seen a yellow-necked mouse Apodemus flavicollis; being nocturnal, largely confined to mature woodland and superficially similar to the much more common wood mouse Apodemus sylvaticus, they are easy to overlook. So perhaps it is not too surprising that the yellow-necked mouse was the last small mammal to be found in Britain. It was not recognised as a separate species until 1894. Yet over a century later we still know remarkably little about yellow-necked mice.

So in 1998 The Mammal Society organized a survey and a national trapping programme to establish their exact range and habitat requirements. A substantial number of records were of yellow-necked mice entering houses, somewhat unexpected for a rodent that is reputedly a woodland specialist. So The Mammal Society then undertook a more detailed survey to try to find out why yellow-necked mice and other small mammals enter houses. Here we present the results of that survey and discuss these in the context of yellow-necked mouse ecology.

The Project Background

In Britain, yellow-necked mice are restricted to southern England and Wales, yet even here their distribution is patchy; they are either very rare or absent from Devon and Cornwall, west Wales and central England. This limited geographical range, their comparative rarity even where they do occur, and their elusive nature have led to the perception that yellow-necked mice may be declining. As part of a national survey to establish their status and distribution, The Mammal Society collected yellow-necked mouse records from organisations, academic researchers and other individuals. It was immediately evident that many people believed yellow-necked mice were living in their homes.

It has been known for some time that yellow-necked mice enter houses and outbuildings. However, the magnitude of this trend was not known, nor was the comparative frequency with which yellow-necked mice entered houses compared to other small mammals. To try to understand how often and why yellow-necked mice enter houses, The Mammal Society distributed a questionnaire entitled A Mouse In Your House? during the winter and spring of 1998-1999. This was sent to around 55,000 natural history enthusiasts all over Britain.

Volunteers were asked to record all the small mammals found in their homes. However, one of the difficulties with any study of the yellow-necked mouse is its close similarity to the much more common wood mouse; it is particularly difficult for the casual observer to separate these two species. The main differences in appearance, behaviour and ecology are:

Ecology of yellow-necked mice

We wanted to establish the location of the houses occupied by yellow-necked mice, whether these houses are particularly near to woodland or predominantly rural, when yellow-necked mice enter houses and whether they cause any damage. To appreciate the significance of these questions and interpret the findings we first need to understand the ecology of this species.

Yellow-necked mice are woodland specialists and it is unusual to catch one other than in deciduous or mixed woodland. They are occasionally caught in mature hedgerows, cereal fields, coniferous areas and other habitats. One has recently been caught on a roadside verge during The Mammal Society's project to establish the importance of road verges as habitats for small mammals. However, such records are generally rare, and usually refer to individual mice rather than substantial resident populations.

The general impression is that yellow-necked mice prefer mature woodland, especially those with a good diversity of seed and fruit producing trees and shrubs. This may mean that yellow-necked mice are particularly vulnerable to woodland loss and fragmentation, and this may be a key factor accounting for their perceived rarity. However, we still do not know the impact of woodland fragmentation on the number and/or distribution of yellow-necked mice, although The Mammal Society's national trapping study suggested that the most isolated woodlands were less likely to contain populations of yellow-necked mice, supporting the idea that woodland fragmentation may limit their abundance.

Yellow-necked mice may be dependent on older woodland because they have a more specialized diet than wood mice, but surprisingly we do not know for certain what they eat. Yellow-necked mice are thought to specialize on tree seeds and to be reluctant to switch to other types of food, such as insects. In contrast, wood mice display a more catholic diet, readily switching to new food sources as necessary. Our recent investigations into the diets of yellow-necked and wood mice living together in the same woodland support these ideas, but suggest a greater degree of interspecific differentiation in their diet than is found in Europe. Given this apparent reliance on tree seed, it is even more surprising that yellow-necked mice appear to enter houses so readily.

Mice in houses

In total, 1796 responses were received from households around the country. This represents an approximate questionnaire return rate of 3.3 percent. The distribution of the 1755 households for which a location was supplied is shown in Figure 1a. In total, 1240 people reported mice in their home at some point during the last five years, mainly wood or house mice (Figures 1b, c). Only 181 people reported having yellow-necked mice in their homes. The distribution of records (Figure 1d) closely reflects the national distribution of the yellow-necked mouse; any unusual records were followed up to try to establish their authenticity. Since it is easy to confuse wood and yellow-necked mice, it is possible that a few misidentifications may have occurred. However, these will not affect the overall results.

