CHAPTER 2 GREY PARTRIDGE
Nicholas J. Aebischer
The Game Conservancy Trust (GCT), Fordingbridge, Hampshire, SP6 1EF
2.1 Summary
Brood production rate and female winter survival rate of the Grey Partridge are dependent on density whereas chick survival rate is not. Density dependence is bound up with the availability of nesting cover.
The annual rate of change in the Grey Partridge population over the last decade, calculated from longterm partridge demographic data from Sussex, was very similar to the value of 0.926 obtained from BTO national CBC data.
Based on data from the GCT Conservation Headland research programme, Grey Partridge chick survival rate was recalibrated in terms of the percentage of arable area made up of insectrich broodrearing habitat. Chick survival rate increased by 0.04 for every 1% increase in percentage arable area that is insectrich.
Ignoring density dependence in the partridge model leaves unrecognised the role of nesting cover as an important determinant of Grey Partridge density in the agricultural environment.
Stabilising the population at its current level requires an increase in average amount of nesting cover from 4 to 4.3 km/km^{2}, or management of 3% of arable area as insectrich broodrearing habitat.
Recovery of the population to its 1996 level depends on achieving 6.5 km/km^{2} of nesting cover together with 3% of arable area as insectrich broodrearing habitat. Setaside could be managed towards this end under existing prescriptions.
2.2 Introduction
The Grey Partridge Perdix perdix is one of the 20 species incorporated in the UK Government’s farmland bird headline qualityoflife indicator. Because it is a quarry species and therefore of considerable economic importance (at least until its decline), it has been the subject of over 70 years of research. As a result, it is the moststudied of the 20 farmland bird species, and the only one for which detailed information on density dependence is available (Potts 1980, 1986). This report reviews the densitydependent relationships underpinning a full population model of Grey Partridge population dynamics, compares model predictions with ones obtained in the absence of density dependence, and examines how changes in the agricultural environment affect Grey Partridge population density.
This study draws heavily on the detailed information published by Dr G.R. Potts in the course of his longterm study of Grey Partridge population dynamics in a 62km^{2} area of the South Downs, Sussex (e.g. Potts 1986). This study began in 1968 and is ongoing. Every year after harvest, partridges have been counted, aged as adults or young, and sexed (adults only). Using the formulae in Potts (1986), these counts enable the following demographic rates to be estimated:
 brood production rate (number of spring pairs that produce chicks)
 nest survival rate (probability that a female that survives the summer will produce chicks)
 female summer survival rate (proportion of females present in the spring alive in the autumn)
 chick survival rate (proportion of chicks that survive from hatching to six weeks of age)
female winter survival rate (proportion of females present in the autumn alive the following spring – includes immigration and emigration)
On the basis of these definitions, brood production rate is equal to the product of female summer survival rate and nest survival rate. Using key factor analysis (Varley & Gradwell 1963), Potts (1980, 1986) found that brood production rate and female winter survival rate varied with partridge density, whereas chick survival rate did not. He discovered that the relationship between brood production rate and density changed depending on whether predators were controlled or not, and that density dependence was bound up with the availability of nesting cover. As predator control is now much less widely practised than in the past, and as the focus here is on how Grey Partridge density responds to changes in the agricultural environment, this paper restricts itself to the situation without predator control. The densitydependent relationships reported by Potts (1980, 1986) are as follows, noting that k denotes a kfactor, i.e. log_{10}(survival rate):
Brood production rate:
k_{1} + k_{2} = 0.22 + 0.58 log_{10}(C pairs / km cover per km^{2}) (1)
where k_{1} is the kfactor for nest survival, k_{2} is the kfactor for female summer survival and C is the mean clutch size (14 eggs).
Female summer survival rate:
k_{2} = 0.31 (k_{1} + k_{2}) (2)
Female winter survival rate:
k_{5} = 0.07 + 0.39 log_{10}((autumn females / km cover per km^{2}) + 1) (3)
Taken together with shooting losses, which are assumed to occur (if they occur) before winter losses, and which have also been shown to be densitydependent, these relationships have been brought together in a model of partridge population dynamics (Potts 1980, 1986) that forms the basis for the current work. The original model took as input observed annual chick survival rates, but these were replaced here by a constant average rate so that effects of varying the average rate could be examined.
Typically, little is known about density dependence in the population dynamics of the other 19 species that make up the farmland bird qualityoflife indicator, so it is instructive to compare results from the Grey Partridge densitydependent model with ones from a Grey Partridge model that ignores density dependence.
Demographic parameters for such a model were estimated from the annual Sussex partridge counts for the last decade (19902000), which was thought to be representative of the current agricultural environment of the Grey Partridge. Demographic parameters was calculated on an annual basis, then averaged across years to give a mean and standard deviation. For comparability with BTO demographic parameters for other species, adult survival and firstyear survival rates were calculated as the product of female summer survival rate and female winter survival rate, fledglings per attempt was calculated as the product of average clutch size (14 eggs) and nest survival rate, and juvenile survival rate was equated to chick survival rate (Table 1).
Based on these parameters, the annual rate of population change was 0.929, which is very close to the value of 0.926 estimated from BTO national CBC data for the Grey Partridge over the same period. This is equivalent to a decline of 77% over 20 years.
The joint effect of varying average female annual survival rate and annual young production per surviving female (fledglings per attempt x juvenile survival) on the annual rate of change is displayed in Figure 1. The limits of the axes were chosen to reflect approximately 95% of observed variation in annual means. The direction of the contours suggests that the rate of change is equally sensitive to changes in adult annual survival and production.

