SEARCH

SEARCH BY CITATION

Keywords:

  • broods;
  • density-dependent reproduction;
  • ducks;
  • fledged birds;
  • nest box addition;
  • production;
  • resource management

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
     Hole-nesting birds are frequently faced with a shortage of suitable nest sites in regions of intensive forest management. Nest boxes are sometimes provided to alleviate nest-site limitation in cavity-nesting waterfowl and are also recommended for several rare and endangered species. However, the impacts on effective breeding numbers and breeding success have rarely been considered, particularly in instances where density dependence might operate.
  • 2
     We experimentally manipulated nest sites to assess limits on the population size of a secondary cavity-nesting species, the common goldeneye Bucephala clangula, living on freshwater lakes. We also examined density dependence in their reproductive output.
  • 3
     Breeding pairs were counted in experimental and control areas over a 12-year period; for 4 years (1988–91) before nest box addition (1992–94 in the experimental area) and for 5 years (1995–99) afterwards. Broods were counted each year between 1988 and 1999 to study reproductive output.
  • 4
     Mean number of pairs per lake increased after the addition of nest boxes in the experimental area but not in the control area. However, neither the mean number of broods per lake nor the mean number of fledged birds per lake increased significantly in the experimental area.
  • 5
     When the whole period of 1988–99 was considered and data pooled from all the lakes, the numbers of broods and fledged birds showed negative density dependence of reproductive output.
  • 6
     Our results indicate that nest sites limit the population size of breeding common goldeneye, but show also that density-dependent factors operate to limit reproductive output. The possibility that density dependence may negate management actions directed at increasing breeding numbers in cavity-nesting waterfowl should be considered carefully before taking these actions. This also applies to nest box provisioning programmes aiming to manage populations of endangered species.

Introduction

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

Resource limitation has fundamental ecological consequences for individual performance, population size and community structure (Wiens 1984; Begon, Harper & Townsend 1996). Food is of primary importance for animals and its role as a limiting factor has been studied intensively (reviews in Newton 1980; Martin 1987; Wiens 1989a,b; Boutin 1990). However, other resources may also be limiting. Nest sites, in particular, have been suggested to be an important resource for hole-nesting animals (Price 1984; Wiens 1989a; Begon, Harper & Townsend 1996). Studies of nest-site limitation in mammals are rare (Brady, Risch & Dobson 2001) but the role of nest sites in limiting breeding densities of birds has been demonstrated in several observational studies and experiments (reviews in Newton 1994, 1998). Nest-site limitation is particularly acute in regions of intensive forest management, where especially large hollow-using species may be excluded or kept at low levels by clearfelling or the selective removal of old and dying trees (Newton 1994; Williams et al. 2001). Furthermore, not only the number but the quality of cavities may play an important role in nest-site limitation of cavity-dependent species (Sedgeley 2001). Interactions between different types of resources or between different types of limiting factors such as food, nest sites and predators may also be important in limiting breeding density and reproductive output (Newton 1998).

The shortage of nest sites can regulate breeding numbers and reproductive output in a density-dependent manner (Newton 1998). An essential part in the study of population regulation is the empirical identification of processes stabilizing populations (Murdoch 1994; Harrison & Cappuccino 1995; Turchin 1995). Given that several limiting factors may be involved, provision of nest sites enables the experimental assessment of density dependence in reproduction, a potentially important regulatory factor. Several experimental studies on small hole-nesting passerines have found density dependence in some reproductive parameters (Alatalo & Lundberg 1984; Török & Tóth 1988; Dhondt, Kempenaers & Adriaensen 1992).

