• demography;
  • geographical range;
  • macroecology;
  • spatial structure


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Few studies have sought to document the variation in demographic rates exhibited by local populations across the geographical range of a species. None the less, such information has the potential to yield insights into the factors that determine the internal abundance structure of ranges.
  • 2
    Major components of demographic rates are reported here for the holly leaf-miner Phytomyza ilicis Curtis across its natural geographical range in Europe, including areas at each of the extreme limits and at the core of its occurrence.
  • 3
    Correlograms reveal that a number of these components exhibit significant spatial structure across the geographical range and, in conjunction with interpolated maps, suggest that some of these patterns are, in terms of broad trends, reasonably simple.
  • 4
    Across the whole range, individual mortality components are largely independent of one another and of leaf-miner population densities, often exhibiting very different spatial patterns. Thus, P. ilicis populations experience markedly different mortality profiles across the range.
  • 5
    While some correlations between broad-scale environmental variation and the components of demographic rates were found, it was not possible to separate these effects from shared spatial structure.
  • 6
    In sum, the contribution of different components of demographic rates to local densities varies markedly across the geographical range of P. ilicis, and in isolation the pattern observed in any one locality provides limited information on what is occurring elsewhere, but these components none the less often exhibit systematic geographical patterns of variation.


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

It is well documented that species show significant variation in local population densities at a wide variety of spatial scales (e.g. Taylor, Woiwod & Perry 1980; Root 1988; Gaston 1994; Brown, Mehlman & Stevens 1995; Price, Droege & Price 1995). The population density in any particular area is necessarily the net outcome of four basic demographic processes: births and immigration minus deaths and emigration. One potentially useful approach to understanding the causes of spatial variation in local population densities therefore may be to investigate spatial variation in underlying demo-graphic rates and how these interact with environmental conditions (Maurer & Brown 1989; Lawton 1993; Holt et al. 1997; Holt & Keitt 2000). Unfortunately, for most species, accurate measurement of demographic rates for wild populations is extremely difficult, if not impossible. As a result, there are few data on how they may vary between areas and affect population densities at all but relatively small spatial scales (e.g. Whittaker 1971; Randall 1982). However, this type of information certainly has the potential to increase understanding of the factors that determine the internal structure and boundaries of entire geographical ranges.

Studies have shown that there can be systematic changes in particular demographic rates, or components thereof, across geographical ranges. Most notably a number have examined patterns of latitudinal, longitudinal or altitudinal variation in clutch size and reproductive success in birds (e.g. Peakall 1970; Ricklefs 1972; Koenig 1986; Yom-tov, Christie & Iglesias 1994; Hendricks 1997; Sanz 1997; Dunn et al. 2000), and some other groups (e.g. Fleming & Gross 1990; Bradford, Taylor & Allan 1997). Equivalent data on spatial variation in levels of mortality, and in the major components of that mortality, are much more scarce (but see Rogers 1979; Rogers & Randolph 1986; Randolph 1994). Thus, to date, even for systems in which immigration and emigration are unlikely to constitute major influences on local populations, there have been few studies that have attempted to begin to build a more complete picture of variation in demographic rates across the range of any species.

The holly leaf-miner Phytomyza ilicis Curtis exhibits a distinct pattern of variation in mean local densities across its geographical range at a broad spatial scale (Brewer & Gaston 2002). This pattern is apparent despite the high degree of variation in densities that is found typically between trees within a local area and between nearby sites (Heads & Lawton 1983; Valladares & Lawton 1991; McGeoch & Gaston 2000). Several previous studies of range structure have claimed that most species tend to exhibit higher local population densities in the centre of the range that gradually decrease towards the edges (Brown et al. 1995). However, the holly leaf-miner exhibits a structure that departs markedly from this pattern (Brewer & Gaston 2002). Peak densities lie at the north-eastern edge of the range, with high to moderate densities predominantly falling in a band running north-east to south-west across its geographical distribution. With a few notable deviations, densities largely decline away from this region. This pattern can to some extent be correlated with broad scale environmental variation across the range of the holly leaf-miner. It can potentially be understood in terms of a modified version of Brown's (1984; 1995; see also Brown & Lomolino 1998) niche-based theory of the spatial structure of the geographical range. This proposes that if (i) the abundance and distribution of a species are determined by combinations of many physical and biotic variables, and that spatial variation in population density reflects the probability density distribution of the required combinations of these variables, and (ii) some sets of environmental variables are distributed independently of each other, and environmental variation is spatially autocorrelated, then density should be highest at the centre of the range of a species and should decline towards the boundary. The spatial structure of the environment to which the holly leaf-miner responds does not seem to follow this simple pattern, but otherwise the basic model seems likely to apply (Brewer & Gaston 2002).

In this paper we investigate whether there are also changes in components of demographic rates between local populations of the holly leaf-miner across broad geographical scales and whether these changes can in turn be correlated with differences in population density and environmental variation.

study system

European (or English) holly Ilex aquifolium L. is a relatively small, dioecious, evergreen tree. It has distinctive, dark green, glossy leaves that are usually spiny with a relatively thick cuticle. Its natural range extends throughout north-western, central and southern Europe (Peterken & Lloyd 1967; Hultén & Fries 1986). It can also be found less commonly in scattered localities in parts of North Africa and has been reported as having a narrow band of distribution extending into Asia Minor although there is some doubt as to whether this is indeed so.

