SEARCH

SEARCH BY CITATION

Keywords:

  • flooding;
  • godwit;
  • option selection;
  • population viability;
  • re-nesting model

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Godwits of the Limosa l. limosa race have declined throughout northern Europe because of changing agricultural practices. The relict UK population is now mostly confined to two reserves within flood-defence structures, and numbers have declined at one of these. This study diagnosed the cause of this decline and evaluated options for remedial management.
  • 2
    Re-nesting models showed that productivity varied among sites and years in relation to flooding patterns. Floods caused breeding failure by forcing godwits to nest on nearby arable fields where nest and chick survival rates were low.
  • 3
    A population model showed that flood-dependent variations in productivity were sufficient to explain the contrasting population trends at the two sites.
  • 4
    The relative merits of various options for mitigating the effect of floods on godwits were investigated using a combination of hydrological, re-nesting and population models.
  • 5
    Models assuming a closed population resulted in numbers of godwit pairs at one site, the Ouse Washes in eastern England, declining and becoming extirpated within 30 years under current conditions. Some management options improved productivity, population growth and persistence likelihood, but the chances of extirpation were still high and conservation targets would not be met.
  • 6
    Models assuming an open population showed that target populations would only be achieved within 30 years if all of the available flood mitigation options were combined. However, habitat creation outside the Ouse Washes resulted in comparable productivity and population growth at a fraction of the cost.
  • 7
    Synthesis and applications. Provision of compensatory habitat is likely to be a more parsimonious means of conserving black-tailed godwits at the Ouse Washes than flood mitigation. However, reliance on the creation of new habitat is a more risky strategy as the godwits may continue to use traditional arable fields in favour of grassland alternatives, and because their productivity on created grassland is unknown.

Introduction

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

Black-tailed godwits Limosa l. limosa L. nest on wet meadows throughout northern Europe (Beintema & Melter 1998), although their numbers have declined since the mid-1960s because of drainage of their wetland habitat and losses of nests and chicks to trampling by stock and machinery used for mowing (den Boer 1995; Kruk, Noordervliet & ter Keurs 1997; Schekkerman & Müskens 2000). The relatively small UK population declined to extirpation during the 17th century (Cadbury & Kirby 1993). Recolonization occurred at the Ouse Washes (east England) in 1952, and numbers increased rapidly to a peak of 65 pairs in 1970 (Cottier & Lea 1969) before declining to three pairs in 2004. The Nene Washes (east England) was colonized in 1971 and has increased steadily to 42 pairs in 2004. The only other sites regularly used by breeding L. l. limosa in the UK are the North Kent Marshes (south-east England) and Ribble Marshes (north-west England; five and two pairs, respectively). The restricted range and small, declining population has led to black-tailed godwit being included on the UK Red List of birds of conservation concern (Gibbons et al. 1996).

Both the Ouse and Nene Washes are floodwater storage and conveyancing structures designed to prevent inundation of surrounding farmland and property during peak river flows. Storage of floodwater during spring suppresses breeding success by displacing birds onto unsuitable habitat, reducing opportunities to replace nests lost to predation and inundating clutches (Green 1986; Green, Cadbury & Williams 1987). The incidence of spring flooding at the Ouse increased during the same period that the godwit population there declined (Green 1985; Green, Cadbury & Williams 1987) while the population at the Nene (which rarely floods in spring) was increasing. This combination of correlative and demographic evidence suggests that flooding is likely to be responsible for the decline of godwits at the Ouse Washes.

The Ouse Washes are listed as a Special Protection Area, Ramsar Site and Site of Special Scientific Interest. The UK government has a legal responsibility to restore black-tailed godwit breeding numbers at the site, and this will require measures to ameliorate the effects of flooding on godwit productivity. This could be achieved by altering components of the Ouse Washes or catchment to reduce flooding (Posford Duvivier 2000) or by creating breeding habitat on adjacent land where godwits can breed successfully during flood years. All these options are expensive, and appraising their cost-effectiveness is a prerequisite to arriving at a parsimonious conservation strategy.

This study presents a re-nesting model that estimates flood-dependent productivity from empirically generated breeding parameters. A population model was used to elucidate whether the incidence of flood-induced breeding failure was adequate to explain the contrasting population trends at the Ouse and Nene Washes. The effects of various flood-mitigation and habitat-creation options on the demography and population trends of godwits were examined using stochastic models. The cost of the options and their likelihood of meeting conservation targets were then examined and a parsimonious management strategy proposed.

Methods

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

study sites

The Ouse (N 52°28′E 0°12′) and Nene (N 52°34′E 0°4′) Washes were created during the 17th century in order to facilitate drainage of the fens for agriculture and settlement (NRA 1993). Both sites act as floodwater conveyancing channels and storage reservoirs, and comprise long, narrow strips of land bounded on each side by barrier banks. During periods when peak river flows coincide with high spring tides, water is diverted into the washes to prevent flooding of surrounding farmland and property. The floodwater is stored in the washes until flows subside, and is then released back into the channel in a controlled manner. The sites are only 24 km apart but lie within separate catchments, such that flood patterns at each are independent beyond the influences of regional rainfall and tide levels.

