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

The population of common guillemots Uria aalge on Skomer Island, Wales has been monitored since 1963, and in the last 30 yr has increased at an almost constant rate of 5% yr−1. A previous attempt to model the population based on intrinsic demographic parameters estimated over just five years failed to explain the observed population increase, probably because the estimate of juvenile survival was too low. This raised the possibility that immigration fuelled the population increase. Here we use > 30 yr of detailed field observations to re-estimate key population parameters (productivity, adult survival and juvenile survival) in order to model the population. We show that the observed rate of increase can be explained by these intrinsic parameters, and we therefore conclude that immigration is not necessary to generate the observed population growth.

Understanding the processes that determine population size is central to the conservation of vulnerable species and to ecosystem management. Fluctuations in populations of top-predators such as seabirds can be indicative of conditions over large spatial and temporal scales, thus making them an important indicator of environmental change (Croxall and Rothery 1991). The common guillemot Uria aalge is one of the most abundant seabird species in the North Atlantic (Wilson et al. 2004) and several breeding populations in Britain and Europe have been closely monitored since the 1960s (Birkhead and Ashcroft 1975, Harris and Wanless 1988, Swann et al. 1989, Sandvik et al. 2005).

Population trajectories of individual guillemot colonies have varied dramatically depending upon their location. Colonies in the Norwegian and Barents Seas for example, shrank by 88% from 1964 to 2006, a decline attributed to drowning in fishing nets, hunting of breeding adults and food shortages (Barrett et al. 2006). Some Scottish popu lations have also recently declined (Kokko et al. 2004), the decrease being attributed to very low productivity caused by a reduction in the energetic value of prey fed to chicks (Wanless et al. 2005). In contrast, guillemot populations in south-west England and Wales have increased over the same period (Mavor et al. 2006).

This geographical variation in guillemot population trajectory may be linked to a geographical variation in both the species provisioned to chicks and in fishery type. In the south of the British Isles sprats Sprattus sprattus make up a large proportion of chicks’ diets (Harris and Hislop 1978, Hatchwell 1991), whereas guillemots breeding in northern Britain and the North Sea rely heavily on sandeels Ammodytes marinus (Pearson 1968, Daan et al. 1990, Monaghan et al. 1992). The sandeel fishery in Shetland that targets small species and individuals, had a very strong negative impact on guillemot populations (Mavor et al. 2004, 2005), because of a direct reduction in prey and a lack of adequate alternative prey (Monaghan et al. 1992, Furness and Tasker 2000). In contrast, in the Celtic Sea surrounding Skomer, there is evidence of a shift in the size structure of the fish community resulting in an increase in the abundance of small fish that are potential guillemot prey (Blanchard et al. 2005). This is thought to have arisen from the indirect effects of fishing because reductions in the abundance of larger predatory fish can cause an increase in their smaller prey (Shin et al. 2005, Blanchard et al. 2009).

Outside the breeding season the availability of prey-fish is hard to quantify because monitoring schemes may not be targeted at relevant fish species, and the foraging location of seabirds is often poorly understood (Cairns 1992, Cook and Reeves 1993), especially in species such as guillemots that range widely outside the breeding season (Votier et al. 2008). Together, these features mean that it is often difficult to ascertain the factors responsible for observed seabird population changes (Reynolds et al. 2011).

The difficulty in understanding the factors underlying population changes in seabirds is exacerbated by their ‘slow’ life history, which makes estimation of population para meters extremely challenging. Guillemots are long-lived and adult birds typically have a survival rate of > 90% (Hatchwell and Birkhead 1991, Reynolds et al. 2011). They have a two-stage recruitment process where immature birds return to the breeding colony aged 3–5, and begin to breed at the earliest at age 4 (Harris et al. 1994), although the majority begin to breed at age 6 (Lindner 2000). Because of this complicated age-related recruitment (Votier et al. 2005) it is possible to model guillemot populations only following intensive field work to estimate productivity and adult and juvenile survival.