Phytomyza ilicis, the holly leaf-miner, is the most common insect herbivore of European holly. It is strictly monophagous, so its geographical range is limited ultimately by the availability of holly trees. The life histories of P. ilicis and its natural enemies in Britain have been described in detail by Cameron (1939) and Lewis & Taylor (1967). Put briefly, the holly leaf-miner exhibits a univoltine lifecycle. Eggs are laid in June on new holly leaves (the tree has just one flush of new leaves per year) into the base of the underside of the midrib. When laying an egg, the adult female P. ilicis leaves a characteristic scar due to the insertion of her ovipositor into the midrib. The presence of these scars mean that oviposition density for a local population may easily be censused. The vast majority of the life history of each individual is spent inhabiting a single holly leaf. After hatching, the larvae eat through the midrib and enter the outer parenchyma of the leaf lamina during the autumn. They feed throughout the following winter months and pupate in the mine in March, emerging from the leaf as adults in late May or June. The short lifespan of the adults and the tight synchronization between their emergence and the flush of new leaves severely restricts the influence of immigration and emigration on local population dynamics.

During the period spent within holly leaves, a leaf-miner population may be subject to a number of potential mortalities that are largely sequential, albeit with some overlap. They have been relatively well documented (Cameron 1939; Lewis & Taylor 1967; Heads & Lawton 1983) and include miscellaneous larval deaths, larval parasitism by the parasitoid Chrysocharis gemma Walker (Hymenoptera: Eulophidae), bird predation, pupal parasitism by at least eight species of hymenopteran parasitoid and miscellaneous pupal mortality. These causes of mortality can be identified by dissection of the mine at the end of the lifecycle of the leaf-miner. In addition, larvae that successfully complete their development and emerge as adult flies can also be identified from the characteristic emergence holes on the leaf surface. Therefore, the holly leaf-miner has a lifecycle that is more suitable than most for measuring a number of the components of the basic demographic rates of a local population. The amount of oviposition (births), a number of components of mortality and the rate of successful emergence for a local population in any one year can all be assessed by rapid censusing techniques and dissection of a representative sample of leaves from the holly trees at a site.


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

data collection

Between June and December 1998 we conducted a survey of leaf-miner densities throughout most of the natural range of holly. Our aim was to obtain a good coverage, sampling both the centre and the range edges. The survey was organized as a series of routes of varying duration: Spain and Portugal; Norway; Germany, Denmark, Switzerland, Italy and France; United Kingdom; Ireland; Greece. During each survey we sampled leaf-miner densities at as many sites as time allowed.

Throughout its range, holly occurs naturally predominantly as a subordinate or understorey plant of deciduous woodland. Wherever possible it was sampled in this habitat type [in practice, this meant that the vast majority (> 85%) of trees were sampled from deciduous woodland, including at both the centre and the edge of its geographical range]. This served particularly to avoid the confounding effects of the planting of holly in heavily human modified environments (e.g. cemeteries, urbanized areas); holly trees in urban areas, for example, typically support higher mine densities than those in deciduous woodland (Brewer 2000).

Leaf-miner and egg scar density estimates were taken from 10 trees at each sampling site, or as many as possible where less than 10 trees were present. Following Heads & Lawton (1983) and Valladares & Lawton (1991), density was estimated on each tree by haphazardly sampling 200 1-year-old leaves (or all the leaves of this age if less than 200 were present on the tree) from all around the canopy, between ground level and a height of approximately 2 m. This protocol gives an estimate of density for the previous growing season. Leaves of a suitable age can readily be distinguished by their position between the annual nodes on the branch. For each leaf, the number of mines and egg scars present were recorded. Typically only one mine per leaf is found except in areas of relatively high leaf-miner density. The number of egg scars is often more variable (Heads & Lawton 1983; Brewer 2000).

For trees on which no mines or egg scars were found in a sample of 200 leaves, an exhaustive search was made of the rest of the tree. This occurred in only four trees of the 703 sampled, in which case densities were treated as an arbitrarily small number (0·1 mines or egg scars per 200 leaves).

When mines were found, a haphazardly chosen sample of 50 mined leaves was removed from each tree. If 50 mines could not be found on a particular tree, all the visible mines of the appropriate age were removed. These mines were dissected to establish the number of the holly leaf-miner individuals that were subject to the different mortalities or emerged successfully as adults, following the procedures given by Heads & Lawton (1983).

The unit for the analyses was the site. The spatial relations of sites, which can be viewed as one form of potential nonindependence of data points (Legendre 1993), were addressed explicitly in the analyses (see below). The geographical location (expressed as decimal degrees longitude and latitude) of each site and its altitude (m) were obtained from appropriate maps.

meteorological data

Regional climatic data over a 0·5° × 0·5° grid were obtained from a coverage of 20th-century terrestrial surface climate (New, Hulme & Jones 2000). From this coverage, mean values for winter (September–February) and summer (March–August) temperatures, precipitation and humidity were calculated from data for the 10-year period between 1986 and 1995 inclusive.

data analysis

The demographic data sets

Because the mortalities acting on the holly leaf-miner are largely sequential in this system, the number of mines that can potentially succumb to a particular mortality is the total number of mines minus the number killed by previous mortalities. This smaller proportion of mines is used to calculate apparent mortality (Bellows, Van Driesche & Elkinton 1992) as opposed to real mortality (which is the proportion of all mines). In all the analyses below, apparent mortalities were used unless otherwise stated. These were calculated as the total number of each mortality recorded divided by the number of mines available to that mortality across all trees that were sampled. Oviposition rate was measured as the total number of egg scars divided by the number of leaves examined during density estimation at each site. Successful emergence was measured as the proportion of the total number of leaf-miners collected, as it will always be 100% of the mines that have not succumbed to any previous mortality.