Species of the genus Bucephala are large secondary cavity-nesters sometimes limited by nest sites, i.e. they do not excavate the nesting cavity but are dependent on cavities provided by large woodpeckers or physical damage of trees. Indeed, the provision of nest boxes has increased the use of new boxes in the bufflehead Bucephala albeola L. (Gauthier & Smith 1987), Barrow’s goldeneye Bucephala islandica G. (Savard 1988a) and common goldeneye Bucephala clangula L. (Sirén 1951; Johnson 1967; Eriksson 1982; Dennis & Dow 1984), suggesting that breeding populations may be limited by nest-site availability (Newton 1994, 1998). However, caution is needed with this conclusion because earlier studies lacked control areas in which there was no nest-box provision (Newton 1994, 1998). Moreover, most of the earlier studies only checked the use of nest boxes but did not include data on breeding numbers. The studies by Gauthier & Smith (1987) on bufflehead and Savard (1988a) on Barrow’s goldeneye are exceptions, and those authors reached mixed conclusions. Gauthier & Smith (1987) concluded that the addition of nest boxes did not increase the density of breeding pairs, whereas Savard (1988a) observed an increase. If changes in population size are not studied simultaneously, a potential pitfall is that, after nest boxes are erected, females breeding in the area only switch from natural cavities to boxes and actual breeding numbers probably remain unchanged. This kind of switch was observed by Gauthier & Smith (1987) in bufflehead and by Petty, Shaw & Anderson (1994) in the tawny owl Tyto alba. Most important, earlier studies of the effects of nest-box addition in Bucephala species did not examine simultaneously the effect of nest-site provisioning on production (but see Savard 1988a), let alone the density dependence of reproductive output. This information is important because nest-box provision may cause a switch in the most limiting resource (e.g. from nest-site limitation to duckling food limitation) and, hence, the management action may not result in a desired outcome (Bradbury et al. 2001). Finally, to be useful widely, a particular management action needs to deliver consistent results wherever it is applied (cf. Sanders 2000).

Despite gaps in basic knowledge of the usefulness and efficiency of nest-box provision, the technique has been used to enhance production of cavity-nesting waterfowl (references in Savard 1988a; Eadie, Sherman & Semel 1998) and to manage populations of other hole-nesting species (Petty 1985; Caine & Marion 1991; Newton 1994). In some species, nest-box provision has caused undesired changes in intraspecific social interactions and population dynamics (Eadie, Sherman & Semel 1998). This finding is of particular concern because nest-box programmes have been recommended as a conservation tool for several rare and endangered cavity-nesting species (see table 12.2 in Eadie, Sherman & Semel 1998).

In general, the application of sound ecological principles is of primary and increasing importance in successful ecological management (Ormerod & Watkinson 2000). One of these central ecological principles is the occurrence of density-dependent mechanisms, the neglect of which may lead to overly optimistic expectations of the effects of management actions (Frederiksen, Lebreton & Bregnballe 2001). Clearly, there is a need for more information about the impact of nest-box provision on the effective population size and breeding output of different species in different ecological contexts. Here we report on a nest-box addition experiment addressing the following questions. First, does nest-site availability limit the breeding population size of common goldeneye? Secondly, does nest-box addition increase the reproductive output or does it rather reveal density dependence of reproductive output in common goldeneye? To investigate whether there is a shift in nest-site use of common goldeneye females, potentially a problem in previous nest-box addition studies (see above), we also examined the dynamics of occupation of old and new nest boxes at the beginning of the nest-box provision experiment.

Materials and methods

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

study areas and nest-box addition

The study was conducted in two areas in southern Finland, an experimental area (61°35′ N, 29°40′ E) and a control area (61°32′ N, 26°23′ E) about 170 km to the west (Table 1). Both study areas are dominated by pine Pinus sylvestris L. or mixed (pine, birch Betula spp. and spruce Picea abies (L.) Karst) forests interspersed with lakes of varying size. The experimental area included 35 small lakes (mean size 3·7 ha, range 0·2–24·0 ha) and the control area 17 small lakes (mean size 6·8 ha, range 1·0–20·0 ha). All lakes were open water. Distances between extreme lakes on the south–north and east–west gradient in the experimental area were 11·8 km and 6·8 km, respectively, the corresponding distances being 8·0 km and 8·5 km in the control area. Old black woodpecker Dryocopus martius L. cavities are the most important natural nest sites of common goldeneye and are available in both study areas.

Table 1.  The number of lakes studied, the number of nest boxes available and statistics of the number of breeding pairs in the experimental and control area for the periods before (1988–91) and after (1995–99) nest-box addition in the experimental area
 Experimental areaControl area
 BeforeAfterBeforeAfter
Number of lakes35351717
Number of nest boxes14641414
Number of breeding pairs
 Mean27·538·414·016·0
 SE 3·0 1·7 1·8 1·0
 Range21–3334–4311–1815–20