The land between the barrier banks is mainly unimproved neutral grassland, divided into fields by a system of drainage ditches. A large proportion of both the Ouse and Nene Washes comprises nature reserves, with management being designed to benefit breeding waders. Grazing and mowing are conducted late in the year (15 May and 15 July onwards, respectively) to minimize trampling and mechanical destruction of nests and chicks, and high water tables and shallow pools are maintained to provide invertebrate prey. Winter wheat, legumes and potatoes are grown in rotation on the surrounding arable land. For further descriptions of the sites, see Benstead et al. (1997).

estimation of breeding parameters

Data were collected at the Ouse and Nene Washes in each year between 1999 and 2003 inclusive. Nests were located by watching off-duty or disturbed birds returning to incubate (Green 1986). The small number of pairs and high search effort meant that almost all nesting attempts in each site-year were included in the study. The clutch size was recorded for all nests that survived the laying period. The length and breadth of all eggs in the clutch were measured to the nearest 0·1 mm and were weighed to the nearest 0·1 g. The hatching date of the clutch was estimated using the nomogram presented in Green (1986). Clutch completion dates were estimated by backdating observed hatching dates (using an incubation period of 23 days; Cramp & Simmons 1983) for nests that hatched or predictions from the nomogram for those that failed.

Nest fates were determined by checking their contents every 3–5 days. Daily nest survival rates were estimated using Mayfield logistic regression (Hazler 2004). Daily rates of predation and abandonment were also estimated to assess their relative contribution to overall nest failure. The fate of all nests were determined, and so there were no biases associated with treatment of nests of unknown fate (Manolis, Andersen & Cuthbert 2000). The effects of year, site and habitat (i.e. whether the nest was located on arable land or the washes) on nest survival rates were tested and survival estimates were derived from the minimal adequate model. Hatching success was estimated by raising the daily nest survival rates to the power of 26 (the combined average laying and incubation period; Cramp & Simmons 1983).

The likelihood of an egg that survived the incubation period being infertile or addled (henceforth termed unviable) was estimated using a GLM with a binomial error distribution and logit link, with the response variable being the number of eggs in a clutch that were unviable and the binomial denominator being the total number of eggs in the clutch. Effects of site, habitat and year on the likelihood of an egg being unviable were examined, and estimates of addling rates were derived from the minimal adequate model.

Nests were visited daily as their predicted hatching dates approached, and one or more chicks in each brood was fitted with 0·5-g Holohil (Ontario, Canada) or Biotrack (Wareham, UK) radio-tags prior to dispersal. The tags had an operating life of 21 days (Holohil) or 27 days (Biotrack). The tags were activated and mounted on an oval piece of black gauze (c. 1·5 × 1 cm) with super-glue. The gauze was stuck to the down over the chicks’ synsacrum and pelvic girdle with latex-based glue. Grant (2002) found no adverse effects of mounting radio-tags in this manner on the survival of Eurasian curlew Numenius arquata L. chicks.

The survival rates of tagged chicks were determined from checks every 1–3 days with a radio-receiver until the tags ceased transmitting or were recovered. Estimating survival from radio-tags fitted to chicks was complicated by the fact that 45% of tags stopped transmitting before the expiry age given by the manufacturer and were not recovered (henceforth termed as disappearing). Such fates are equivocal, as the disappearance of a tag could be the result of it either being destroyed when a chick was eaten or a tag failing for technical reasons on a living chick. A simulation model was therefore used to correct survival rates for technical tag failure. The age-dependent daily likelihood of each possible fate of a tag (recorded on a live chick, recovered on a dead one, found detached from the chick or disappeared) was estimated using GLM with a binomial error distribution and logit link function, with the estimates being bootstrapped to allow for non-independence of chicks within broods (Grant et al. 1999). These likelihoods were used to randomly assign daily fates until the chick fledged or the tag was recovered or ceased transmitting for whatever reason. The age-dependent likelihood of a tag failing on a live chick was estimated using data from lapwing Vanellus vanellus L. chicks (M. Bolton, unpublished data), and this was used to class those chicks that disappeared as either dying as a result of predation or surviving with a failed tag. The daily survival rate was calculated as the number of chicks that survived each day divided by the total number of chick days simulated. This was repeated 999 times and the average, 2·5th and 97·5th percentiles were taken as the daily survival rate and the lower and upper confidence limits, respectively. The simulations were conducted using a program written in Microsoft Visual Basic 6·0.

estimating productivity

Simulation models allowing for re-nesting (Beintema & Müskens 1987; Green 1988; Green et al. 1997) were used to estimate godwit productivity. Females were randomly allocated a start date (the date of completion of their first clutch) and a stop date (the date after which no further clutches would be laid). These were drawn randomly from a normal distribution defined by the mean and SD of the parameter. The start and stop dates and their SD were estimated by iteratively minimizing the sum of the squared deviations between observed clutch completion dates (including replacements) and those simulated by the re-nesting model described below (Green 1988).

Clutch size was four unless a randomly generated probability exceeded the estimated proportion of four-egg clutches, in which case the clutch size was three eggs. The laying period was calculated as the clutch size minus 1 day. The start of a female's breeding season was then calculated as her allocated clutch completion date minus the laying period.

The dates that the Ouse and Nene Washes were flooded were deduced from daily records of water levels in the main drainage ditch of each reserve. The flood mitigation options are listed in Appendix S1 and daily water levels that would have occurred in the main ditch (and hence the dates the washes were flooded) under each were calculated using hydraulic formulae (Posford Duvivier 2000). Birds were assumed to always breed on the washes if this habitat was available on their allotted laying date. If the washes were flooded on this date the pairs’ nesting date was delayed. If the flood drained before the pairs’ allotted laying date on arable they would nest on the washes on the date the flood subsided. If the flood persisted beyond their laying date on arable they would nest on arable fields instead.