Spatial autocorrelation

Spatial autocorrelation analysis (Cliff & Ord 1973; see also Legendre & Fortin 1989) was used to characterize the spatial structure of the demographic rates of the holly leaf-miner that we measured. Moran's I was calculated for 15 equal-distance intervals, and spatial correlograms were produced and tested for significant spatial dependence. As the survey covered a relatively large geographical area, site coordinates (measured as decimal degrees longitude and latitude) were not treated as Cartesian coordinates when measuring distances between them. Instead, distances along great circles were calculated to take into account the curvature of the earth's surface. Bonferroni's correction for multiple comparisons was used when assessing overall correlogram significance.


The demographic rates were interpolated across space to give a visual impression of the changes in these rates across the geographical range. Each demographic rate was entered as a point coverage into a geographical information system (GIS; ArcInfo and ArcView, Environmental Systems Research Institute Inc. 1997a,b), and interpolated using an inverse distance weighting technique (IDW) to create a grid comprising 0·1° × 0·1° cells over the study area. Seven nearest neighbours were used to estimate the values, with weights inversely proportional to the square of the distance from the estimated cell.

Partial regression analysis

We used partial regression analysis to model oviposition, the mortality components and successful emergence in terms of the environmental variables. Following the methodology suggested by Legendre (1993), we used the technique to estimate how much variation in the components of leaf-miner demographic rates could be attributed to regional climatic variation and altitude once the effect of spatial location had been taken into account. The spatial component of the holly leaf-miner demographic data was modelled using a third-order polynomial of the form:

  • f(x,y) = b0 + b1x + b2y + b3x2 + b4xy + b5y2 + b6x3 + b7x2y + b8xy2 + b9y3

where x and y represent longitude and latitude, respectively. Significant terms, determined by conducting a stepwise regression of leaf-miner demographic rates on a matrix of all of the terms in the expression, were retained to construct a new matrix of spatial variables to be used in the subsequent analysis. Similarly, the leaf-miner demographic rates were regressed onto a matrix containing the climate and altitude data and significant terms, determined after stepwise regression by elimination, were retained.

The combined effect of both the environmental and spatial variables on leaf-miner demographic rates was calculated by multiple regression of the rates onto both sets of predictive variables combined. The explanatory potential of the environmental variables, after correcting for spatial dependence, was calculated by measuring the change in deviance explained by the regression model after the environmental variables were removed. This fraction was then tested to determine whether the change in deviance was statistically significant.

At the end of the partial regression analysis, variation in leaf-miner demographic rates could be divided into four components: (a) non-spatial environmental −the fraction that can be explained by the environmental variables independent of any spatial structure; (b) spatially structured environmental − spatial structuring in leaf-miner demographic data that is shared with the environmental data; (c) non-environmental spatial −spatial structure in leaf-miner demographic data that is not shared with the environmental variables; and (d) unexplained (residual) variation. These components and their associated probabilities allow several hypotheses about the causal relationships between environmental variation, spatial position and leaf-miner demographic rates to be tested.

The method employed here is essentially the same as Legendre's (1993) third extension to partial regression analysis with some minor modifications (see also Legendre & Legendre 1998). Because the demographic data were in the form of proportions (e.g. apparent mortality rate per number of mines available), binomial errors were modelled. This results in proportions of deviance explained, which are analogous to partial r2 statistics at the end of the analysis. In addition, during the stepwise regressions, any overdispersion in the residual deviance was corrected for using William's procedure where appropriate (Collett 1991; Crawley 1993) before testing for parameter significance. All the partial regression analyses were conducted using glim version 3·77 (Royal Statistical Society, London).


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

spatial structure in the demographic data

Spatial autocorrelation

The correlograms show that significant spatial dependence is apparent in oviposition, successful emergence and a number of the components of leaf-miner mortality that we measured across the geographical range of the holly leaf-miner (Fig. 1). At least some significant positive autocorrelation was found at shorter lags for all of the rates measured, indicating that sites near to each other tend to have more similar levels of natality and mortality compared to those further apart. This correlation falls close to or below zero at intermediate lags followed by a considerable amount of variation between the rates at the longest separation distances between sites. At these long lags it should be noted that Moran's I is often calculated from a relatively low number of pairs of sites (Fig. 1h) and so care should be taken when interpreting these values. It is clear that where some of the mortality rates measured show what might be considered strong systematic spatial structure (e.g. larval parasitism) others showed a lot weaker (e.g. bird predation) or no (pupal parasitism) spatial structure across the whole of the geographical range. Oviposition rates show a similar pattern of autocorrelation to population densities (Brewer & Gaston 2002) with positive autocorrelation up to approximately 400 km and marked negative autocorrelation between sites separated by the longest distances, representing sites at opposite edges of the geographical range.