Fourteen nest boxes on six lakes in the experimental area, and 14 nest boxes on nine lakes in the control area, were available for common goldeneye for several years before 1992. In the experimental area, new nest boxes were erected in three phases (Pöysä 1999): 15 boxes in spring 1992, before the break-up of ice cover and arrival of migrating common goldeneyes, another 15 boxes in winter 1992–93, and a further 20 boxes in winter 1993–94 (together with old boxes totalling 64 nest boxes on 30 lakes from 1994 onwards). The boxes (25 × 26 × 70 cm, hole diameter of 9 cm) were made of Scots pine Pinus sylvestris board and they were burned slightly (charred) to make them less conspicuous. The boxes were placed in pairs so that one box was erected at the shoreline and the other inside the forest at 30–140 m from the shoreline (Pöysäet al. 1999; Intsilä study area). Based on a nest-site prospecting experiment, common goldeneye females were known to prospect shore and forest boxes equally (Pöysäet al. 1999), meaning that visibility per se may not play a central role in the selection of a nest site in this species (Eadie, Sherman & Semel 1998). Hence, all new nest boxes were potentially suitable for common goldeneye. Irrespective of the availability of nest sites, breeding common goldeneye use several lakes within a given area during the egg-laying, incubation and brood-rearing stages of the breeding cycle (Eriksson 1979; Mallory et al. 1993; Pöysä & Virtanen 1994; Wayland & McNicol 1994; H. Pöysä, unpublished data). Therefore, five small lakes (0·2–1·4 ha) that did not have nest boxes but were situated close to the experimental lakes were also included in the study area. The mean number of nest boxes per lake in the experimental area was 0·4 (SE = 0·2, n= 35) before the nest-box addition, and 1·8 (SE = 0·3) subsequently. No nest boxes were added to the control area.

In the experimental area, the occupation of the nest boxes was assessed by visiting each box at least three times between late April and early June each year; ≥ 1 egg laid was interpreted as a breeding attempt in a given box (Pöysä 1999). Deserted clutches and unhatched eggs were removed from the boxes after each breeding season.

breeding numbers and reproductive output

The number of breeding common goldeneye pairs was estimated on each lake in both study areas each year between 1988 and 1999 using standard waterfowl point counts (Koskimies & Pöysä 1989; Koskimies & Väisänen 1991). According to recommendations about optimal census times in Pöysä (1996), censuses were conducted in the middle of May. If necessary, several census points were established per lake to ensure that all the open water and shoreline of a given lake was surveyed. Observations were translated into breeding pair numbers according to standard criteria, i.e. pairs and lone adult males were counted as breeding pairs (Koskimies & Väisänen 1991).

Data on broods and fledged birds were not gathered in the control area. In the experimental area, number of broods and fledged birds was estimated using the point count method. Three censuses were conducted on each lake in each year. The first brood census was done around mid-June, the second in early July and the last one in late July–early August. In addition to these censuses, lakes having common goldeneye broods were visited several times in August to obtain the final number of fledged birds. Whenever possible the number of young and their age class (seven classes: Ia–c, IIa–c, III; length of an age class 7·5 days; Pirkola & Högmander 1974) in different broods was recorded on each census. Due to the gradual development of the plumage and the shape of young (Pirkola & Högmander 1974), age is easy to determine in the field. Offspring ≥ 40 days old were considered fledged (for definition of fledging age, see Pöysä, Virtanen & Milonoff 1997). Data from the first two censuses were used to estimate the number of broods produced (i.e. mean number of broods per census, hereafter simply the number of broods). The number of fledged birds was estimated on the basis of the third census and additional visits.

Canadian beaver Castor canadensis Kuhl has occupied one of the lakes on the edge of the experimental area since 1992. The beavers caused a considerable increase in water level, which was unusually high in most years between 1992 and 1999. Common goldeneye broods prefer flooded areas that harbour a high abundance of invertebrates (Nummi & Pöysä 1995) and, in general, food plays an important role in lake selection by broods, inducing some movement between lakes (Eriksson 1978; Pöysä & Virtanen 1994; Pöysä, Rask & Nummi 1994; Wayland & McNicol 1994). There were indications that the lake flooded by beavers attracted common goldeneye broods from lakes outside the experimental area. Data from individually marked females (coloured wing-tags) also confirmed that broods moved between lakes within the experimental area (H. Pöysä, unpublished data). Therefore, to ensure that the beaver-flooded lake did not confound our results and conclusions, we conducted a separate analysis of broods and fledged birds from all lakes excluding that lake; below we give results analysed in both ways.

statistical analyses

To study the effect of nest-box addition in the experimental area between 1992 and 1994 (see above) on breeding numbers, we divided the time series of experimental and control areas into two parts, i.e. before (1988–91) and after (1995–99) nest-box addition (note that nest boxes were not added to the control area). For each lake we calculated the mean number of common goldeneye pairs for the period before nest-box addition and for the period after nest-box addition, and used these lake-specific paired values in tests. Because the data did not meet the requirements of parametric tests even after appropriate transformation, we tested the change in mean pair numbers by using the non-parametric Wilcoxon paired-sample test (Zar 1996). Note that, because differences of zero are ignored in the Wilcoxon paired-sample test (Zar 1996), actual sample sizes in the tests were smaller than the total number of lakes studied.