During each day of the laying and incubation period (26 days; Cramp & Simmons 1983), the clutch was subjected to a habitat-specific likelihood of failure by testing whether a random probability exceeded the daily nest survival rate until it failed or hatched. Any clutches present on the washes on the day a flood started were classed as failing because of inundation. Pairs that failed were allowed to re-lay if the date of failure plus the replacement period (12 days; Hegyi & Sasvári 1998) was earlier than the allotted date on which they stopped laying. Birds would re-lay on washland if this was available when their replacement date arrived, otherwise they would renest on arable.

If the nest survived the laying and incubation period, a random likelihood was generated for each egg in the clutch and the egg was classed as being infertile or addled if this was below the probability of an egg being unviable. Each chick in the brood was subjected to a habitat-specific daily mortality rate until its fledging period (27 days; Cramp & Simmons 1983) elapsed or it died. Chicks on arable were assumed to remain on this habitat as no chicks were recorded moving onto the washes. Chicks were assumed to be unaffected by flooding as they are sufficiently mobile to avoid rising floodwater (Green, Cadbury & Williams 1987). Once the fledging period elapsed, one was added to the total number of juveniles fledged from the population. In the event of complete brood failure, the date upon which the last chick in the brood died was determined. The replacement period was added to this date, and if this date fell before the pairs allotted stop date they would re-lay, otherwise they failed. Double-brooding has never been recorded in black-tailed godwits and so successful pairs did not make second breeding attempts.

This procedure was repeated for each pair in each year, and productivity was calculated by dividing the number of chicks fledged by the number of pairs. Productivity of the population was estimated 999 times, and the mean and SD (equivalent to SE) of these bootstrapped replicates are presented to illustrate annual variation at each site since 1970. For comparisons of long-term performance of flood mitigation options, productivity was estimated for each of the 20 years for which predicted flooding levels were available (Posford Duvivier 2000). The total number of chicks fledged during the 20 years of study was divided by the total that would have fledged in the absence of flooding to provide a simple single-figure index of reproductive performance. A program written in Microsoft Visual Basic 6·0 was used to perform the simulations.

population size and modelling

The number of breeding pairs at each site was determined by three visits approximately 1 week apart in late April and early May. During these visits, the number of paired godwits in each field was determined by scanning with a telescope. Counts were conducted at both sites in all years since 1975 at the Ouse Washes and since 1984 at the Nene Washes.

An age-structured stochastic population model was used to simulate population trends under various management options. This incorporated annual variation in productivity estimated by the re-nesting model. The adult survival rate used in the model was 87·0% based on colour-ring resightings of East Anglian wintering islandica race godwits and dead recoveries of Netherlands ringed limosa race godwits (Beintema & Drost 1986; Gill et al. 2001). Immature survival was estimated as 75·0% by iteratively varying the value of this parameter until the sum of the squared deviations of predicted and observed population estimates in each year and each site were minimized. Years during which flooding occurred on the Ouse were omitted from this analysis as the numbers counted were abnormally low (Fig. 2), probably as a result of birds deferring breeding. A breeding likelihood of 0·43 during flood years and 1·00 during others minimized the sum of the squared deviation between the previously fitted model and the observed counts. All birds were assumed to breed for the first time at 2 years old (Cramp & Simmons 1983).

image

Figure 2. Observed (filled circles) population trends of breeding black-tailed godwits at the Ouse Washes (a) and Nene Washes (b) and those predicted (open circles) from a stochastic population model ± 1 SE. Open triangles denote years in which floods that significantly disrupted breeding occurred at each site.

Download figure to PowerPoint

The number of breeding female birds in year t+ 1 was estimated by summing the number of adult and immature females in year t and subjecting each to the annual rate of adult mortality. Birds were classed as dying (and were subtracted from the total surviving) if a randomly generated probability exceeded 0·870. The number of immature females in year t+ 1 was calculated by subjecting each juvenile female in year t to a survival rate of 0·750 in a similar manner. This procedure was repeated for each male in the population. Black-tailed godwits are monogamous, and so the number of breeding pairs at the start of each breeding season was set as the number of adults of the least abundant sex. All pairs were assumed to attempt breeding except in flood years, when pairs were only permitted to breed if a randomly generated probability was less than or equal to 0·43.

The number of juveniles fledged at the end of year t+ 1 was calculated from the likelihood of each breeding pair present in that year producing zero, one, two, three or four chicks. The re-nesting model was used to estimate the proportion of pairs producing each brood size in each of the 20 years of study for each of the management options, and one of these sets of values was chosen at random for each year of simulation. These were used to randomly allocate a brood size at fledging to each pair breeding in that year. The number of chicks fledged by each pair was summed to calculate the number of juveniles. Juveniles were allocated a sex at fledging (assuming a 1 : 1 sex ratio) by generating a random probability and classing the birds as male if this exceeded 0·5 and female if it were equal to or lower than this.

A closed model with the above structure was run for the Nene and Ouse Washes separately to examine population trends and persistence likelihoods when each site was dependent entirely on natal recruitment. An open model was also run for both sites simultaneously that allowed movements of recruiting birds in both directions. The likelihood of a bird recruiting away from its natal site was assumed to be 7%, based on colour-ring resightings of birds ringed as chicks at the Ouse and Nene Washes (n = 64; S. A. Schmitt, unpublished data) and figures presented in Kruk, Noordervliet & ter Keurs (1998). Birds surviving to breeding age were classed as emigrating if a random probability was less than or equal to 0·07, and were added to the total number of breeding males or females at the destination site. During years of flooding at the Ouse, it was assumed no immigration from the Nene occurred because of the absence of favourable nesting habitat. Birds were assumed to be site-faithful following recruitment because adult godwits with previous breeding experience rarely move more than 20 km in the Netherlands (Groen 1993). However, in 2000 significant movements of mature birds from the Ouse to the Nene did appear to occur in response to an exceptionally deep flood that inundated all marginal foraging habitat. Numbers at the Ouse and Nene permanently declined and increased by c. 7 pairs, respectively, in 2000, with anecdotal colour-ring resightings supporting the supposition that this was the result of movements among sites. Such events were not included in the model as their incidence was unknown (although clearly less than 1 : 30). Occurrence of such events would reduce population growth, viability and likelihood of meeting conservation targets.