Adult survival

Firstyear survival

Juvenile survival

Postfledging survival

Fledglings per attempt

Number
of attempts

Mean

0.411

0.411

0.294



8.57

1

SD

0.094

0.094

0.065



1.62

0

Table 1 Demographic parameter estimates for the Grey Partridge. Estimates are averages of annual values derived from autumn count data from the Sussex study, 19902000.
2.5 Comparison Between Models with and Without Density Dependence
Effects of Changes in the Agricultural Environment
In the presence of density dependence, it becomes meaningless to talk of rate of population change because the modelled population no longer changes with a constant rate, but seeks to return to an equilibrium level where, by definition, the rate of change is zero (Aebischer 1991). Because population size at equilibrium is stable, the densitydependent demographic parameters become fixed, so that it is not possible to present a figure akin to Figure 1 for the densitydependent model (for the Grey Partridge, both adult survival and production contain densitydependent elements).
To compare outputs from the two types of model, what is needed therefore is a representation of population change in relation to variables that do not themselves depend on density. An obvious one is chick survival rate, the sole partridge demographic parameter that does not vary with density (Potts 1980, 1986). Another one is the amount of nesting cover (km/km^{2}), an environmental variable that enters into the densitydependent survival relationships (equations (1) and (3)).
Using nesting cover has the further advantage that it provides a direct link to changes in the agricultural environment, as it represents the density of noncropped linear features in the landscape that are suitable for nesting, such as hedgerows and grass banks. Rands (1986a, 1987) showed how the attractiveness of different areas to Grey Partridges was determined by the amount and quality of available nesting habitat. It is also known that chick survival rate varies in relation to the availability of chickfood invertebrates in cereals, itself dependent on the pesticide regime (Potts 1973, Potts 1977, Green 1984, Potts 1986, Rands 1986b, Potts & Aebischer 1991). It seems therefore appropriate to attempt a recalibration of chick survival rate in terms of the amount of insectrich broodrearing habitat available in the landscape, to provide a second link to changes in the agricultural environment.
The GCT Conservation Headland research programme (Boatman & Sotherton 1988, Sotherton et al. 1989, Sotherton 1991, Sotherton et al. 1993) offers a sound scientific basis for the recalibration exercise. Conservation Headlands are the outer 6 m of a cereal crop that receive only selective herbicide treatment and no summer insecticide treatment, so that the crop understorey is rich in broadleaved weeds and associated invertebrates. This constitutes an ideal insectrich broodrearing habitat for Grey Partridges. Sotherton et al. (1993) found experimentally that, over eight years, chicksurvival rate averaged 0.23 on fully sprayed areas and 0.39 on areas with Conservation Headlands within the same East Anglian farms. They also reported similar data on mean brood size at six weeks in Hampshire and eastern England, with an average of 4.8 chicks on fully sprayed blocks and 7.3 chicks on blocks with Conservation Headlands. Using the conversion formula in Potts (1986), these are equivalent to chick survival rates of 0.19 and 0.37 respectively. Further information on chick survival rate in a fully sprayed situation come from Sussex, where chick survival rate averaged 0.22 over six years when broadspectrum herbicides and insecticides had been used intensively on an area of 7 km^{2} (Aebischer & Potts 1998).