By using the same division of the time series we calculated, for each lake in the experimental area, the mean number of broods and fledged birds for the periods before and after nest-box addition and tested for differences between the periods by using the non-parametric Wilcoxon paired-sample test. To avoid the sampling error of lake-specific data due to brood movements (see above) and to increase the power of the test, we also looked for possible density dependence of reproductive output in the experimental area by using data from the whole time series 1988–99 and pooled from all lakes. We assessed density dependence separately for the nesting phase and the brood-rearing phase and used model II regression (geometric mean regression; Sokal & Rohlf 1981) in the tests. For the nesting phase, we regressed annual number of broods against annual number of breeding pairs and, for the brood-rearing phase, annual number of fledged birds against annual number of broods, and calculated 95% confidence limits (CL) for the slopes of these regressions. We compared the slopes of these regressions with expected slopes that were calculated as the mean annual number of broods per breeding pair and the mean annual number of fledged birds per brood. If the upper 95% limit of the slope from the regression is smaller than the expected slope, the rate of the increase of the dependent variable (i.e. number of broods or number of fledged birds) is smaller than the rate of the increase of the independent variable (i.e. number of breeding pairs or number of broods, as appropriate), indicating negative density dependence of reproductive output. It should be noted that the expected slopes used in these comparisons make our test of negative density dependence very conservative. Biologically sound expected slopes would be 1 for the number of broods (i.e. one pair will never have more than one brood) and 5·9 for the number of fledged birds (i.e. the mean brood size in the data from the first brood census).

We studied the dynamics of occupation of old (available before 1992, see above; one of the 14 old boxes was removed after the breeding season 1992 so only 13 boxes were included in this analysis) and new nest boxes between 1992 and 1995 as follows. There were three sets of new boxes; the first set (15 boxes) was erected before the breeding season in 1992, the second set (15 boxes) before 1993 and the third set (20 boxes) before 1994 (see above). For each set of new boxes we calculated the proportion of occupied boxes in years t1 (first year available) and t2 (second year available). For each set of new boxes we formed a comparison group of ‘old’ boxes (i.e. boxes that had been available before the year t1) as follows. The 13 old boxes served as the comparison group (first set of old boxes) for the first set of new boxes, these 13 old boxes plus the first set of new boxes comprised the comparison group (second set of old boxes totalling 28 boxes) for the second set of new boxes, and the 13 old boxes plus the first and second set of new boxes comprised the comparison group (third set of old boxes totalling 43 boxes) for the third set of new boxes. For each set of old boxes we calculated the proportion of occupied boxes in years t1 and t2 that were the same years as those in the corresponding set of new boxes. By using these sets of new and old boxes, we tested the change in the proportion of occupied boxes from year t1 to year t2 with repeated-measures analysis of variance where box type (new or old) was the grouping factor and year-specific occupation rate (t1 or t2) the repeated measure.

All statistical analyses were run with systat procedures (Wilkinson 1992) except that all calculations needed in model II regressions were done according to Sokal & Rohlf (1981). All probability values were two-tailed, according to systat procedures or Zar (1996), as appropriate.

Results

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

breeding numbers

The mean number of breeding pairs per lake increased significantly in the experimental area after the addition of nest boxes (before: mean = 0·8, SE = 0·2; after: mean = 1·1, SE = 0·3; Z= 1·965, P < 0·05, n= 30 in the test), but not in the control area (before: mean = 0·8, SE = 0·3; after: mean = 0·9, SE = 0·2; Z= 0·336, P > 0·70, n= 16 in the test). Also, the yearly total number of breeding pairs was consistently higher after nest-box addition than before nest-box addition in the experimental area but not in the control area (Table 1).

nest-box occupation

The use of new nest boxes increased from the first year in the experimental area (four in 1992, 17 in 1993 and 20 in 1994). The mean number of yearly breeding attempts between 1995 and 1999 in the new boxes was 11·0 (SE = 1·3), i.e. close to the increase in total number of breeding pairs from the period before nest-box addition to the period after nest-box addition (10·9 pairs, see above). The number of nesting attempts in the old boxes decreased slightly when new boxes were added (from 11 in 1992 to nine in 1993 and 1994), although the decrease was less than the simultaneous increase of nesting attempts in the new boxes. The mean number of breeding attempts in the old boxes was 6·8 (SE = 0·5) between 1995 and 1999.