Figure 1. Spatial correlograms of the holly leaf-miner demographic rate components from across its geographical range. Scale on the x-axis represents the maximum in each distance class. Filled circles indicate significant values for Moran's I (P < 0·05). Overall correlogram significance was tested using Bonferroni's correction for multiple comparisons. (a) oviposition density (P < 0·01); (b) miscellaneous larval mortality (P < 0·01); (c) larval parasitism (P < 0·01); (d) bird predation (P < 0·05); (e) pupal parasitism (P > 0·05); (f) miscellaneous pupal mortality (P < 0·05); (g) successful emergence (P < 0·01); (h) the number of pairs of comparisons within each distance class for the autocorrelation.

Download figure to PowerPoint


The exclusive use of interpolated maps to demonstrate spatial structure in a data set can be a misleading exercise. However, as the results of our spatial autocorrelation analyses demonstrated statistically significant spatial trends in the leaf-miner demographic data, we present such maps to serve as a useful way of visualizing this spatial structure (Fig. 2). The similarity between the correlograms of local population densities of the holly leaf-miner (Brewer & Gaston 2002) and its oviposition rates is borne out in the interpolated map. Highest rates of oviposition are found in the north-eastern extreme of the range (Norway) with a band of moderate densities running towards the south-west. Oviposition tends to decline away from this band. However, as with population densities, there are exceptions to this, most notably at several sites in Italy.


Figure 2. Interpolated surfaces of the holly leaf-miner demographic rate components across the range of European holly. Points on the map indicate the locations of the sites where holly leaves were sampled. (a) oviposition densities; (b) miscellaneous larval mortality; (c) larval parasitism; (d) bird predation; (e) pupal parasitism; (f) successful emergence; (g) relative density of mines that emerge successfully.

Download figure to PowerPoint

Some of the demographic data show more obvious trends across space than others. Both larval parasitism and miscellaneous larval mortality are at their highest in what might be considered the geographical centre of the range of the holly leaf-miner. Successful emergence is highest in the eastern part of the range. Simple patterns in pupal parasitism and bird predation, however, are not so obvious.

A comparison of the interpolated maps of oviposition density (Fig. 2a) and the density of adults (calculated as proportion of successful emergence × mine density, Fig. 2g) shows that the pattern of relative leaf-miner densities remains relatively stable throughout the lifecycle. Sites with a high level of oviposition also have the highest levels of successful emergence and vice versa.

Comparison between range edges and the rest of the geographical range

The geographical range of the holly leaf-miner is rather irregular and spreads across several distinct land masses in Europe. This makes it difficult to separate edge and centre populations in a purely objective manner for the purposes of comparison. However, we identified a number of sites that could reasonably be considered edge sites. In the north we chose all sites from Norway, Denmark and the northernmost sites in Scotland. In the east we chose the limit of the range of holly in Germany and south to Italy, and in the west, Spanish sites, where leaf-miners were present, and sites in the west of Ireland. These sites were compared to the remaining sites where leaf-miners were present which were considered ‘non-edge’ sites. A one-way anova, modelling binomial errors, was used to compare the levels of each demographic component between these two sets of sites, again correcting for any over-dispersion before conducting the significance tests (Crawley 1993). Edge sites tend to have lower levels of larval parasitism and miscellaneous larval mortality than other sites but higher rates of bird predation and successful emergence (Table 1). However, there were no significant differences in miscellaneous larval mortality or pupal parasitism.

Table 1.  Comparison of levels of holly leaf-miner density and the demographic rates components between edge and centre sites using anova, modelling binomial errors
Demographic rate componentSignificance test 
Mine densityF1,61 = 0·08, NS 
Oviposition densityF1,61 = 1·137, NS 
Miscellaneous larval mortalityF1,54 = 7·666, P < 0·05Edge sites have lower rates of miscellaneous larval mortality
Larval parasitismF1,54 = 22·296, P < 0·05Edge sites have lower rates of larval parasitism
Bird predationF1,54 = 4·804, P < 0·05Edge sites have higher rates of bird predation
Pupal parasitismF1,54 = 0·072, NS 
Miscellaneous pupal mortalityF 1,54 = 0·377, NS 
Successful emergenceF1,54 = 6·117, P < 0·05Edge sites have higher rates of successful emergence

correlations between demographic rate components and local population densities

The extent to which the mortality rates covary across the whole of the range was calculated. Significant positive correlation was found between miscellaneous larval mortality and larval parasitism (r = 0·679, n = 57, P < 0·001, Fig. 3), but no other mortality rates were correlated between sites. No significant correlations were found between each of the mortality rates and leaf-miner population densities at this scale. Successful emergence showed a weak negative (r = −0·294; n= 57; P > 0·05) correlation with mine density but this was not significantly different from zero. However, when considering all trees, Fig. 4 shows that while there is not a simple linear relationship between the two variables there does appear to be an upper limit on the rate of successful emergence that declines with leaf-miner density. At low densities successful emergence is very variable but as the proportion of mined leaves increases, the maximum rate of successful emergence drops consistently.