It seems reasonable to conclude that under an intensive spraying regime, chick survival rate will be close to 0.20, which can be taken as the rate corresponding to zero availability of insectrich broodrearing habitat within the arable area. With Conservation Headlands, chick survival rate was around 0.38. Landscape data collated by Oxford under this DEFRA project shows that in a typical arable situation, 54% of land is arable, of which 69% is in cereals, with an average field size of 8.24 ha. Assuming square cereal fields with Conservation Headlands along three sides of each field, the amount of insectrich broodrearing habitat would be 6.3% of cereal area, or 4.3% of arable area. Thus a 1% increase in such habitat on the arable area was equivalent to an increase in chick survival rate of around 0.04. By extrapolation, having 6% of arable area as insectrich broodrearing habitat would result in a chick survival rate of around 0.44, close to that recorded during the prepesticide era (Potts 1986). The current average chick survival rate of 29.4% observed in Sussex (Table 1) corresponds to 2.35% of the arable area being insectrich.
The Oxford landscape data also indicated that the average amount of hedgerow in a typical arable situation was 4 km/km^{2}. Using this value as a measure of nesting cover, together with an average chick survival rate of 29.4% (Table 1), resulted in an equilibrium density of 5.45 pairs/km^{2} according to the densitydependent model. This was taken as the standard reference level against which to compare equilibrium levels for all combinations of nesting cover (from 2 to 10 km/km^{2}) and percentage of insectrich arable area (from 0% to 6%), so that the latter were expressed as percentage differences relative to the reference level (Figure 2). The same combinations of nesting cover and percentage of insectrich arable area were used as input into the densityindependent model for comparison (Figure 3). To improve comparability, the output, which was an annual rate of change, was transformed to a relative difference over five years, this being the approximate time for the densitydependent model to reach equilibrium.
When density dependence was included in the model, the contours indicated that when insectrich broodrearing habitat was scarce, increasing its proportional coverage had a similar positive effect on density as increasing the amount of nesting cover. When the insectrich broodrearing habitat reached around 3% of arable area, the positive response in density was strongest in relation to increasing nesting cover. The model predicted that a 1% increase in insectrich broodrearing habitat, from 2.35% (equivalent to the current average chick survival rate) to 3.35%, would yield a 13.8% increase in population equilibrium when nesting cover is kept at its average of 4 km/km^{2}. Conversely, keeping insectrich habitat constant at 2.35% of arable area, a 1% increase in nesting cover (from 4 to 4.04 km/km^{2}) would produce a population increase of 1.1%.
When density dependence was omitted from the model, the pattern of contours was very different. Availability of nesting cover no longer played a role in determining density (all contours horizontal). Generally the amount of insectrich broodrearing habitat required to avoid a decline was higher than when density dependence was present.
2.6 Conclusions
1. Ignoring density dependence in the partridge model leaves unrecognised the role of nesting cover as an important determinant of Grey Partridge density in the agricultural environment.
2. To improve prospects for the recovery of the Grey Partridge, equal emphasis should be given to improving the amount of arable land containing insectrich broodrearing cover and nesting cover.
3. Considering that the average annual rate of population change over the last decade was 0.926, stabilising the population (BAP Objective 4.2, Anon. 1995) requires a relative increase of 8% (100/0.926100). This can be achieved by having 3% of arable area as insectrich broodrearing habitat, or by increasing nesting cover from 4 to 4.3 km/km^{2}.
4. The average annual rate of population change over the last decade of 0.926 amounts to a change of 0.58 (42% decline) between 1995, when the BAP targets were set, and 2002. Recovery of the population to its 1995 level (BAP target 4.3, Anon. 1995) requires a relative increase of 72% (100/0.58100). It depends on achieving 6.5 km/km^{2} of nesting cover together with 3% of arable area being insectrich.
5. Given that setaside currently represents on average 11% of arable land (Oxford data, this DEFRA project), its judicious use could satisfy these requirements, e.g. by managing it as adjacent strips of tussocky grass for nesting and cereal mixtures for broodrearing (total width 20 m) distributed over the farm. Such management is already permitted under existing prescriptions.