Of those households with mice, 29.5% of people reported that more than one species of small mammal was present. The number of reports for the six most commonly reported species is shown in Figure 2; bank voles Clethrionomys glareolus, common shrews Sorex araneus and pygmy shrews Sorex minutus were only reported occasionally. While snap-traps remain the favoured means of control, live-capture traps are now widely used (Figure 3), although since the questionnaires were mainly sent to natural history enthusiasts, the use of non-destructive control methods may be higher than average.

What type of houses do mice prefer?

We asked people whether their homes were detached, semi-detached, terraced or a flat, and whether they were built pre 1800, 1800-1900, 1900-1945, 1945-1970, or post-1970. The first analyses were to determine whether house type influenced the presence or absence of the three commonest household invaders, the wood, house and yellow-necked mouse.

There was a fairly even response from householders living in properties dating from pre-1800 through to those living in modern post-1970s houses. The age of the house did not influence the presence or absence of mice. However, the results of a log-linear analysis showed that house type (detached, semi-detached or terraced) did have a significant effect on both house and wood mice. Wood mice were less common in terraced houses than detached or semi-detached houses, while house mice were more common in terraced houses. Yellow-necked mice showed a similar trend to wood mice but this was not significant, possibly due to the smaller sample size. Given that terraced housing tends to be more common in urban areas, it is perhaps understandable that house mice were more common visitors to these properties than wood mice.

The effect of house location

We also asked people to estimate whether the distances to the nearest countryside and nearest woodland were less than 100 m, 100-500 m, 500-1500 m or more than 1500 m. In addition, we asked for the size and composition of the nearest woodland, i.e. was it largely deciduous, mixed or largely coniferous. The majority of returns were from people living in relatively rural settings, with 1020 reporting that they were within 100 m of countryside. However, there were sufficient households at a greater distance to show that the further a house was from countryside, the lower the likelihood that wood or yellow-necked mice would be present (Figure 4a). There was a significant difference in the distance to the nearest woodland between the houses that did and did not contain wood mice (p<0.001). The same was true for yellow-necked mice (p<0.001). House mice were more likely to be found in houses further from the countryside, but this trend was not significant (Figure 4b). Wood (p<0.001) and yellow-necked mice (p<0.001) were less likely, while house mice (p<0.01) more likely, to be present in houses further away from the nearest woodland.

So, the distance to the nearest woodland was important to all three species of mice, although the size of the woodland did not appear to matter. Woodland type (deciduous, mixed or coniferous) was included in a log linear analysis but was not found to be a significant factor in explaining the presence or absence of any of the species of mouse. However, proportionately fewer people in houses near coniferous woodland reported yellow-necked mice (6.5%) compared to people in houses near mixed (13.0%) or deciduous woodland (15.8%). While these differences were not significant, there was a clear trend indicating a preference for deciduous woodland. Overall, these results support the belief that yellow-necked mice are closely linked to deciduous woodland but at the same time this makes why they enter houses all the more intriguing.

Winter retreats for yellow-necked mice?

One of the key aims of The Mammal Society's survey was to clarify exactly when yellow-necked mice enter houses and whether any pattern existed nationally. The vast majority of people reported that autumn and winter were the key periods when yellow-necked mice entered their homes, with few cases during the summer (Figure 5). So the appearance of yellow-necked mice was highly seasonal. Once in the house, yellow-necked mice were most commonly found in the roof space (41%), although reports from the ground floor were also quite common (30%) (Figure 6a). They were also frequently found in other parts of the property, including outbuildings (47%) (Figure 6b).