The models were used to predict the population multiplication rate (λ) according to Lebreton & Clobert (1991), the likelihood of the population persisting and the likelihood of it increasing to the levels observed at the time of designation (46 pairs). The percentage annual rates of population growth were estimated as the population growth rate (r) (Lebreton & Clobert 1991) multiplied by 100. The time period over which simulations were run was varied (from 1 to 30 years in 1-year increments) to assess time scales over which extirpation or recovery of the population would occur. Populations were classed as persisting if one or more birds of both sexes was alive at the end of the simulated time period. In the open models, no recolonization following extirpation was assumed. This is plausible as godwits are semi-colonial nesters, and a loss of breeding pairs might result in a cessation of social attraction to immigrants. This procedure was repeated 999 times to determine the likelihood of a population persisting and recovering. The population persistence and recovery likelihoods over the varying time periods were examined in this manner for each of the flood mitigation options described in Appendix S1 (see Supplementary material).

Results

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

breeding parameters

Relative sample sizes within habitats varied among sites because of different population sizes and habitat use at each. Of the breeding attempts sampled, 9% and 54% were from Ouse washland and arable, respectively, and it should therefore be noted that the power to detect inter-site differences was low.

The first date of incubation was determined for 115 clutches (23 on arable). The average date on which incubation started on washland was 26 April (SD 9 days) and the average stop date was 20 May (SD 9). On arable, the start and stop dates were 1 May (SD 5) and 15 May (SD 4), respectively. The average duration of the laying season was therefore 25 days on washland and 15 on arable.

The modal clutch size was four eggs (83·8%, n= l30), with three-egg clutches comprising 12·2%. The remainder of clutches comprised two eggs (1·9%) or were super-normal clutches of five and seven eggs (0·6% and 1·3%, respectively). Excluding these, there were no effects of site (inline image = 0·4, P > 0·5), habitat (inline image = 1·7, P > 0·1) or year (inline image = 4·1, P > 0·3) on the proportion of four-egg clutches.

The fates of 156 clutches were recorded (22 on arable). Daily survival rates varied significantly among wet grassland and arable habitats (inline image = 15·1, P < 0·001). There were no effects of year (inline image = 2·9, P > 0·5), site (inline image = 0·76, P > 0·3) or calendar date (inline image = 0·5, P > 0·4) once the habitat-specific deviance was explained. The daily nest survival rates were 98·5% (Lower 95% Confidence Interval (LCI) 97·9, Upper 95% Confidence Interval (UCI) 98·9) on washland and 94·4% (LCI 90·0, UCI 96·9) on arable. Over the 26-day average laying and incubation period, the hatching success on washland was 67·4% and that on arable was 22·1%. Of the daily losses on washland, predation accounted for 96·2% and desertion for 3·8%, while on arable these figures were 66·6% and 33·3%, respectively. Predation accounts for 56% and desertion for 44% of the difference in overall nest survival among habitats.

The proportion of unviable eggs in a clutch was recorded for 100 nests. No eggs were unviable on arable, so analyses were confined to clutches laid on washland. The proportion of unviable eggs in a clutch varied among years (inline image = 11·49, P < 0·05). A higher proportion of eggs was unviable in those years in which floodwater remained on the washes beyond the first half of April than in years when it had drained by early April (inline image = 7·29, P < 0·01), probably because of damp soil chilling the eggs or increased humidity. Once this deviance was explained, there was no significant residual year effect (inline image = 5·95, P > 0·2). There were no significant effects of site after controlling for flooding effects (inline image = 1·9, P > 0·1). The percentage of unviable eggs in a clutch was 5·4% (LCI 3·2%, UCI 9·1%) in non-flood years and 13·3% (LCI 6·9%, UCI 24·2%) in flood years.

The survival of 116 chicks in 90 broods and 11 chicks in seven broods was recorded on washland and arable, respectively. Daily chick survival rate was 94·7% (LCI 93·6, UCI 95·6) on washland and 77·6 (LCI 66·7, UCI 82·3) on arable, with the lack of overlap between these estimates and their bootstrapped confidence intervals showing the differences to be significant. There were no effects of year, site, calendar date or chick age on survival rates within habitats. Over the 27-day average fledging period, the fledging success on wet grassland was 22·9% and that on arable was 0·1%. None of the chicks that hatched on arable subsequently moved to washland. Of the daily losses on washland, predation accounted for 94·6% and starvation for 5·4%, while on arable these were 81·9% and 18·1%, respectively. Predation accounted for 79% and starvation for 21% of the overall difference in chick survival among habitats.

productivity

The productivity of black-tailed godwits nesting solely on washland in years when flooding did not disrupt nesting was 0·65 (SE 0·23) chicks pair−1. The productivity of pairs nesting exclusively on arable was 0·01 chicks pair−1 (SE 0·002). The simulated annual productivity of black-tailed godwits breeding at the Ouse and Nene Washes between 1975 and 2003 are presented in Fig. 1. Floods caused productivity at the Ouse to be significantly lower than the level of 0·395 chicks pair−1 required to maintain a stable population in 59% of years (n = 29) and complete failure occurred in 76% of these. At the Nene, productivity was suppressed to approximately this level in only 10% of years (n = 20).

image

Figure 1. The simulated productivity of black-tailed godwits at the Nene (open circles) and Ouse Washes (filled circles) ± 1 SE. The dotted line represents the level of productivity required to maintain a stable population.