A previous attempt to model population growth on Skomer Island, Wales, using intrinsic demographic para meters collected over the relatively short time span of 5 yr failed to explain the observed population change, probably because the estimate of juvenile survival was too low (Hatchwell and Birkhead 1991). In that study, none of the > 1000 chicks that had been ringed on Skomer in the preceding few years had recruited into the breeding population, so juvenile survival was based on the only estimates that were available from over 10 yr earlier in 1972–1975 (Birkhead and Hudson 1977), before the population had started to increase. Furthermore, the great majority of guillemots on Skomer and elsewhere in the region are not ringed, so it is very difficult to determine whether observed population changes are fuelled by intrinsic demographic processes or by immigration/emigration. In some instances changes in guillemot numbers have been so rapid they could not possibly be explained by an increase in breeding success or survival, and instead it was suggested that massive movements between colonies offered a possible explanation (Tuck 1961, Southern et al. 1965). However, changes in population size were poorly quantified in those studies and there was no evidence for movement between colonies. On the other hand, studies of other alcid species provide convincing evidence that immigration and emigration may be important for population dynamics. For example, immigration from nearby colonies was shown to be important for the rapid growth of Atlantic puffin Fratercula arctica (Harris 1976) and black guillemot Cepphus grille (Petersen 1981) colonies. Moreover, there is some movement of immature guillemots between Skomer and other colonies (Lindner 2000), and the previous lack of fit between the observed and projected population size raised the possibility of substantial net immigration into the Skomer Island population. In addition, Hatchwell and Birkhead (1991) estimated productivity and adult survival from data collected over a short time-span, so these estimated parameters could also have been inaccurate.

The aims of the present study were: 1) to use > 30 yr of data to estimate demographic parameters and produce a model of the common guillemot population on Skomer Island. 2) To determine whether the observed population increase can be explained by the intrinsic demographic parameter estimates or whether net immigration must have occurred to explain the population increase.


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

Study area and data collection

Photographs of guillemot breeding habitat on Skomer Island, Wales (51°40’N, 05°15′W) provide clear anecdotal evidence that guillemot numbers were substantially larger in the 1930s than has been recorded since (Birkhead and Ashcroft 1975). Since 1963, guillemot numbers have been monitored using a ‘whole island’ count where the numbers of individuals present on ledges at colonies around the entire island are counted. Since 1973 these counts have been supplemented by ‘study plot’ counts carried out during the time between peak laying and peak fledging (usually early June) at three large colonies: Bull Hole, South Stream and High Cliff (comprising on average 27% ± 0.01 SE of the whole population). The number of guillemots present on ledges at these three colonies is counted 5–15 times each year within a 21-d period. Since 1973, the whole island count includes the mean values for the study plot counts, but most other colonies are counted just once in a given year.

Since 1985, breeding adult guillemots and guillemot chicks have been ringed annually at sub-colonies around the island (for numbers see Table 1). The birds are ringed with a metal British Trust for Ornithology (BTO) ring and unique colour rings that allow the birds to be subsequently identified using a ×20–60 telescope from a distance of up to 300 m in good conditions. From 1991–2011, a sub- colony of guillemots has been observed, and the timing of breeding and productivity of colour-ringed birds within this plot has been recorded using standard protocols (Birkhead and Nettleship 1980).

Table 1.  Raw data and estimated parameters, where available, from 1985 to 2011.
YearWhole island countMean of study plot countsNo. adults ringedNo. chicks ringedEstimated adult survivalEstimated cohort survival to age 3Productivity (chicks pair21)
198561811351.168150 0.28 
198658351300.111264 0.550.79
198761921434.412345 0.550.79
19886532148824317 0.39 
198955561271.610308 0.28 
199060511514.37313 0.54 
199175161911.11325 0.610.88
199810 8993295.713180.940.410.76
199912 1253728.0152970.970.490.90
200013 8523867.1233020.920.420.86
200114 2814101.672731.000.460.82
200214 4344128.492800.900.210.89
200314 6764074.0243020.960.410.90
200519 7114897.7172970.92 0.78
200616 9745038.532930.95 0.82
200717 5445140.3292990.87 0.80
200817 0885078.0102700.90 0.75
200919 5126458.3102920.94 0.84
201019 9626400.4372770.90 0.76
201121 6886360.5572420.84 0.87

Stage-structured population models

We constructed stage-structured matrix models (Lefkovitch 1965, Caswell 2001, Cooch et al. 2012) to project the population of adult female guillemots on Skomer. This is similar to an age-structured Leslie matrix model (Leslie 1945) but differs because individuals of different ages that have not yet recruited into the breeding population can be grouped together in a single category. This stage-based matrix is constructed from underlying parameter estimates of newborn survival (So), adult survival (Sa), the probability of recruiting at a stage i (ai), and the probability of repro ducing at a given stage (Bi, Cooch et al. 2012). Our stage- structured model (A, Fig. 1) is relatively simple because it includes only 3 stages, and assumes that all juveniles recruit at age 6 and that all recruits breed every year.