The dynamics of nest-box occupation suggests that there was a shift in nest-site use from old boxes to new ones. Indeed, when the change in the proportion of occupied boxes from year t1 to year t2 was compared between new and old boxes, the trends were significantly the opposite: the use of new boxes increased from t1 to t2, whereas that of old boxes decreased (Table 2).

Table 2.  The proportion (%) of occupied boxes in year t1 and year t2 in three sets of new boxes (first time available in year t1) and old boxes (available before t1) and the results of repeated-measures anova of the change in the proportion of occupied boxes from year t1 to year t2. The number of nest boxes in each set is given in parentheses (see the Materials and methods, statistical analyses)
Box typePercentage occupied
Year t1Year t2
New boxes
Set 1 (n = 15)26·780·0
Set 2 (n = 15)33·360·0
Set 3 (n = 20)30·035·0
Old boxes
Set 1 (n = 13)84·669·2
Set 2 (n = 28)75·050·0
Set 3 (n = 43)53·534·9
Sourced.f.MSFP
Between subjects
Box type1 870·403 2·280·206
Error4 381·896  
Within subjects
Year1  56·333 0·370·576
Year × box type11728·011·3480·028
Error4 152·279  

reproductive output

The number of broods and the number of fledged birds did not increase significantly from the period before nest-box addition to the period after nest-box addition, although there was an increase in numbers of fledged birds (Table 3). Because the number of breeding pairs increased (see above), this finding suggests that reproductive output was at least to some degree density-dependent. Brood size did not differ between the two periods (brood observations from the first brood census included; before nest-box addition: mean = 6·3, SE = 0·7, n= 26; after nest-box addition: mean = 5·6, SE = 0·6, n= 30; t= 0·866, d.f. = 54, P= 0·39). The total number of fledged birds varied between 14 and 33 (mean = 20·5, SE = 4·3) before nest-box addition (1988–91) and between 25 and 35 (mean = 30·2, SE = 2·0) after nest-box addition (1995–99) when the beaver-flooded lake was included; the corresponding figures were 12–22 (mean = 17·0, SE = 2·1) before nest-box addition and 16–25 (mean = 21·6, SE = 1·6) after nest-box addition when the beaver-flooded lake was excluded.

Table 3.  Mean values of reproductive parameters for the periods before (1988–91) and after (1995–99) nest-box addition in the experimental area. For number of broods and number of fledged birds the values are means of 35 (A, beaver-flooded lake included) or 34 (B, beaver-flooded lake excluded; see the Materials and methods, breeding numbers and reproductive output) lakes. Z denotes Wilcoxon paired-sample test statistics and n actual sample size in the test
 Before nest-box additionAfter nest-box additionZPn
 MeanSEMeanSE
Number of broods
 A0·170·050·190·070·2400·8118
 B0·170·050·140·040·6870·4917
Number of fledged birds
 A0·590·250·860·340·8290·4117
 B0·490·230·640·260·5170·6116

When the whole period of 1988–99 was considered and data from all lakes were pooled, the number of broods did not increase with the number of breeding pairs (Fig. 1a). The number of fledged birds tended to increase with the number of broods when the beaver-flooded lake was included but not when it was excluded (Fig. 1b; note that the positive relationship was largely due to the data point in the upper right corner of Fig. 1b, i.e. a year when the number of both broods and fledged birds was exceptionally high in the beaver-flooded lake). The slopes of the corresponding regressions testing for density dependence were significantly smaller than the expected slopes, except for fledged birds when the beaver-flooded lake was included (upper 95% CL vs. expected; Table 4), indicating negative density dependence of reproductive output. Also, it should be noted that the slope of the regression was negative, not positive, in three out of four cases (Table 4). As a corollary, the annual number of broods per breeding pair decreased with increasing annual number of breeding pairs (beaver-flooded lake included: r = −0·693, P < 0·02; beaver-flooded lake excluded: r = −0·865, P < 0·001; n= 12 years in both cases) and the annual number of fledged birds per brood decreased with annual number of broods, although the correlation was not significant when the beaver-flooded lake was included (beaver-flooded lake included: r = −0·259, P > 0·20; beaver-flooded lake excluded: r = −0·831, P < 0·001; n= 12 years in both cases).