Download figure to PowerPoint

The productivity indices at the Ouse in relation to flood mitigation options are presented in Table 1. Productivity under current management was 42% of that in the absence of flooding. Changing a single component of the Ouse Washes or the catchment produced only marginal improvements in the productivity index of between 4% and 10%. Options involving pairing of changes to the Ouse Washes or the catchment caused the productivity index to improve to between 57% and 63%. Combining all available flood mitigation options resulted in the productivity index rising to 89%, which was more than double that resulting from current management.

Table 1.  The effect of flood mitigation options (see Appendix S1 for details) on the productivity and population multiplication rate (λ) of black-tailed godwits at the Ouse Washes. Productivity is expressed as a percentage of that which would have been attained if no floods had occurred during the 20 years of study
OptionProductivityOpen populationClosed population
Without option 12With option 12λSDλSD
Option 042·193·61·0330·0230·9020·048
Option 149·593·61·0460·0190·9230·044
Option 2a49·195·51·0400·0210·9160·044
Option 2b49·496·71·0400·0220·9170·043
Option 348·796·11·0400·0210·9210·044
Option 448·596·51·0390·0220·9230·043
Option 548·996·51·0390·0210·9210·042
Option 646·595·41·0380·0210·9120·045
Option 751·996·81·0420·0210·9240·043
Option 860·196·21·0590·0170·9380·039
Option 957·496·51·0490·0190·9390·038
Option 1063·096·21·0570·0190·9470·039
Option 1188·996·31·0870·0131·0000·029
Option 121·0920·0121·0100·029

Creating wet meadows outside the Ouse Washes resulted in the productivity index rising further to between 93% and 97% (Table 1). Combining habitat creation with the various flood mitigation options produced only marginal increases in the productivity index and so further modelling of habitat creation was conducted in isolation.

population trends

The population trends at the Ouse and Nene Washes differed markedly (Fig. 2). Numbers at the Ouse declined from 55 pairs in 1975 to five pairs in 1992, with fluctuations associated with the occurrence of spring flooding. The population recovered to a peak of 19 pairs in 1997 during a flood-free period, but declined to four pairs in 2004 as floods recurred. The Nene population increased at an average rate of 11% pa, with a more rapid increase of 32% between 1999 and 2000 and between 2003 and 2004.

The predictions of the population models incorporating constant adult and immature survival and flood-dependent variation in annual productivity and breeding likelihood broadly described the contrasting population trends at the Ouse and Nene (Fig. 2). This provided strong evidence that flood-induced breeding failure was responsible for the decline in godwit numbers at the Ouse Washes. It also suggested that the parameters used in the model were reasonably accurate and provided a quantitative link between flood patterns and population growth. The use of the models to predict future population trends under various management options is therefore justifiable.

Population growth rates at the Nene Washes were 5·1% pa (SD 0·7) and 4·1% pa (SD 0·8) for the closed and open models, respectively. These rates of growth resulted in the population in 30 years time being 195 (SD 40) and 145 (SD 31) pairs, respectively. The slower growth at the Nene Washes in the open model was because of net emigration of birds to the smaller and less productive Ouse Washes population.

Population multiplication rates of godwits at the Ouse under various management options are presented in Table 1. The closed population model resulted in the population declining at a rate of −10·3% pa under current conditions. Single flood mitigation options resulted in numbers declining at slower rates of around −8% pa, and paired options resulted in growth rates of −6% pa. Option 11 resulted in a population that was on average stable, while option 12 produced a population that increased at 1% pa. For the open population model, numbers increased at a rate of 3·2% pa under current conditions. Single in situ flood mitigation options produced population growth rates in the order of 4% pa. Paired mitigation options produced population growth in the order of 5% p.a. Options 11 and 12 produced population growth rates of 8·3% and 8·8% pa, respectively. The contrasting predictions of the closed and open models suggested that the Ouse population was a sink, dependent on immigration from the Nene Washes source in order to persist, under all management options other than 11 and 12.

likelihood of population persistence and meeting objectives

The likelihood of a closed population persisting over varying time periods in relation to the flood mitigation options employed are presented in Fig. 3a. The persistence likelihood was 91% for all options up to 5 years but decreased rapidly under most options thereafter. Only 8% of simulated populations persisted for 30 years under current conditions, 15% under single flood mitigation options and 30% under paired options. The rate of decline in population persistence with time under options 11 and 12 was slower, at 52% and 60% after 30 years, respectively.

image

Figure 3. The simulated likelihood of the black-tailed godwit population persisting at the Ouse Washes in relation to the number of years over which persistence is measured and flood mitigation options for (a) closed populations and (b) open populations. Single options are options 1–7 and paired options are options 8–10. See Appendix S1 for a list of mitigation options.

Download figure to PowerPoint

Allowing immigration from the Nene improved persistence likelihoods considerably for all options examined (Fig. 3b). Persistence likelihoods over 30 years were in excess of 75% under current conditions, 80% for single flood mitigation options, 90% for paired options and approached 100% for options 11 and 12.

The likelihood of meeting the target of 46 pairs over any time period up to 30 years was less than 0·03 when employing options 0–10 for either the open or closed population models. The closed population model showed that the likelihood of the target being met was zero under all management options. For open models under options 11 and 12 the likelihood of meeting the target was zero up to 20 years but this increased rapidly thereafter to 0·56 and 0·71 (respectively) by 30 years.