Figure 1. Stage-structured matrix model. ‘B’ is productivity (the number of female offspring produced per adult female guillemot), ‘So’ juvenile survival to age 3. ‘Sa' is adult survival, a2 is the probability of recruiting into the breeding population from the juvenile stage.

Download figure to PowerPoint

  • image

Census counts of known numbers of breeding pairs on Skomer showed that during the census period the pair:individual ratio was 0.68 (Hatchwell and Birkhead 1991), a ratio that appears stable at a range of densities (Hatchwell 1988). We used this conversion factor to estimate the number of breeding females present on Skomer from whole island counts assuming a 1:1 sex ratio (Lindner 2000). From 1963–1980, the population of guillemots on Skomer was stable, but from 1980 onwards it increased at an almost constant rate of 5% yr21 (Fig. 2). Because of this clear change in population trajectory we modelled the population using two stage-structured models. The first (model 1: 1963–1980) used estimates of productivity and adult survival from Birkhead and Hudson (1977). The second (model 2: 1980–2011) used parameters estimated as follows. Productivity – the mean breeding success per pair was calculated (Table 1) and halved to give an estimate of the number of female offspring produced per female breeder. Adult survival – we defined ‘adult’ as birds of age ≥ 6 yr because 6 is the modal breeding age (Lindner 2000). We used ProgramMARK (White and Burnham 1999) to estimate a mean survival rate from capture–mark–recapture data of adults from 1992 to 2010. We allowed the re- sighting probability ‘p’ to vary with time, but we constrained the survival rate to a single parameter because this was necessary to construct a stage-structured matrix model of the guillemot population. Because there is some evidence that the disturbance associated with capture and ringing may cause birds to move breeding site (Birkhead and Hudson 1977), birds ringed as adults were included only if they had been seen at least once in a year following first ringing. Similarly, individuals ringed as chicks entered the ‘adult’ dataset only if they were seen in at least two years aged 6+.


Figure 2. The number of guillemots (± SD) on Skomer from 1963–2011. Whole island and total study plot counts are shown. The markers indicate the years modelled by the two stage- structured matrix models.

Download figure to PowerPoint

Estimates of the probability of ‘newborn’ or chick survival are complicated by the guillemots’ slow life history and two-stage recruitment process. Juvenile guillemots return to colonies only at ≥ 2 yr old (Birkhead and Hudson 1977). An increasing number return each successive year, while some birds will be lost to mortality and others will progressively join the breeding population, perhaps at non- natal colonies (Lindner 2000). Therefore there is no single time to census juveniles. Instead, for both time periods we calculated juvenile survival to age 3. For 1963–1980 we used estimates of juvenile survival from Birkhead and Hudson (1977), and for 1980–2011 we used the proportion of a cohort ringed in year n that were sighted from year n+ 3 – year n+7 (from 1985 to 2004). We used this seven- year window (from time of ringing), to avoid negatively biasing survival estimates of more recent cohorts (negligible ringing occurred before 1985). We therefore replaced ‘newborn’ survival with ‘juvenile survival to age 3’ in our stage structured models and assumed adult survival from then on. The probability of recruitment at stage 2 (a2) was 1/3 because the juvenile stage included only three, four and five year old guillemots and we would expect only the five-year-old guillemots to recruit. We ran the models and used the number of guillemots in the adult stage to predict of the number of adult female guillemots on the island.


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

Population changes

The population size of guillemots on Skomer remained relatively constant from 1963 to 1980, but from 1980 onwards increased at a rate of ca 5% yr21 (Fig. 2). Counts at each of the three study plots are closely correlated to the residual whole island count (whole island–study plot) from 1980–2011 (Pearson's product moment correlations, South Stream: r = 0.94, DF = 30, p < 0.001, High Cliff: r = 0.97, DF = 30, p < 0.001, Bull Hole: r = 0.97, DF = 30, p < 0.001). The two methods of measuring population size therefore give very similar results.