Figure 3. Relationship between larval parasitism and miscellaneous larval mortality for all sites (r = 0·679, n= 57, P < 0·01).

Download figure to PowerPoint


Figure 4. Relationship between the proportion of successful emergence and the proportion of mined leaves for all trees sampled in the survey.

Download figure to PowerPoint

demographic rate components and environmental variation

Partial regression analyses

Using stepwise regressions, variation in all of the demographic rates could, to some extent, be described significantly by the environmental data when considered alone (Table 2). However, a large proportion of the environmental variation that contributed to explaining the demographic rates is spatially structured itself. As a result, when spatial position is taken into account using the terms from the coordinates polynomial, a significant environmental effect was detected for oviposition only (Table 2).

Table 2.  Results of the partial regressions of the holly leaf-miner oviposition, mortality and successful emergence on environmental data taking into account a spatial component. Environmental variables: wt= winter temperature; st= summer temperature; wp= winter precipitation; sp= summer precipitation; wh= winter humidity; sh= summer humidity; alt= altitude. See text for details
Leaf-miner dataEnvironmental variablesCoordinate termsDeviance accounted for in leaf-miner data (%)Significance test
TotalEnvEnv * SpaceSpace
Ovipositionwt st wpspshx y x2y2xy xy269·314·135·819·4F5,80 = 4·64, P < 0·05
Miscellaneous larval mortalitywtx xy x2x2y xy2x3y364·6 1·612·250·8F1,48 = 0·561, NS
Larval parasitismwt whx y y277·9 0·326·950·7F2,49 = 0·280, NS
Bird predationstspwhx2x2y y347·113·632·7 0·8F3,50 = 2·67, NS
Pupal parasitismspwhx xy xy239·4 2·416·920·1F2,51 = 0·560, NS
Miscellaneous pupal mortalityaltx y xy xy236·0 018·317·7F1,50 = 1·69, NS
Successful emergencewt stx x2xy xy2x368·8 1·711·955·2F2,49 = 0·937, NS


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Holly leaf-miner densities exhibit a distinct spatial structure across the geographical range of the species (Brewer & Gaston 2002). In addition, oviposition density, a number of the components of mortality and successful emergence are positively spatially autocorrelated at short separation distances and also show a broad scale spatial structure across the whole geographical range. However, with the exception of miscellaneous larval mortality and larval parasitism, each of the mortality components are largely independent of one another. The consequence of this is that sites close together are likely to have similar mortality profiles whereas populations separated by large distances experience very different patterns of mortality. The general picture that emerges then is that, at the broadest spatial scale, the range of the holly leaf-miner is set upon a complex yet structured template of mortality with populations experiencing different mortality profiles dependent upon their location.

There does not appear to be a simple relationship between any of the components of mortality and population densities between sites. No single component of mortality that we were able to measure during the survey can be responsible for regulating leaf-miner populations across the whole range and hence creating the spatial abundance structure that we have recorded. This should not be surprising, given the geographical variation in the distribution of natural enemies and climate. In addition, previous studies have shown that different factors may be responsible for determining the range limits of a species in different regions (MacArthur 1972; see Hoffmann & Blows 1994 for a brief review) and this may also be true for determining variation in local population densities.

Perhaps more important is that under the equilibrium model presented by Maurer & Brown (1989) a relationship between per capita birth or death rates and differences in density between sites should not be expected once local population densities have reached equilibrium levels. At equilibrium densities, density-dependent biotic factors would serve to reduce per capita birth rates and increase per capita death rates. Therefore, comparisons between survival and birth rates between populations in different parts of the range may not give a direct indication of the reasons for differences in density. However, qualitative differences in the components of mortality may certainly provide clues to range structure, for example the potential regulatory effect of larval parasitism in some parts of the range of the holly leaf-miner (see below).


Oviposition levels of the holly leaf-miner necessarily set an upper bound to leaf-miner densities, as the two were measured together during the survey. Across all sites, 48% of all eggs, as indicated by oviposition scars, successfully hatched to become larvae visible in the leaf lamina. This is consistent with the findings of the local study of the holly leaf-miner by Valladares & Lawton (1991). However, a plot for all of the trees in the range-wide survey shows that the proportion of eggs that hatch successfully tends to decrease at higher leaf-miner densities (Fig. 5).


Figure 5. Relationship between the total number of mines and total oviposition for all trees (y = 0·0004x2 + 0·4262x, r2 = 0·8403, P < 0·001).

Download figure to PowerPoint

The results of the partial regression analysis (Table 2) show that a significant effect of environmental variation on oviposition rate can be separated from the potentially confounding effect of spatial position at this geographical scale. This is in accordance with the results from the density survey (Brewer & Gaston 2002).

miscellaneous larval mortality

Miscellaneous larval mortality is thought to occur relatively early in the lifecycle of the holly leaf-miner as mines suffering from this mortality are typically small. It may be due to a number of possible causes, including death during the overwintering phase of the lifecycle, as the result of a physiological response of the host tree or from attacks by natural enemies. Miscellaneous larval mortality of the holly leaf-miner has been found previously to be density-independent at local spatial scales (Heads & Lawton 1983) so that it seems unlikely that it is a consequence of intraspecific competition. We did not find any evidence for density-dependence in miscellaneous larval mortality across the whole range of the holly leaf-miner (r = 0·002, n= 57, NS).