Discussion

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

diagnosis of the decline

Productivity of godwits at the Ouse and Nene Washes was 0·65 chicks pair−1 if flooding did not disrupt breeding. This level of productivity was similar to those of godwits nesting on reserves in the Netherlands (Kruk, Noordervliet & ter Keurs 1997; Schekkerman & Müskens 2000; Groen & Hemerik 2002) and generally higher than those nesting on dairy farms there (Kruk, Noordervliet & ter Keurs 1997; Schekkerman & Müskens 2000). It is also above the level of productivity required to maintain a stable population, given the immature and adult survival rates used in this study. It is likely that sympathetic reserve management at the Ouse and Nene, such as manipulation of water levels and delayed grazing or mowing, has contributed to the high productivity of godwits breeding there.

Productivity at the Ouse and Nene Washes exhibited striking spatial and temporal variation, with the Ouse experiencing complete breeding failures in 42% of years and depressed productivity in a further 12% of these. In contrast, at the Nene complete breeding failures did not occur and reduced productivity was only evident in 10% of years. Differences among sites and years were explained by the frequency and duration of flooding on the washes between mid-April and mid-May. A population model demonstrated that flood-dependent variation in productivity was sufficient to explain the contrasting population trends at the Ouse and Nene. This provides strong evidence that flooding is responsible for the decline in numbers of breeding godwits at the Ouse Washes, and that effective mitigation of the effects of flooding on productivity will promote a population recovery.

accuracy of model predictions

Predictions derived from population models must be interpreted with caution as these are only as robust as the data and parameter estimates upon which they are based (for a review see Bessinger 1998). The parameters used in the model provided a reasonably accurate description of the population trends of black-tailed godwits at the Ouse and Nene Washes over the previous 30 years (see the Results) but may not continue to do so in the future. The model assumes that the only parameter to change in the future is flood-dependent rates of productivity. However, any changes in survival rates in response to changes in hunting pressure or habitat availability in wintering grounds or along migration routes (Beintema & Drost 1986) will reduce the accuracy of the predictions. As godwits have relatively long generation times, their population trends will be particularly sensitive to changes in adult survival rates (Lebreton & Clobert 1991). Variations in productivity as a result of variations in predation rates could also cause inaccuracies in model predictions. Furthermore, most of the data for washland were collected at the Nene and, because of small sample sizes at the Ouse, the power to test for site effects was low. If these values for the Ouse were in reality lower than those for the Nene, the models would produce optimistic predictions.

The predictions of the open models are sensitive to variations in movement rates, but the available estimate of this is based on small sample sizes and factors affecting it are unknown. Studies of site fidelity in black-tailed godwits (Groen 1993) and other waders (Redmond & Jenni 1982; Grant 1991; Berg 1994; Jackson 1994) show that females exhibit lower site fidelity than males. Imbalances in the sex ratio of immigrants will tend to reduce population growth rates at the Ouse. Furthermore, site fidelity in waders is related to habitat quality and previous breeding success (Groen 1993; Berg 1994), such that flooding may result in the Ouse receiving less immigration and greater emigration than estimated in the models. This would further reduce the efficacy of some in situ flood mitigation options. Further research into movement rates and the factors that affect these are required in order to refine the models predictions.

The patterns of flooding experienced in the past may also differ from those that occur in the future independently of flood mitigation measures used. Flooding of the Ouse Washes generally follows periods of heavy rainfall coinciding with high tides, and climatic models predict rising sea levels (Hulme, Hossell & Parry 1993; Shennan 1993) and heavier summer rainfall events (Christensen & Christensen 2003). As such, in situ flood mitigation measures may be less effective than predicted by the models.

Although the predictions of population models are likely to be inaccurate, they provide powerful heuristic tools for assessing the relative merits of various conservation options (Bessinger 1998). In this study, models have allowed the relative benefits of conservation measures and godwit population dynamics to be quantified, and their use in selecting a parsimonious management option is justifiable despite the reservations outlined above.

selection of a management option

The efficacy of the flood mitigation options in conserving black-tailed godwits at the Ouse Washes varied markedly. The single in situ flood mitigation options had only marginal effects on the incidence of flooding, and so were inadequate to improve godwit productivity to a level that allowed the population to recover to the target level. Furthermore, the population was a sink under these options and was likely to decline in the absence of immigration from the Nene. This is highly undesirable as any factors affecting the population at the Nene would also be manifested at the Ouse because of a reduction in immigration. Pairing flood mitigation options produced marked increases in productivity and population growth compared with current conditions and single options, but still failed to meet the conservation targets within 30 years or to make persistence at the Ouse independent of events at the Nene. As such, the single and paired mitigation options can be regarded as poor value for money.

Combining all of the single in situ flood mitigation options (option 11) and creating compensatory wet meadow habitat on arable fields outside the Ouse Washes (option 12) both resulted in notable improvements in productivity. These were sufficient to produce a population that would be self-sustaining in the absence of immigration and had a high likelihood of persisting and recovering to target levels within 30 years. Option 12 is the most attractive for conserving black-tailed godwits as it is cheaper than option 11 and the time to implementation is likely to be much shorter. This option is also less susceptible to the increasing incidence and severity of flooding predicted for the future that could reduce the efficacy of in situ measures.