Population projection models

Two stage-structured matrix models were constructed, one for the period 1963–1980 when numbers were relatively constant (model 1, Fig. 2), and one from 1980–2011 (model 2, Fig. 2). The following parameter estimates were used. Productivity – pre-1980 productivity (female chicks only) was estimated to be 0.359 (Birkhead and Hudson 1977). The available productivity estimates post-1980 are shown in Table 1. The mean observed productivity from 1980 to 2011 was calculated and divided by 2 to give an estimate of the mean number of female offspring produced per female over this period of 0.410. Adult survival – pre-1980 adult survival was estimated to be 0.915 (Birkhead and Hudson 1977). Post-1980 adult survival was modelled using ProgramMARK. A bootstrapped estimate of the variance inflation factor c was 1.02, indicating a good model fit. The mean estimate of adult survival across years was 0.930 ± 0.01 SE. Juvenile survival to age 3 – pre-1980, Birkhead and Hudson (1977) estimated juvenile survival to age 3 based on known age birds viewed at clubs (chicks were not individually identifiable) to be 0.246. Post-1980: using re-sighting data as detailed in the methods, mean cohort survival to age 3 from 1985 to 2004 was 0.430 (Table 1). The probability of recruiting (a2) for both models was 1/3. The stage-based models are therefore structured as follows for 1963–1980 (B) and for 1980–2011 (C).

  • image

As starting points for the two stage-structured models we converted whole island counts into counts of females using the conversion factor of 0.68 females:1 individual (Birkhead 1978, Hatchwell and Birkhead 1991). This meant that the starting point was 3302 adult females in 1963 and 3157 adult females in 1980. Parameter estimates and population growth rates for each of the matrices are shown in Table 2.

Table 2.  Population growth rate (λ), parameter estimates and measures of elasticity rates for each parameter in the two stage-structured matrix models.
Model 1: 1963–19800.987Productivity (B)0.3590.058
  Adult survival (Sa)0.9150.884
  Juvenile survival to age 3 (So)0.2640.058
  Probability of recruiting at stage 2 (a2)0.30.011
Model 2: 1980–20111.051Productivity (B)0.4100.082
  Adult survival (Sa)0.9300.835
  Juvenile survival to age 3 (So)0.4300.082
  Probability of recruiting at stage 2 (a2)0.30.023

We calculated the geometric mean of the observed and predicted population sizes for both time periods. This gives a measure of the median population size for any point in time (Morris and Doak 2002). For the period 1963–1980 the geometric mean of the observed values was −0.003, indicating a population growth rate of 0.997, a close fit to the model prediction of 0.987. In the second stage-structured model (1980–2011), the observed geo metric mean was 0.050, indicating a population growth rate of 1.050 which was very close to the estimate from the model of 1.051 (Fig. 3).


Figure 3. Shows predicted values from model 2 (white dots) and observed numbers of guillemots (black dots) in whole island counts on Skomer from 1980–2011 (using the correction factor of 0.68:1 to estimate the numbers of females present).

Download figure to PowerPoint

Elasticity of parameter estimates

We calculated the elasticity, defined as the relative impact of a demographic parameter on population growth rate (λ, Morris and Doak 2002), for each parameter in each of our stage-structured models (Table 2). Adult survival had the greatest relative effect on λ (model 1: 0.884; model 2: 0.835), whereas productivity, juvenile survival to age 3 and probability of recruiting at stage 2 had relatively weak effects on λ (Table 2).


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

We modelled the population of common guillemots on Skomer Island using two stage-structured matrix population models. The first (model 1: 1963–1980), used estimates of annual productivity (0.718), adult survival (0.915), and juvenile survival to age 3 (0.246) from Birkhead and Hudson (1977) and predicted a stable population. The second (model 2: 1980–2011), used estimates of productivity (0.820), adult survival (0.930), and juvenile survival to age 3 (0.430) based on > 30 yr of data collected since 1980, and projected a population increasing by 5% yr21. Both models were a close fit to the observed data. These findings suggest that the dynamics of the guillemot popu lation on Skomer, and in particular the observed increase in numbers over the past 30 yr, can be adequately explained by intrinsic estimates of adult survival, juvenile survival to age 3, and productivity, without the need for net immigration into the population.