There is, however, a positive correlation between miscellaneous larval mortality and larval parasitism by Chrysocharis gemma (Fig. 3). While some degree of miscellaneous mortality is present at most sites, regions with high rates of larval parasitism tend also to have higher rates of miscellaneous larval mortality. It is therefore possible that a component of miscellaneous larval mortality may be due to attacks by C. gemma, when it is ovipositing into the leaf-miner larva, but this would probably be in addition to other factors.

larval parasitism

Chrysocharis gemma is the only recorded larval parasite of the holly leaf-miner and has been introduced into Canada as a biological control agent in commercial holly crops (Hansson 1985). Auerbach, Connor & Mopper (1995) point out that in nearly every study examining the mortality of leaf-miners, larval parasitism is an important factor, and our data support this observation over a considerable proportion of the range of the holly leaf-miner. Highest levels of larval parasitism are found in the centre of the range. Larval parasitism is almost completely absent from north-eastern regions. Leaf-miner densities are at their highest in the north-east which suggests that C. gemma may contribute towards reducing leaf-miner densities elsewhere. However, across the whole of the range there is no significant relationship between leaf-miner density and larval parasitism between sites (r = 0·002, n = 56, NS). Heads & Lawton (1983) found that C. gemma exhibited an aggregative response to leaf-miners within a single habitat patch. We conducted regression analyses of C. gemma attack rates on both real and apparent leaf-miner densities between trees separately for 21 sites where C. gemma was present, but no significant positive relationships were found. However, the power of each statistical test was not high as we were limited to the relatively low number of trees sampled at each site.

The absence of mortality caused by C. gemma in the north-east of the range of the holly leaf-miner is in accordance with the known distribution of this parasitoid (Hansson 1985). C. gemma overwinters as an adult in the evergreen host plants of the leaf-miners that it parasitizes. It is possible that its absence from northern parts of the leaf-miner's range may result from winter conditions that are too extreme for it to survive in these regions. Winter temperature and humidity contributed significantly to the incidence of larval parasitism when considered alone, but this effect could not be disentangled from the confounding effect of spatial position.

bird predation

The holly leaf-miner can suffer high levels of bird predation in some habitats (Heads & Lawton 1983; McGeoch & Gaston 2000). Heads & Lawton (1983) found that bird predation accounted for approximately 40% of total mortality in their local study, although there was a great deal of variation between samples. In addition, they found a surprising negative relationship between mine density and levels of bird predation. However, they point out that the ability to detect density-dependent mortality and the nature of the relationship may depend on the spatial scale over which the samples are taken. We did not find such a relationship over a much broader spatial scale.

Despite the fact that no obvious spatial structure could be discerned from the correlogram and the spatial component of variation in the partial regression analysis was small, bird predation was significantly higher at the edges of the geographical range of the holly leaf-miner than elsewhere. This is highlighted well in the interpolated map (Fig. 2d).

pupal parasitism

Pupal parasitism can be an important source of mortality for leaf-miners (Auerbach et al. 1995), often with species-rich parasitoid communities associated with a single leaf-miner species. The holly leaf-miner is no exception. In the United Kingdom, eight species of parasitoids have been recorded attacking the pupae of the holly leaf-miner with considerable variation in the species found between regions (Cameron 1939; K.J. Gaston, personal observation). We did not find significant spatial autocorrelation in the amount of pupal parasitism across the range. This is, perhaps, not surprising because, if this component of mortality can be subdivided into those mortalities caused by each parasitoid species, and if each of these species has a range largely independent of the others, then one would not expect to find spatial structure in pupal parasitism across the whole range.

The interpolated map of pupal parasitism (Fig. 2e) does show an elevated incidence in the north-east of the range of the holly leaf-miner. This may be due partly to the absence of larval parasitism leaving more mines available to the pupal parasitoids.

miscellaneous pupal mortality

The causes of miscellaneous pupal mortality of the holly leaf-miner are unknown but may be due to attack by pupal parasitoids, predators or pathogenic fungi. We found significant spatial autocorrelation in levels of this mortality at short lags. However, there does not appear to be a simple pattern at larger scales.

successful emergence

Relatively high levels of successful emergence were found in the east of the range of the holly leaf-miner. The weak negative relationship between maximum levels of successful emergence and local population densities is somewhat surprising. Intuitively, one might expect sites supporting high population densities to have a higher per capita rate of successful emergence, indicative of their suitability as habitats for the holly leaf-miner. However, the converse seems to be true. All sites supporting high densities of leaf-miners have relatively high rates of mortality (per mine) (Fig. 4). However, the density of successful emergences (the number of adults produced at each site) is still generally higher at these sites (Figs 2g & 6).


The population dynamics of the holly leaf-miner across Europe are complex and there is still some way to go before the spatial structure of its geographical range is properly understood. No single source of mortality appears to be responsible for limiting population numbers across the whole range. At any site, the mortality that a population suffers is the sum of largely independent yet spatially structured components. These components may be, to some extent, determined by the environmental variables we used but it has not been possible to separate these environmental effects from the spatial structure they share with the rates themselves, so that other abiotic or biotic factors cannot be ruled out. This should not be surprising, yet it highlights the difficulties that face extensive surveys of this nature.