Although option 12 appears the most parsimonious means of conserving godwits at the Ouse, there are more assumptions associated with modelling its effects than for the in situ options. Godwits were assumed to nest on these fields if the washes were flooded, but the high site fidelity of the species (Groen 1993) may result in them continuing to select the arable fields adjacent to their preferred breeding washes. An experimental pilot project to create habitat on 44 ha of arable land by the Ouse Washes was initiated in 2002. Although the site supported suitable habitat for breeding waders during a flood in 2004, godwits chose to nest, and fail, on the arable fields they used during previous flood years. If birds continue to use their current arable site, the performance of option 12 would be identical to that under current conditions. Relocating birds by social facilitation has been successful for other colonially nesting Charadriiformes (Kress 1997; Roby et al. 2002) and will be attempted for godwits at created sites. If this fails, the acquisition and creation of wet grassland on the arable fields currently used by breeding godwits is likely to be the only means of producing an effective ex situ management option. While this may be necessary to ensure reproductive success of established breeding pairs at the Ouse, recruiting birds do move over greater distances (see the Results and Kruk, Noordervliet & ter Keurs 1998) and this provides greater flexibility in selection of sites for habitat creation.

The model assumes that productivity on created wet grassland during floods is similar to that on the washes, but it is unknown whether this will be the case. The reduced productivity of godwits on arable is partially attributable to desertion of nests and starvation of chicks (see the Results). Creating wet grassland on arable land used by godwits for laying would be expected to ameliorate both forms of loss as food availability for both adults and chicks would be improved. However, the main explanation for lower productivity on arable is elevated predation. The higher predation on arable fields might be habitat-specific, with nests and chicks being less concealed in crop monocultures (Skeel 1983; Baines 1990) or levels of parental nest defence (Byrkjedal 1985; Green 1986) being reduced because of off-duty parents having to forage away from their territory. Alternatively, elevated predation on arable could be situation-specific if floods displace both godwits and their predators into smaller areas outside the washes, such that predator encounter rates with godwit nests and chicks are higher irrespective of habitat. Floods could also reduce the abundance of alternative prey, such that predators become more reliant on wader eggs and chicks for food. In either case, the relatively small size of the created grasslands and absence of flooding will allow more intense predator exclusion or control than would be feasible on the washes. For example, an electrified perimeter fence has been erected around the pilot project to reduce red fox Vulpes vulpes L. incursion. In the event of colonization of created grasslands, research into predation on godwit nests and chicks will be required to design an effective predator management strategy.