The proportional effect of each demographic parameter on population growth rate (λ) was also estimated. In both stage-structured matrix models, adult survival had the greatest relative effect on λ (model 1: 0.884; model 2: 0.835), whereas productivity and juvenile survival had very similar, lower, elasticities of ca 0.07 (Table 2). The likelihood of recruiting into the breeding population, for both models had very small elasticities of < 0.03. This means that for both models a proportional increase in adult survival would have more than 10 times the effect of the same proportional increase in any of the other parameters on λ. Between the two periods of time modelled, there was a proportional increase in adult survival of 1.6%, a proportional increase in juvenile survival to age 3 of 57%, and a proportional increase in productivity of 14.2%. The probability of recruitment at a given time was the same for both models (1/3). By substituting the parameters from model 2 into model 1 one by one and calculating λ we find that 63% of the observed population increase is due to this large increase in juvenile survival to age 3, 24% due to the increase in adult survival and just 13% to the increase in productivity between the two periods. Given the clear evidence that in general bird populations are limited by food availability, it is possible that the increase in productivity, adult survival, juvenile survival and overall numbers between the two time periods modelled, is linked in some way to the increased abundance of potential guillemot prey (fish < 20 cm) in the Celtic sea (Blanchard et al. 2005).

By the nature of matrix projection models, the para meters used are crucially important in terms of the predictions of the model. How confident are we in our parameter estimates? First, we consider productivity. Estimates from both periods were from detailed, almost daily, observations of colonies throughout the breeding season based on the protocol by Birkhead and Nettleship (1980) and are therefore likely to be both accurate, and consistent between the two time periods presented. Second, although the estimates of adult survival for each model are calculated in different ways (pre-1980 by percentage observed survival rather than using capture mark recapture techniques) the former is nevertheless likely to be an accurate estimate, since in the 1970s the colonies were very much smaller and less dense so that it was much less likely that individually colour marked birds could be overlooked (and assumed not to have survived) than in the current very dense colonies on Skomer. This difference in the visibility of marked birds is a consequence of population growth through the packing of existing colonies rather than the founding of many new ones. One of the main limitations of capture mark recapture analyses is that ‘local’ survival is estimated – i.e. emigration and survival is not distinguished (White and Burnham 1999). However for the current purpose of modelling the observed population size based on the observed parameters, an estimate of ‘local’ survival is appropriate. Site philopatry for breeding adults is very high (Birkhead 1977, Harris et al. 1996) with > 90% of surviving breeders returning to the same part of the cliff, so it is unlikely that many of the colour ringed breeders would have been missed. We therefore feel that our estimates of adult survival are accurate for both periods modelled. They are also consistent with earlier analyses of data from the same study population (Votier et al. 2005).

In contrast, ‘local’ survival (or recruitment) of juvenile guillemots is much more difficult to measure (Birkhead and Hudson 1977, Hatchwell and Birkhead 1991, Harris et al. 1992, Lindner 2000, Votier et al. 2008). A previous attempt to model the guillemot population on Skomer (Hatchwell and Birkhead 1991) predicted population growth that was substantially lower than that observed, probably because the estimate of juvenile survival to age 5 (20.6%) was too low. When a higher estimate (41.1%) was used instead the model more accurately fitted the observed values. However, this value was based on a composite estimate of survival to age 5 from ringing recoveries in North America (Lack 1951, Birkhead and Hudson 1977), and at that time was regarded as unfeasibly high. It is possible that the estimate of juvenile survival for the period 1963–1980 is an underestimate since the chicks ringed in the early 1970s were not individually identifiable, and this meant that the estimates had to be based on maximum numbers seen on clubs in a day. A recent study by Crespin et al. (2006) estimated survival to age 2 of guillemots on the Isle of May to be 58%, and this combined with an immature survival rate lower than that of adults, gave an estimate of survival to age 4 of 42.9%. Using an adult survival rate between ages 3 and 4 we would calculate survival to age 4 to be similar for Skomer guille mots at 39.9%. Harris et al. (2007) suggest that a lower than adult ‘local’ survival rate from age 2 is due to birds either emigrating from the population, or ‘otherwise becoming unobservable’. Several authors have noted that the probability of resighting breeding birds is much lower than the probability of resighting non-breeders due to the crowded nature of breeding colonies (Lindner 2000, Crespin et al. 2006, Harris et al. 2007). It is very difficult to determine whether birds that are not sighted after age 3 have died, permanently emigrated from the population, or have simply recruited into the breeding population at locations where they cannot be seen by observers. Similarly, despite our finding that observed population growth can be accounted for by intrinsic population parameters, it is very difficult to rule out the possibility of immigration into the study population.