Our picture of holly leaf-miner population dynamics across the range, although better than for most animal species, is, of course, not complete. The results reported are necessarily a ‘snap-shot’ of a single year and one would expect levels of oviposition, mortality and successful emergence to exhibit temporal fluctuations. However, there is evidence that, at least in broad terms, our findings provide a realistic picture of the patterns of demographic rates across the range. For example, surveys of holly leaf-miner populations at smaller scales have shown that relative levels of densities and mortality components are consistent between years (Valladares & Lawton 1991; Brewer 2000). Similarly, the absence of C. gemma from the north-eastern part of the range is reasonably well known and documented (T. Hofsvang, personal communication, Hansson 1985).

There are currently no data on the longevity, dispersal and potential agents of mortality of adults of the holly leaf-miner. This stage necessarily represents a small proportion of the total lifecycle as the availability of leaves for oviposition is limited to at most only a few weeks in the year. However, it should not be assumed automatically that this means that the stage is unimportant in determining population densities. In a review of leaf-miner population dynamics, Auerbach et al. (1995) note that host–plant phenology and adult fecundity may have important consequences for population sizes. These are difficult parameters to measure, although there may be some value in measuring oviposition rates for two successive generations (the one in which leaf-miner densities and mortalities are recorded and the subsequent one). This would at least provide a measure of the mean number of eggs produced from each successful emergence and hence an estimate of the effects of all of these factors considered together.

Another potentially important factor that might influence leaf-miner densities and demographic rate components is variation in the local density of holly trees across the range. While it is unlikely that the effects of emigration and immigration of adult leaf-miners would be important across the broad spatial scale of the study, local movements of adults could have an important effect at smaller scales. One might expect that in regions where holly trees are sparse, densities of leaf-miners would be lower as a result of ‘metapopulation’ processes. Large variation in the density of holly trees across the range was strongly evident during the survey (although it could not be measured readily) but there did not appear to be any consistent relationship between this variation and leaf-miner densities. For example, in Norway low holly density is accompanied by high leaf-miner density, whereas in parts of Germany low holly density was often accompanied by low leaf-miner density.

Despite these considerations, the results of this survey are encouraging. The strong spatial structure that the holly leaf-miner exhibits in a number of components of its demographic rates suggest that it may indeed be possible to characterize range structure in such terms. Testing the generality of these results for most other taxa may be more difficult, as the holly leaf-miner is an unusually amenable system for studies of this nature.


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was supported by the Natural Environmental Research Council (grant GR9/03449/) and the University of Sheffield. We would like to thank Rob and Kerry Briers, Sian Gaston, Trond Hofsvang, Jens Rolff, Walter Wimmer and, most especially, John Marçal for assistance with the fieldwork.