image

Figure 1. (a) Annual (1988–99) number of broods per census in relation to annual total number of breeding pairs and (b) annual total number of fledged birds in relation to annual number of broods per census. Filled circles give the data including the beaver-flooded lake, open circles excluding the beaver-flooded lake (see the Materials and methods, breeding numbers and reproductive output). Model II regressions of these relationships are given in Table 4.

Download figure to PowerPoint

Table 4.  Model II regressions of annual (1988–99) number of broods per census against annual total number of breeding pairs and annual total number of fledged birds against annual number of broods per census (A, beaver-flooded lake included; B, beaver-flooded lake excluded; see the Materials and methods, breeding numbers and reproductive output). 95% CL for the slopes of the regressions and expected slopes are given (see the Materials and methods, statistical analyses). n= 12 years in all cases. Data are shown in Fig. 1
 95% CLExpected slope
 LowerUpper
Number of broods vs. number of breeding pairs
 A y = 15·512 − 0·279x−0·473−0·0850·205
 B y = 11·669 − 0·232x−0·377−0·0870·209
Number of fledged birds vs. number of broods
 A y = −4·425 + 4·485x 1·901 7·0693·874
 B y = 34·517 − 3·019x−5·095−0·9434·039

Discussion

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

nest-site limitation

The role of nest sites in limiting densities of hole-nesting birds has been studied intensively (Newton 1994, 1998). However, nest-box provisioning experiments in large hole-nesting species are rare and firm conclusions are difficult to draw because of problems in experimental design (see the Introduction). Our nest-site provision experiment revealed that the availability of nest sites limits the breeding population size of common goldeneye: the number of breeding pairs increased significantly in the experimental area but did not increase in the control area where new nest boxes were not provided. Our result thus confirms the suggestion of earlier studies that common goldeneye breeding populations may be limited by nest-site availability (Sirén 1951; Johnson 1967; Eriksson 1982; Dennis & Dow 1984).

Based on a similar nest-box provision experiment, Gauthier & Smith (1987) concluded that nest sites did not limit breeding density of bufflehead in their study area, whereas Savard (1988a) found support for nest-site limitation in Barrow’s goldeneye. Both of these studies were of relatively short duration. Gauthier & Smith’s (1987) study lasted for 4 years (1982–85, new boxes added between 1982 and 1984) and Savard’s (1988a) study for 5 years (1980–84, new boxes added between 1981 and 1984), data from the first year serving as a background for comparisons in both studies. Our study involved 12 years, and we carefully separated the background years (1988–91) from the ‘effective’ years (1995–99). Hence, our experimental design was probably more efficient in detecting an effect than that of Gauthier & Smith (1987), which may explain the difference in our results. On the other hand, Savard (1988a) found that numbers of Barrow’s goldeneye pairs increased in response to nest-box addition, although not until 2 years after the erection of the first boxes. Gauthier & Smith (1987) did not find evidence that suitable natural cavities were in short supply for bufflehead in their area. In fact, the use of boxes by bufflehead was mostly compensatory: after nest-box provision females preferred boxes over natural cavities, and there was simply a shift in nest-site use, not an increase in breeding numbers (Gauthier & Smith 1987). Savard (1988a) did not address this possibility in Barrow’s goldeneye, although he mentioned that there was a scarcity of large natural cavities in his study area. We did not study the use of natural cavities by common goldeneye. However, in the beginning of the experiment there was a shift in nest-site use from old to new boxes, which may mirror a corresponding switch from nesting in natural cavities to using provided boxes. If that kind of switch occurred, it probably took place immediately after new boxes were available, i.e. mostly during the intervening years 1992–94 of nest-box addition that were excluded from our analysis (see above). In any case, the effect of nest-box provision on breeding numbers was clear and a possible switch in nest-site use did not confound this effect in our study.