Figure 1 Contour map of the annual rate of population change for the Grey Partridge in relation to female annual production and survival, assuming no density dependence in either production or survival. The contour for the 19902000 annual rate observed in the BTO CBC data is superimposed (0.926), as is the point corresponding to the average parameters in Table 1.
Figure 2 Densitydependent model. Contour map of relative differences (%) in equilibrium levels according to the amount of nesting cover and the percentage of arable area made up of insectrich broodrearing habitat. The baseline (contour 0) passes through the density predicted on an “average” arable farm over the last decade.
Figure 3 Densityindependent model. Contour map of relative differences (%) after five years according to the amount of nesting cover and the percentage of arable area made up of insectrich broodrearing habitat.
References
Aebischer, N.J. (1991) Sustainable yields: gamebirds as a harvestable resource. Gibier Faune Sauvage, 8, 335351.
Aebischer, N.J. & Potts, G.R. (1998) Spatial changes in Grey Partridge (Perdix perdix) distribution in relation to 25 years of changing agriculture in Sussex, U.K. Gibier Faune Sauvage, 15, 293308.
Anon. (1995) Biodiversity: The UK Steering Group Report. Vol. 2: Action Plans. Her Majesty’s Stationery Office, London.
Boatman, N.D. & Sotherton, N.W. (1988) The agronomic consequences and costs of managing field margins for game and wildlife conservation. Aspects of Applied Biology, 17, 4756.
Green, R.E. (1984) The feeding ecology and survival of partridge chicks (Alectoris rufa and Perdix perdix) on arable farmland in East Anglia. Journal of Applied Ecology, 21, 817830.
Potts, G.R. (1973) Factors governing the chick survival rate of the grey partridge (Perdix perdix). Proceedings of the Xth International Congress of Game Biologists, 10, 8596.
Potts, G.R. (1977) Population dynamics of the grey partridge: overall effects of herbicides and insecticides on chick survival rates. In: Proceedings of the XIIIth International Congress of Game Biologists, pp. 203211. International Congress of Game Biologists, Atlanta, Georgia.
Potts, G.R. (1980) The effects of modern agriculture, nest predation and game management
on the population ecology of partridges (Perdix perdix and Alectoris rufa). Advances in Ecological Research, 11, 179.
Potts, G.R. (1986) The Partridge: Pesticides, Predation and Conservation. Collins, London.
Potts, G.R. & Aebischer, N.J. (1991) Modelling the population dynamics of the Grey Partridge: conservation and management. In: Bird Population Studies: Their Relevance to Conservation Management (eds C.M. Perrins, J.D. Lebreton & G.J.M. Hirons), pp. 373390. Oxford University Press, Oxford.
Rands, M.R.W. (1986a) Effect of hedgerow characteristics on partridge breeding densities. Journal of Applied Ecology, 23, 479487.
Rands, M.R.W. (1986b) The survival of gamebird (Galliformes) chicks in relation to pesticide use on cereals. Ibis, 128, 5764.
Rands, M.R.W. (1987) Hedgerow management for the conservation of partridges Perdix perdix and Alectoris rufa. Biological Conservation, 40, 127139.
Sotherton, N.W. (1991) Conservation Headlands: a practical combination of intensive cereal farming and conservation. In: The Ecology of Temperate Cereal Fields (eds L.G. Firbank, N. Carter, J.F. Derbyshire & G.R. Potts), pp. 373397. Blackwell Scientific Publications, Oxford.
Sotherton, N.W., Boatman, N.D. & Rands, M.R.W. (1989) The 'Conservation Headland' experiment in cereal ecosystems. The Entomologist, 108, 135143.
Sotherton, N.W., Robertson, P.A. & Dowell, S.D. (1993) Manipulating pesticide use to increase the production of wild game birds in Britain. In: Quail III: National Quail Symposium (ed. by K.E. Church & T.V. Dailey), pp. 92101. Kansas Department of Wildlife and Parks, Pratt.
Varley, G.C. & Gradwell, G.R. (1963) The interpretation of insect population changes. Proceedings of the Ceylon Association for the Advancement of Science, 18(D), 142156.
Appendices to CSG15
May 2003