An obvious explanation for this winter invasion could be that yellow-necked mice retreat to houses for warmth during the colder months of the year. However, this is unlikely to be the case, since we know that yellow-necked mice tolerate very cold winters in Scandinavia and eastern Europe. A second possibility is that yellow-necked mice are escaping the wet conditions associated with this time of year. It has been suggested that yellow-necked mice avoid areas with wet soils (Montgomery 1978) and its European range appears to encompass areas that experience cold, but not necessarily wet, winters. However, the survey showed that these mice enter homes throughout their range in Britain, yet we know that areas such as East Anglia experience a much drier, more continental climate, than other parts of the country. Since there was no geographical trend in house occupancy in Britain, we believe that shelter from the wet may not be the crucial factor either.

Perhaps the answer lies in some of the other statistics from the survey. Although 18% of householders reported that yellow-necked were not responsible for any damage to their house, furnishings or contents, 82% thought otherwise (Figure 6c). In particular, 31% thought that yellow-necked mice were responsible for spoiling food. Maybe yellow-necked mice enter homes in search for food, or maybe they simply utilize the food they find once they are inside.

Yet this explanation is equally unconvincing. If yellow-necked mice enter houses in search of food, why should this occur more in the autumn and winter, when their natural diet of tree seeds is most abundant, than in late spring and summer when this food source is scarce? Furthermore, some people reported that yellow-necked mice go to considerable lengths to collect and store hazelnuts, acorns or other seeds inside their houses. This also suggests that they are not attracted to our homes for food. We clearly need more detailed studies before we can provide a satisfactory answer to this enigma.

Conclusions

It is clear that yellow-necked mice enter houses throughout their range in Britain, largely during the colder months of the year, and that this behaviour does not appear to be affected by the presence of either wood or house mice. This survey supports our current understanding of the ecology of all three species and confirms our view of the yellow-necked mouse as a rural, woodland species. Based on the results of the national trapping study, we now believe that populations of the yellow-necked mouse in Britain are stable; in some places they may even be common. Nonetheless, across the wider countryside the wood mouse is by far the most abundant mouse and appears to enter houses frequently.

In an earlier study John Flowerdew said the yellow-necked mouse was renowned for entering houses, although in this survey there was no indication that they showed any greater tendency to enter houses than wood mice. However, confirmation that the yellow-necked mouse is indeed a regular visitor to houses throughout its range suggests that it may be more adaptable to new environments than previously thought.

So what exactly do yellow-necked mice gain by entering houses? The Mammal Society's survey has shown that yellow-necked mice in Britain are largely confined to areas with the warmest summers. This preference for hot summers means that yellow-necked mice are one species that may increase in numbers and/or range in Britain with the current trend for hotter summers. At first we thought that the occupation of houses over winter might reflect some aspect of these climatic requirements. However, so far the evidence for this is not convincing and perhaps yellow-necked mice are simply exploring novel environments as they disperse from their natal woods.

When small mammal densities peak in the autumn, competition occurs for space and other resources, leading to widespread dispersal. The simplest explanation for the occurrence of both wood and yellow-necked mice in houses is that dispersing animals are investigating new habitats.

The results do clarify a number of misconceptions. Anecdotal evidence had suggested that, where present, yellow-necked mice tended to be found in the roof space of houses. While this was the favoured location, the majority of householders actually reported them from other parts of the house. Equally, the general impression was that yellow-necked mice were less destructive occupants of houses than other species, rarely damaging food, products or furnishings. So it was a surprise to discover that over four-fifths of respondents held yellow-necked mice accountable for damage within the home.

This survey provided a limited snap shot of one aspect of yellow-necked mouse behaviour. It has shown that house occupancy by yellow-necked mice occurs throughout their range and that the proximity of a house to countryside and woodland are two important factors that may determine the probability of a winter invasion by this species. Although more common than we suspected, houses invaded by yellow-necked mice are still very much in the minority. Perhaps the realisation that the mouse in your house might well be a yellow-necked mouse will encourage further studies of this little known British mammal.

Further information

People interested in obtaining the full report on The Mammal Society's yellow-necked mouse trapping study (price £2.00 to include post and packing), or helping with further mammal studies, should write to: Georgette Shearer, The Mammal Society, 15 Cloisters House, 8 Battersea Park Road, London SW8 4BG.

Acknowledgements

The Mammal Society is grateful to English Nature for a grant to support the distribution of this questionnaire and to everyone who took part. Thanks also to Annemarie Greenwood and Kate Pennell for assistance with data entry.