Of the 191 guillemots sighted on Skomer that were ringed as chicks elsewhere, only 9% (18) were ever observed breeding on Skomer. These birds comprise < 2% of ringed recruits subsequently observed breeding on Skomer. Fourteen immigrants were from Great Saltee, 60 km to the north-west, the closest colony where significant ringing effort occurs, two were from Ceann Ousdale in the north of Scotland, and there was one bird from each of Bardsey Island, Stora Karlsö in the Baltic, the Isle of Canna, and Sanda Island. Of the birds ringed on Skomer that have been sighted elsewhere, one has been reported breeding on Lundy Island (60 km south-east) and one on Skokholm Island (3 km south). Because the projections from the population models fit the observed data so well, we regard it as very unlikely that there has been substantial net immigration into the breeding population on Skomer since the population increase is adequately explained by intrinsic parameters. Furthermore, our results also indicate that natal and breeding dispersal between colonies must be very limited, because such dispersal would have the effect of reducing measured survival rates, thereby generating a mismatch between observed and predicted population growth. It is also interesting to note that previous increases in alcid populations shown to be driven by immigration were associated with decreasing populations at nearby colonies (Harris 1976, Petersen 1981), in this case other populations in colonies bordering the Irish Sea have also tended to increase over the same time period as that observed on Skomer (Mavor et al. 2006), so there is no obvious local source of immigrants.

In conclusion, it seems that the population increase of guillemots on Skomer can be explained by intrinsic estimates of productivity, and adult and juvenile survival. It is clear that there is some limited exchange of individuals between Skomer and other guillemot populations, but it seems that the levels of immigration and emigration must be on a small scale.


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

We are grateful to many people who have been involved with the Skomer guillemot study: C. Barton, K. Bowgen, D. Boyle, K. Bradley, T. Clarke, B. Dean, P. Eady, B. Eatwell, K. Fellerman, C. Gray, R. Humpidge, H. Kirk, R. Lindner, G. McKnight, K. Munro, G. Ovenden, S. Patrick, J. Pellatt, D. Porter, S. Price, H. Watson, L. Wilson, S. Wilson and R. Woodburn, as well as the wardens on Skomer: M. Alexander, J. Brown, P. Corkhill, J. and D. Milborrow, S. Smith, S. and A. Sutcliffe, and especially the current warden C. Taylor. We are grateful to the Wildlife Trust of South and West Wales for allowing us to work on Skomer, and the crew of the Dale Princess for their help. We are grateful to the Centre for Environment Fisheries and Aquaculture Sciences (CEFAS) for provision of the English Celtic Sea groundfish survey data. We are grateful to J. Biggins and S. Votier for statistical advice, and to M. Harris and S. Wanless for valuable discussion. We thank M. Alexander and T. Hellawell for their support and encouragement. This work was funded by the Natural Environment Research council, Countryside Council for Wales, Wildlife Trust of South and West Wales and Joint Nature Conservation Committee.


