General considerations for the interpretation of NGC trends

The general purpose of species monitoring is to identify long-term population trends. Trends derived from bags are unusual because the data analysed represent numbers of animals killed rather than counts of live animals. Nevertheless, they have been shown generally to provide a good index of population change where it has been possible to match up bag data with count data (e.g. red grouse: Cattadori et al. 2003; fox: Jarnemo & Liberg 2005). There are potential biases associated with bag data that do not occur with count data, such that underlying trends can be obscured or apparent changes in abundance created where none has occurred. The main potential sources of bias to bear in mind when considering trends in bags are presented below.

Interactions between animal abundance and harvesting effort

Crucially, the number of animals killed depends on the number of animals present and on the effort invested in harvesting them. For game animals, there are annual changes in the number of shooting days per site and number of shooters per day that cause effort to vary over time. This is well illustrated by the lower bag sizes of many species during the two World Wars, when game shooting effort was low. For predatory species, the number of gamekeepers per site, number of traps set, type of trap and duration of trapping will all influence effort and contribute to variation in numbers of animals killed. In practice, it appears that much of this variation adds noise to an underlying trend that reflects species abundance (Whitlock et al. 2003). Comparisons with surveys based on counts of live animals are one way to look for potential bias of this sort.

Legislation often initiates changes in shooting and culling practices. Temporal and regional differences in hunting regulations and their enforcement were reasons for noise in the relationship between counts and bags in Finnish woodland grouse (Ranta et al. 2008). For predators, a number of trapping methods have progressively been outlawed over the last 100 years because they have been judged inhumane. Variation in pest control methods over time, resulting from either changing legislation or improvements in efficiency, influences effort and hence bags over varying timescales. Developments in predator control practice can also affect the seasonality of culling. In recent decades, there has been a shift from culling foxes in spring and summer using snares and terriers to autumn and winter culling by lamping with a rifle. This is likely to affect fox bags because the culling period now coincides with an annual peak in fox density and dispersal before high natural mortality occurs over the winter (Reynolds 2000). Legislation may also change the status of formerly unprotected species, e.g. protection of the wildcat and partial protection of the hedgehog under the 1981 Wildlife and Countryside Act, causing a systematic decrease in numbers recorded on predator sheets.

A more subtle effect stems from potential non-linearity between yield and effort, whereby effort is related to abundance (density-dependence). This occurs if less effort is invested in culling in years of low density, which leads to proportionally smaller bag sizes as abundance falls. A typical example is the suspension of red grouse shooting in years of failed breeding (Hudson 1992). The net result is a steeper decline in bags than in actual density (Hudson 1992, Ranta et al. 2008). Predator bags are less likely to suffer from this problem because predator control tends to be prophylactic, hence independent of predator density in a particular year. However, at high predator density, trap saturation may be possible, especially for single-catch traps. In most cases it is unlikely because of the legislative requirement for daily checking of traps by keepers and typically low trapping rates.

Culling can itself be the cause of changes in species abundance. This poses a problem because the method of data collection then has a direct impact on the quantity it intends to measure. It is impossible to assess the implications of this for population dynamics without prior knowledge of population size and demography. The proportion of a population removed by hunting may be considerable: Hutchings & Harris (1996) estimated that around 60% of brown hares may be removed locally by February culls. Shooting can also alter population age and sex structure, as has been found for red grouse (Bunnefeld et al. 2009). The same may hold for trapping, for example the catchability of mustelids using tunnel traps is dependent on their mass; females, being lighter, may not spring a trap unless it is lightly set (Anon. 1994).

Recording difficulties and representativeness

Two other issues need to be taken into consideration when interpreting trends arising from analyses of NGC bag data. The first is that for some pest species, it is impossible to record the numbers killed over a year accurately because mortality is unobserved and hence unquantifiable. In such cases numbers recorded on NGC forms are either best guesses or the observable fraction (i.e. minimum numbers). This applies especially to brown rat and grey squirrel, for which poisoning is legal.

The second issue is representativeness. The sites contributing records to the NGC do so on a voluntary basis. They cannot be assumed to represent a random sample of shoots across the UK, nor to be typical of the British countryside. Landowners participating in field sports such as fox hunting and gamebird shooting maintain more established woodland and plant more new woodland and hedgerows than non-participants (Oldfield et al. 2003). In the uplands, moors managed for grouse shooting retained considerably more heather than non-managed moors (Robertson et al. 2001). Such management practices mean that estates are likely to be atypical with respect to their numbers of game and non-game species, as they offer good-quality habitat and often cull predators.

Nevertheless, any bias induced by non-randomness is probably reduced by having the same sites contribute records over many years, because records from the same site are directly comparable between years. Toms et al. (1999) considered that using the same sites in consecutive surveys may avoid bias when there is subjectivity in the choice of sites and this subjectivity varies over time. They emphasized that a powerful advantage of historical continuity was in improving the precision of estimates of change by removing between-site varition. The NGC benefits greatly from its historical continuity, which is reinforced by constant attempts to obtain historical records from new participants.


  • Anon. (1994). Predator Control. Game Conservancy Ltd, Fordingbridge.
  • Bunnefeld,N., Baines,D., Newborn,D. & Milner-Gulland,E.J. (2009). Factors affecting unintentional harvesting selectivity in a monomorphic species. J. Anim. Ecol. 78: 485-492.
  • Cattadori,I.M., Haydon,D.T., Thirgood,S.J. & Hudson,P.J. (2003). Are indirect measures of abundance a useful index of population density? The case of red grouse harvesting. Oikos 100: 439-446.
  • Hudson, P.J. (1992). Grouse in Space and Time. The Game Conservancy Ltd, Fordingbridge.
  • Hutchings, M.R.W. & Harris, S. (1996). Current Status of the Brown Hare in Britain. Report to JNCC. University of Bristol, Bristol.
  • Jarnemo,A. & Liberg,O. (2005). Red fox removal and roe deer fawn survival - a 14-year study. Journal of Wildlife Management 69: 1090-1098.
  • Oldfield, T.E.E., Smith, R.J., Harrop, S.R. & Leader-Williams, N. (2003). Field sports and conservation in the United Kingdom. Nature 423: 531-533.
  • Ranta,E., Lindström,E., Lindén,H. & Helle,P. (2008). How reliable are harvesting data for anal;yses of spatio-temporal population dynamics? Oikos 117: 1461-1468.
  • Reynolds, J.C. (2000). Fox Control in the Countryside. The Game Conservancy Trust, Fordingbridge.
  • Robertson, P.A., Park, K.J. & Barton, A.F. (2001). Loss of heather Calluna vulgaris moorland in the Scottish uplands: the role of red grouse Lagopus lagopus scoticus management. Wildl. Biol. 7: 11-16.
  • Toms, M.P., Siriwardena, G.M. & Greenwood, J.J.D.(1999). Developing a Mammal Monitoring Programme for the UK. British Trust for Ornithology, Thetford.
  • Whitlock, R.E., Aebischer, N.J. & Reynolds, J.C. (2003). The National Gamebag Census as a Tool for Monitoring Mammal Abundance in the UK. GCT Research Report to JNCC. The Game Conservancy Trust, Fordingbridge.