All three Bucephala species are territorial (Savard 1982, 1984, 1988b; Gauthier 1987a; Eadie, Mallory & Lumsden 1995), and it has been suggested that territoriality limits breeding density in these species (Fredga & Dow 1984; Gauthier & Smith 1987; Eadie, Sherman & Semel 1998). A relatively large proportion of apparently non-nesting individuals and a high frequency of conspecific nest parasitism in our common goldeneye population (Pöysä 1999; H. Pöysä, unpublished data) suggest that many common goldeneye females use parasitic egg laying as an alternative reproductive tactic. This behavioural trait complicates the assessment of whether territoriality limits the number of actually breeding individuals in our study population.

density-dependent reproductive output

Provisioning of nest sites may increase the target breeding population to a level at which other limiting factors affect reproductive output in a density-dependent manner (Newton 1998). We found that reproductive output, measured in terms of the number of broods and fledged birds produced, was negatively density-dependent in our experimental area. This result was clear, especially considering the fact that our test of density dependence was overly conservative (see the Materials and methods, statistical analyses). Several mechanisms, operating at nesting or brood-rearing stage, may underlie this pattern. These include density-dependent clutch depredation (Fredga & Dow 1984), disturbance caused by parasitically laying females resulting in decreased hatching success with increasing population density (Eadie, Sherman & Semel 1998) and brood-stage territoriality resulting in increased duckling mortality, also potentially a density-dependent mechanism (Savard 1982; Gauthier 1987b; Savard, Smith & Smith 1991). All these mechanisms may apply to the present population because the level of both nest depredation and parasitism is quite high and varies considerably between years in the experimental area (Pöysä 1999), and common goldeneye females are very aggressive during brood rearing (Ruusila & Pöysä 1998). A study focusing on the actual mechanisms, and their relative importance, behind the density dependence is currently in progress but the results so far suggest that density dependence is involved during both the nesting and brood-rearing stage.

management implications

Resource management may not always result in a desired outcome in the target population (Eadie, Sherman & Semel 1998; Bradbury et al. 2001). Our study revealed that, even though nest-box provision increased breeding numbers, density dependence during the nesting and brood-rearing stage largely compensated for its effect on the numbers of fledged birds. This finding has an important management implication: density dependence of reproductive output should be considered in nest-box programmes aimed to increase breeding numbers and production of common goldeneye and other hole-nesting waterfowl. This has not been done so far. Furthermore, as indicated by a shift in nest-site use from old to new boxes in our study and from natural cavities to boxes in bufflehead in the study by Gauthier & Smith (1987), nest-box provision may not necessarily increase the number of breeders. Therefore, simultaneous monitoring of breeding numbers, and not just occupation rate of boxes, should always be included in nest-box addition programmes.

The results of the present study also have implications for the management of populations of endangered species in landscapes of extensive forest harvesting. Dozens of secondary cavity-nesting species currently are or have recently been of conservation concern, and nest-box programmes have been recommended as a management tool for some endangered species (Eadie, Sherman & Semel 1998). Nest-box provisioning may help, especially in areas of intensive forest management where nest-site limitation is acute. However, before extensive nest-box programmes are implemented, it should be ensured that the environment has not deteriorated with respect to other key resources, notably food, that may become the new most-limiting resource. Also, the risk of nest depredation for cavity-nesting species may increase in highly fragmented landscapes (Andrén 1995; Hartley & Hunter 1998; but see Pöysä, Milonoff & Virtanen 1997; Pierre, Bears & Paszkowski 2001), an additional factor that should be taken into account when considering nest-box provisioning as a management tool for endangered species.