  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Barrett R. T., Lorentsen S.-H. and Anker-Nilssen T. 2006. The status of breeding seabirds in mainland Norway. Atl. Seabirds 8: 97126.
  • Birkhead T. R. 1977. The effect of habitat and density on breeding success in the common guillemot Uria aalge. J. Anim. Ecol. 46: 751764.
  • Birkhead T. R. 1978. Attendance patterns of guillemots Uria aalge at breeding colonies on Skomer. Ibis 120: 219229.
  • Birkhead T. R. and Ashcroft R. E. 1975. Auk numbers on Skomer Island. Nat. Wales 14: 223233.
  • Birkhead T. R. and Hudson P. J. 1977. Population parameters for the common guillemot Uria aalge. Ornis Scand. 8: 145154.
  • Birkhead T. R. and Nettleship D. N. 1980. Census methods for murres, Uria spp.: a unified approach. Occasional Paper no. 43, Canadian Wildlife Service, Ottawa.
  • Blanchard J. L., Dulvy N. K., Jennings S., Ellis J. E., Pinnegar J. K., Tidd A. and Kell L. T. 2005. Do climate and fishing influence size-based indicators of Celtic Sea fish community structure? ICES J. Mar. Sci. 62: 405411.
  • Blanchard J. L., Jennings S., Law R., Castle M. D., McCloghrie P., Rochet M.-J. and Benoît E. 2009. How does abundance scale with body size in coupled size-structured food webs? J. Anim. Ecol. 78: 270280.
  • Cairns D. K. 1992. Bridging the gap between ornithology and fisheries science: use of seabird data in stock assessment models. Condor 94: 811824.
  • Caswell H. 2001. Matrix population models construction, analysis, and interpretation, 2nd ed. – Sinauer.
  • Cooch E. G., Cam E. and Caswell H. 2012. Incorporating ‘recruitment’ in matrix projection models: estimation, parameters, and the influence of model structure. J. Ornithol. 152: 585595.
  • Cook R. M. and Reeves S. A. 1993. Assessment of North Sea industrial fish stocks with incomplete catch-at-age data. ICES J. Mar. Sci. 50: 425434.
  • Crespin L., Harris M. P., Lebreton J.-P., Frederiksen M. and Wanless S. 2006. Recruitment to a seabird population depends on environmental factors and on population size. J. Anim. Ecol. 75: 228238.
  • Croxall J. P. and Rothery P. 1991. Population regulation of seabirds: implications of their demography for conservation. – In: Perris C. M., Leberton J.-D. and Hirons G. J. M. (eds), Bird population studies. Oxford Univ. Press, pp. 272296.
  • Daan N., Bromley P. J., Hislop J. R. G. and Nielsen N. A. 1990. Ecology of North Sea fish. Neth. J. Sea Res. 26: 343386.
  • Furness R. W. and Tasker M. L. 2000. Seabird-fishery inter actions: quantifying the sensitivity of seabirds to reductions in sandeel abundance, and identification of key areas for sensitive seabirds in the North Sea. Mar. Ecol. Prog. Ser. 202: 253264.
  • Harris M. P. 1976. Inter-colony movement of Farne Island puffins. Trans. Nat. Hist. Soc. Northumb. 42: 115118.
  • Harris M. P. and Hislop J. R. G. 1978. The food of young puffins (Fratercula arctica). J. Zool. 185: 213239.
  • Harris M. P. and Wanless S. 1988. The breeding biology of guillemots Uria aalge on the Isle of May over a six year period. Ibis 130: 172192.
  • Harris M. P., Halley D. J. and Wanless S. 1992. The post- fledging survival of young guillemots Uria aalge in relation to hatching date and growth. Ibis 134: 335339.
  • Harris M. P., Halley D. J. and Swann R. L. 1994. Age of first breeding in common murres. Auk 111: 207209.
  • Harris M. P., Wanless S. and Barton T. R. 1996. Site use and fidelity in the common guillemots Uria algae. Ibis 138: 399404.
  • Harris M. P., Frederiksen M. and Wanless S. 2007. Within- and between-year variation in the juvenile survival of common guillemots Uria aalge. Ibis 149: 472481.
  • Hatchwell B. J. 1988. Population biology and coloniality of common guillemots Uria aalge. – PhD thesis, Univ. of Sheffield, Sheffield.
  • Hatchwell B. J. 1991. An experimental study of the effects of timing of breeding on the reproductive success of common guillemots (Uria aalge). J. Anim. Ecol. 60: 721736.