  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Auerbach, M.J., Connor, E.F. & Mopper, S. (1995) Minor miners and major miners: population dynamics of leaf-mining insects. Population Dynamics: New Approaches and Synthesis (eds N.Cappuccino & P.W.Price), pp. 83110. Academic Press, San Diego.
  • Bellows, T.S., Jr Van Driesche, , R.G. & Elkinton, , J.S. (1992) Life-table construction and analysis in the evaluation of natural enemies. Annual Review of Entomology, 37, 587614.
  • Bradford, M.J. Taylor, , G.C. & Allan, , J.A. (1997) Empirical review of coho salmon smolt abundance and the prediction of smolt production at the regional level. Transactions of the American Fisheries Society, 126, 4964.
  • Brewer, , A.M. (2000) Interactions between demographic rates, population density and the environment − the spatial structure of the range of the holly leaf-miner. PhD Thesis, University of Sheffield.
  • Brewer, , A.M. & Gaston, , K.J. (2002) The geographic range structure of the holly leaf-miner I: population density. Journal of Animal Ecology, 71, 99111.
  • Brown, , J.H. (1984) On the relationship between abundance and distribution of species. American Naturalist, 124, 255279.
  • Brown, , J.H. (1995) Macroecology. University of Chicago Press, Chicago.
  • Brown, , J.H. & Lomolino, , M.V. (1998) Biogeography, 2nd edn. Sinauer Associates, Sunderland, Massachusetts.
  • Brown, J.H. Mehlman, , D.W. & Stevens, , G.C. (1995) Spatial variation in abundance. Ecology, 76, 20282043.
  • Cameron, , E. (1939) The holly leaf miner (Phytomyza ilicis, Curtis) and its parasites. Bulletin of Entomological Research, 30, 173208.
  • Cliff, , A.D. & Ord, , J.K. (1973) Spatial Autocorrelation. Pion, London.
  • Collett, , D. (1991) Modelling Binary Data. Chapman & Hall, London.
  • Crawley, , M.J. (1993) glim for Ecologists. Blackwell Scientific, London.
  • Dunn, P.O. Thusius, K.J. Kimber, , K. & Winkler, , D.W. (2000) Geographic and ecological variation in clutch size of tree swallows. Auk, 117, 15221.
  • Environmental Systems Research Institute, Inc. (1997a) ArcInfo GIS, Version 7·1·2. Redlands, California USA (
  • Environmental Systems Research Institute, Inc. (1997b) ArcView GIS, Version 3·0. Redlands, California, USA (
  • Fleming, I.A. & Gross, M.R. (1990) Latitudinal clines: a trade-off between egg number and size in Pacific Salmon. Ecology, 71, 111.
  • Gaston, K.J. (1994) Rarity. Chapman & Hall, London.
  • Hansson, C. (1985) Taxonomy and biology of the Palearctic species of Chrysocharis Förster, 1856 (Hymenoptera: Eulophidae). Entomologica Scandinavica Supplement, 26.
  • Heads, P.A. & Lawton, J.H. (1983) Studies on the natural enemy complex of the holly leaf miner: the effects of scale on the detection of aggregative responses and the implications for biological control. Oikos, 40, 267276.
  • Hendricks, P. (1997) Geographical trends in clutch size: a range-wide relationship with laying date in American pipits. Auk, 114, 773778.
  • Hoffmann, A.A. & Blows, M.W. (1994) Species borders: ecological and evolutionary perspectives. Trends in Ecology and Evolution, 9, 223227.
  • Holt, R.D. & Keitt, T.H. (2000) Alternative causes for range limits: a metapopulation perspective. Ecology Letters, 3, 4147.
  • Holt, R.D., Lawton, J.H., Gaston, K.J. & Blackburn, T.M. (1997) On the relationship between range size and local abundance: back to basics. Oikos, 78, 183190.
  • Hultén, E. & Fries, M. (1986) Atlas of North European Vascular Plants North of the Tropic of Cancer. Koeltz Scientific Books, Königstein.
  • Koenig, W.D. (1986) Geographical ecology of clutch size variation in North American woodpeckers. Condor, 88, 499504.
  • Lawton, J.H. (1993) Range, population abundance and conservation. Trends in Ecology and Evolution, 8, 409413.
  • Legendre, P. (1993) Spatial autocorrelation: trouble or new paradigm? Ecology, 74, 16591673.
  • Legendre, P. & Fortin, M.-J. (1989) Spatial pattern and ecological analysis. Vegetatio, 80, 107138.
  • Legendre, P. & Legendre, L. (1998) Numerical Ecology. Developments in Environmental Modelling, 20, 2nd English edn. Elsevier, Amsterdam.
  • Lewis, T. & Taylor, L.R. (1967) Introduction to Experimental Ecology. Academic Press, London.
  • MacArthur, R.H. (1972) Geographical Ecology: Patterns in the Distribution of Species. Harper & Row, New York.
  • Maurer, B.A. & Brown, J.H. (1989) Distributional consequences of spatial variation in local demographic processes. Annales Zoologici Fennici, 26, 121131.
  • McGeoch, M.A. & Gaston, K.J. (2000) Edge effects on the prevalence and mortality factors of Phytomyza ilicis (Diptera, Agromyzidae) in a suburban woodland. Ecology Letters, 3, 2329.
  • New, M.G., Hulme, M. & Jones, P.D. (2000) Representing twentieth-century space-time climate variability. Part II. Development of 1901–96 monthly grids of terrestrial surface climate. Journal of Climate, 13, 22172238.
  • Peakall, D.B. (1970) The eastern bluebird: its breeding season, clutch size, and nesting success. Living Bird, 9, 239256.
  • Peterken, G.F. & Lloyd, P.S. (1967) Biological flora of the British Isles: Ilex aquifolium L. Journal of Ecology, 55, 841858.
  • Price, J., Droege, S. & Price, A. (1995) The Summer Atlas of North American Birds. Academic Press, London.
  • Randall, M.G.M. (1982) The dynamics of an insect population throughout its altitudinal distribution: Coleophora alticolella (Lepidoptera) in northern England. Journal of Animal Ecology, 51, 9931016.
  • Randolph, S.E. (1994) Population dynamics and density-dependent seasonal mortality indices of the tick Rhipicephalus appendiculatus in eastern and southern Africa. Medical and Veterinary Entomology, 8, 351368.
  • Ricklefs, R.E. (1972) Latitudinal variation in breeding productivity of the rough-winged swallow. Auk, 89, 826836.
  • Rogers, D. (1979) Tsetse population dynamics and distribution: a new analytical approach. Journal of Animal Ecology, 48, 825849.
  • Rogers, D.J. & Randolph, S.E. (1986) Distribution and abundance of tsetse flies (Glossina spp.). Journal of Animal Ecology, 55, 10071025.
  • Root, T. (1988) Atlas of Wintering North American Birds. University of Chicago Press, Chicago.
  • Sanz, J.J. (1997) Geographic variation in breeding parameters of the pied flycatcher Ficedula hypoleuca. Ibis, 139, 107114.
  • Taylor, L.R., Woiwod, I.P. & Perry, J.M. (1980) Variance and the large scale spatial stability of aphids, moths and birds. Journal of Animal Ecology, 49, 831854.
  • Valladares, G. & Lawton, J.H. (1991) Host-plant selection in the holly leaf miner: does mother know best? Journal of Animal Ecology, 60, 227240.
  • Whittaker, J.B. (1971) Population changes in Neophilaenus lineatus (L.) (Homoptera: Cercopidae) in different parts of its range. Journal of Animal Ecology, 40, 425443.
  • Yom-tov, Y., Christie, M.I. & Iglesias, G.J. (1994) Clutch size in passerines of southern South America. Condor, 96, 170177.