Acknowledgements

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

We thank the RSPB staff at the Nene and Ouse Washes for their help and support throughout the project, particularly Charlie Kitchin, Jonathan Taylor, Cliff Carson, Dave Suddaby and Leigh Marshall. Jen Smart assisted with collection of data in 2000, and Mark Collier and Kim Fenton assisted in 2001. Thanks to John Kemp and Carl Mitchell from the WWT for providing data from Welney WWT reserve. Sarah Dawkins, Andrew Dodd, Ken Smith and David Gibbons provided comments on an earlier draft. Thanks also to the British Trust for Ornithology and English Nature for granting licenses for fitting radio-tags to godwit chicks.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Baines, D. (1990) The roles of predation, food and agricultural practice in determining the breeding success of the lapwing Vanellus vanellus on upland grasslands. Journal of Animal Ecology, 59, 915929.
  • Beintema, A.J. & Drost, N. (1986) Migration of the black-tailed godwit. Gerfaut, 76, 3762.
  • Beintema, A.J. & Melter, J. (1998) Black-tailed godwit. The EBCC Atlas of European Breeding Birds; Their Distribution and Abundance. (eds W.J.M.Hagemeijer & M.J. Blair). T.&A.D. Poyser, London, UK.
  • Beintema, A.J. & Müskens, G.J.D.M. (1987) Nesting success of birds breeding in Dutch agricultural grasslands. Journal of Applied Ecology, 24, 743758.
  • Benstead, P., Drake, M., Jose, P., Mountford, O., Newbold, C. & Treweek, J. (1997) The Wet Grassland Guide: Managing Floodplain and Coastal Wet Grasslands for Wildlife. RSPB, Sandy, UK.
  • Berg, Å. (1994) Maintenance of populations and causes of population changes of curlews Numenius arquata breeding on farmland. Biological Conservation, 67, 233238.
  • Bessinger, S.R. (1998) On the use of demographic models of population viability in endangered species management. Journal of Wildlife Management, 62, 821841.
  • Den Boer, T.E. (1995) Meadowbirds: Facts for Conservation[Dutch with English summary]. Technical Rapport Vogelbesherming Nederland 16., Zeist, Nederlands.
  • Byrkjedal, I. (1985) Time budget and parental labour division in breeding black-tailed godwits Limosa 1. limosa. Fauna Norvegicus, Series C, Cinculus, 8, 2434.
  • Cadbury, C.J. & Kirby, J. (1993) Black tailed godwit. The New Atlas of Breeding Birds in Britain and Ireland (eds D.W.Gibbons, J.B.Reid & R.A.Chapman), pp. 19881991. T.&A.D. Poyser, London, UK.
  • Christensen, J.H. & Christensen, O.B. (2003) Severe summer flooding in Europe. Nature, 421, 805.
  • Cottier, E.J. & Lea, D. (1969) Black-tailed godwits, ruffs and black terns breeding at the Ouse Washes. British Birds, 62, 259270.
  • Cramp, S. & Simmons, K.E.L. (1983) Handbook of the Birds of Europe, the Middle East and North Africa: The Birds of the Western Palearctic. Volume III. Waders to Gulls. Oxford University Press, Oxford UK.
  • Gibbons, D.W., Avery, M.I., Baillie, S.R., Gregory, R.D., Kirby, J., Porter, G. & Williams, G. (1996) Bird species of conservation concern in the United Kingdom, Channel Islands and Isle of Man: revising the Red Data List. RSPB Conservation Review, 10, 718.
  • Gill, J.A., Norris, K., Potts, P.M., Gunnarsson, T.G., Atkinson, P.W. & Sutherland, W.J. (2001) The buffer effect and large-scale population regulation in migratory birds. Nature, 412, 436438.
  • Grant, M.C. (1991) Nesting densities, productivity and survival of breeding whimbrel Numenius phaeopus. Shetland. Bird Study, 38, 160169.
  • Grant, M.C. (2002) Effects of radio-tagging on the weight gain and survival of curlew Numenius arquata chicks. Bird Study, 49, 172176.
  • Grant, M.C., Orsman, C., Easton, J., Lodge, C., Smith, M., Thompson, G., Rodwell, S. & Moore, N. (1999) Breeding success and causes of breeding failure of curlew Numenius arquata in Northern Ireland. Journal of Applied Ecology, 36, 5974.
  • Green, R.E. (1985) Summer Flooding and Black-tailed Godwits at the Ouse Washes. RSPB, Sandy, UK.
  • Green, R.E. (1986) The Management of Lowland Wet Grassland for Breeding Waders. RSPB, Sandy, UK.
  • Green, R.E. (1988) The effects of environmental factors on the timing and success of breeding of common snipe (Aves: Scolopacidae). Journal of Applied Ecology, 25, 7993.
  • Green, R.E., Cadbury, C.J. & Williams, G. (1987) Floods threaten black-tailed godwits breeding at the Ouse Washes. RSPB Conservation Review, 1, 1416.
  • Green, R.E., Tyler, G.A., Stowe, T.J. & Newton, A.V. (1997) A simulation model of the effect of mowing of agricultural grassland on breeding success of the corncrake (Crex crex). Journal of Zoology, London, 243, 81115.
  • Groen, N.M. (1993) Breeding site tenacity and natal philopatry in the black-tailed godwit Limosa 1. limosa. Ardea, 81, 107113.
  • Groen, N.M. & Hemerik, L. (2002) Reproductive success and survival of black-tailed godwits Limosa limosa in a declining local population in the Netherlands. Ardea, 90, 239248.
  • Hazler, K.R. (2004) Mayfield logistic regression: a practical approach for analysis of nest survival. Auk, 121, 707716.
  • Hegyi, Z. & Sasvári, L. (1998) Components of fitness in lapwings Vanellus vanellus and black-tailed godwits Limosa limosa during the breeding season: do female body mass and egg size matter? Ardea, 86, 4350.
  • Hulme, M., Hossell, J.E. & Parry, M.L. (1993) Future climate change and landuse in the United Kingdom. Geographical Journal, 159, 131147.
  • Jackson, D.B. (1994) Breeding dispersal and site-fidelity in three monogamous wader species in the Western Isles, UK. Ibis, 136, 463473.
  • Kress, S.W. (1997) Using animal behaviour for conservation: case studies in seabird restoration from the Maine coast, USA. Journal of the Yamashina Institute for Ornithology, 29, 126.
  • Kruk, M., Noordervliet, M.A.W. & Ter Keurs, W.J. (1997) Survival of black-tailed godwit chicks Limosa limosa in intensively exploited grassland areas in the Netherlands. Biological Conservation, 80, 127133.
  • Kruk, M., Noordervliet, M.A.W. & Ter Keurs, W.J. (1998) Natal philopatry in the black-tailed godwit Limosa limosa L. and its possible implications for conservation. Ringing and Migration, 19, 1316.
  • Lebreton, J.-D. & Clobert, J. (1991) Bird population dynamics, management and conservation: the role of mathematical modelling. Bird Population Studies, Relevance to Conservation and Management (eds C.M.Perrins, J.D.Lebreton & G.J.M. Hirons), pp. 105125. Oxford University Press, Oxford, UK.
  • Manolis, J.C., Andersen, D.E. & Cuthbert, F.L. (2000) Uncertain nest fates in songbird studies and variation in Mayfield estimation. Auk, 117, 615626.
  • NRA (1993) Ouse Washes Summer Flood Control. NRA Anglian Region, Peterborough, UK.
  • Posford Duvivier (2000) Ouse Washes Habitat Protection and Funding Group: Overview of Various Measures to Alleviate Summer Flooding. Report No. F4667/CP690. Posford Duvivier, Peterborough, UK.
  • Redmond, R.L. & Jenni, D.A. (1982) Natal philopatry and breeding area fidelity of long-billed curlews (Numenius americanus): patterns and evolutionary consequences. Behavioural Ecology and Sociobiology, 10, 277279.
  • Roby, D.R., Collis, K., Lyons, D.E., Craig, D.P., Adkins, J.Y., Myers, A.M. & Suryan, R.M. (2002) Effects of colony relocation on diet and productivity of Caspian terns. Journal of Wildlife Management, 66, 662673.
  • Schekkerman, H. & Müskens, G. (2000) Do black-tailed godwits Limosa limosa breeding in agricultural grasslands produce sufficient young for a stable population? Limosa, 73, 121134.
  • Shennan, I. (1993) Sea level changes and the threat of coastal inundation. Geographical Journal, 159, 148156.
  • Skeel, M.A. (1983) Nesting success, density, philopatry and nest site selection of the whimbrel (Numenius phaeopus) in different habitats. Canadian Journal of Zoology, 61, 218225.

Supporting Information

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

Appendix 1. Flood mitigation options, their cost (millions of pounds) and a brief description of their effects. For full details, see Posford Duvivier (2000). Note that costs only include the cost of creation or installation of the option, and not the cost of maintaining or managing them in perpetuity.

FilenameFormatSizeDescription
JPE_1076_sm_appendix1.doc39KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.