  • Hatchwell B. J. and Birkhead T. R. 1991. Population dynamics of common guillemots Uria aalge on Skomer Island, Wales. Ornis Scand. 22: 5559.
  • Kokko H., Harris M. P. and Wanless S. 2004. Competition for breeding sites and site-dependent population regulation in a highly colonial seabird, the common guillemot Uria aalge. J. Anim. Ecol. 73: 367376.
  • Lack D. 1951. Population ecology in birds. – In: Horstadius S. (ed.), Proceedings of the 10th International Ornithological Congress. Almqvist and Wiksell, pp. 409448.
  • Lefkovitch L. P. 1965. The study of population growth in organisms grouped by stages. Biometrics 21: 118.
  • Leslie P. H. 1945. On the use of matrices in certain population mathematics. Biometrika 33: 183212.
  • Lindner R. 2000. The population biology of the common guillemot, Uria aalge. – PhD thesis, Univ. of Sheffield, UK.
  • Mavor R. A., Parsons M., Heubeck M. and Schmitt S. 2004. Seabird numbers and breeding success in Britain and Ireland, 2003. Peterborough, Joint Nature Conservation Committee, UK Nature Conservation no. 27.
  • Mavor R. A., Parsons M., Heubeck M. and Schmitt S. 2005. Seabird numbers and breeding success in Britain and Ireland, 2004. Peterborough, Joint Nature Conservation Committee, UK Nature Conservation, no. 29.
  • Mavor R. A., Parsons M., Heubeck M. and Schmitt S. 2006. Seabird numbers and breeding success in Britain and Ireland, 2005. Peterborough, Joint Nature Conservation Committee, UK Nature Conservation, no. 30.
  • Monaghan P., Uttley J. D. and Burns M. D. 1992. Effect of changes in food availability on reproductive effort in Arctic terns Sterna paradisaea. Ardea 80: 7181.
  • Morris W. F. and Doak D. F. 2002. Quantitative conservation biology: theory and practice of population viability analysis. – Sinauer.
  • Pearson T. H. 1968. The feeding biology of seabird species breeding on the Farne Islands, Northumberland. J. Anim. Ecol. 37: 521552.
  • Petersen A. 1981. Breeding biology and feeding ecology of black guillemots. PhD thesis, Univ. of Oxford, UK.
  • Reynolds T. J., Harris M. P., King R., Swann R. L., Jardine D. C., Frederiksen M. and Wanless S. 2011. Among-colony synchrony in the survival of common guillemots Uria aalge reflects shared wintering areas. Ibis 153: 818831.
  • Sandvik H., Erikstad K. E., Barrett R. T. and Yoccoz N. G. 2005. The effect of climate on adult survival in five species of North Atlantic seabirds. J. Anim. Ecol. 74: 817831.
  • Shin Y. J., Rochet M.-J., Jennings S., Field J. and Gislason H. 2005. Using size-based indicators to evaluate the ecosystem effects of fishing. ICES J. Mar. Sci. 62: 384396.
  • Southern H. N., Carrick R. and Potter W. G. 1965. The natural history of a population of guillemots (Uria aalge Pont). J. Anim. Ecol. 34: 649661.
  • Swann R. L., Aiton D. G., Carruthers J., Graham R. and Ramsay A. D. K. 1989. Changes in recovery and retrap patterns of guillemots ringed on Canna 1974 75. Ring. Migr. 10: 3540.
  • Tuck L. M. 1961. The murres. – Canadian Wildlife Service, no. 1.
  • Votier S. C., Hatchwell B. J., Beckerman A., McCleery R. H., Hunter F. M., Pellatt J., Trinder M. and Birkhead T. R. 2005. Oil pollution and climate have wide-scale impacts on seabird demographics. Ecol. Lett. 8: 11571164.
  • Votier S. C., Birkhead T. R., Oro D., Trinder M., Grantham M. J., Clark J. A., McCleery R. H. and Hatchwell B. J. 2008. Recruitment and survival of immature seabirds in relation to oil spills and climate variability. J. Anim. Ecol. 77: 974983.
  • Wanless S., Harris M. P., Redman P. and Speakman J. R. 2005. Low energy values of fish as a probable cause of a major seabird breeding failure in the North Sea. Mar. Ecol. Prog. Ser. 294: 18.
  • White G. C. and Burnham K. P. 1999. Programme MARK: survival estimation from populations of marked animals. Bird Study 46: 120139.
  • Wilson L. J., Daunt F. and Wanless S. 2004. Self-feeding and chick provisioning diet differ in the common guillemot Uria aalge. Ardea 92: 197208.