Acknowledgements

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

We thank John Eadie, Markku Milonoff, Vesa Ruusila, Jean-Pierre Savard, Jorma Sorjonen and an anonymous referee for valuable comments on the manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Alatalo, R.V. & Lundberg, A. (1984) Density dependence in breeding success of the pied flycatcher (Ficedula hypoleuca). Journal of Animal Ecology, 53, 969977.
  • Andrén, H. (1995) Effects of landscape composition on predation rates at habitat edges. Mosaic Landscapes and Ecological Processes (eds L.Hansson, L.Fahrig & G.Merriam), pp. 225255. Chapman & Hall, London, UK.
  • Begon, M., Harper, J.L. & Townsend, C.R. (1996) Ecology: Individuals, Populations and Communities, 3rd edn. Blackwell Science, Oxford, UK.
  • Boutin, S. (1990) Food supplementation experiments with terrestrial vertebrates: patterns, problems, and the future. Canadian Journal of Zoology, 68, 203220.
  • Bradbury, R.B., Payne, R.J.H., Wilson, J.D. & Krebs, J.R. (2001) Predicting population responses to resource management. Trends in Ecology and Evolution, 16, 440445.
  • Brady, M.J., Risch, T.S. & Dobson, F.S. (2001) Availability of nest sites does not limit population size of southern flying squirrels. Canadian Journal of Zoology, 78, 11441149.
  • Caine, L.A. & Marion, W.R. (1991) Artificial addition of snags and nest boxes to slash pine plantations. Journal of Field Ornithology, 62, 97106.
  • Dennis, R.H. & Dow, H. (1984) The establishment of a population of goldeneyes Bucephala clangula breeding in Scotland. Bird Study, 31, 217222.
  • Dhondt, A.A., Kempenaers, B. & Adriaensen, F. (1992) Density-dependent clutch size caused by habitat heterogeneity. Journal of Animal Ecology, 61, 643648.
  • Eadie, J.M., Mallory, M.L. & Lumsden, H.G. (1995) Common goldeneye (Bucephala clangula). The Birds of North America, No. 170 (eds A.Poole & F.Gill), pp. 132. The Academy of Natural Sciences, Philadelphia, PA, and The American Ornithologists’ Union, Washington, DC.
  • Eadie, J., Sherman, P. & Semel, B. (1998) Conspecific brood parasitism, population dynamics, and the conservation of cavity-nesting birds. Behavioral Ecology and Conservation Biology (ed. T.Caro), pp. 306340. Oxford University Press, Oxford, UK.
  • Eriksson, M.O.G. (1978) Lake selection by goldeneye ducklings in relation to the abundance of food. Wildfowl, 29, 8185.
  • Eriksson, M.O.G. (1979) Aspects of the breeding biology of the goldeneye Bucephala clangula. Holarctic Ecology, 2, 186194.
  • Eriksson, M.O.G. (1982) Differences between old and newly established goldeneye Bucephala clangula populations. Ornis Fennica, 59, 1319.
  • Frederiksen, M., Lebreton, J.-D. & Bregnballe, T. (2001) The interplay between culling and density-dependence in the great cormorant: a modelling approach. Journal of Applied Ecology, 38, 617627.
  • Fredga, S.G. & Dow, H. (1984) Factors affecting the size of a local population of goldeneye Bucephala clangula (L.) breeding in Sweden. Viltrevy, 13, 225251.
  • Gauthier, G. (1987a) The adaptive significance of territorial behaviour in breeding buffleheads: a test of three hypotheses. Animal Behaviour, 35, 348360.
  • Gauthier, G. (1987b) Brood territories in buffleheads: determinants and correlates of territory size. Canadian Journal of Zoology, 65, 14021410.
  • Gauthier, G. & Smith, J.N.M. (1987) Territorial behaviour, nest-site availability, and breeding density in buffleheads. Journal of Animal Ecology, 56, 171184.
  • Harrison, S. & Cappuccino, N. (1995) Using density-manipulation experiments to study population regulation. Population Dynamics: New Approaches and Synthesis (eds N.Cappuccino & P. W.Price), pp. 131147. Academic Press, San Diego, CA.
  • Hartley, M.J. & Hunter, M.L. Jr (1998) A meta-analysis of forest cover, edge effects, and artificial nest predation rates. Conservation Biology, 12, 465469.
  • Johnson, L.L. (1967) The common goldeneye duck and the role of nesting boxes in its management in north-central Minnesota. Journal of the Minnesota Academy of Science, 34, 110113.
  • Koskimies, P. & Pöysä, H. (1989) Waterfowl censusing in environmental monitoring: a comparison between point and round counts. Annales Zoologici Fennici, 26, 201206.
  • Koskimies, P. & Väisänen, R.A. (1991) Monitoring Bird Populations: A Manual of Methods Applied in Finland. Zoological Museum, Finnish Museum of Natural History, Helsinki, Finland.
  • Mallory, M.L., Weatherhead, P.J., McNicol, D.K. & Wayland, M.E. (1993) Nest site selection by common goldeneyes in response to habitat features influenced by acid precipitation. Ornis Scandinavica, 24, 5964.
  • Martin, T.E. (1987) Food as a limit on breeding birds: a life history perspective. Annual Review of Ecology and Systematics, 18, 453487.
  • Murdoch, W.W. (1994) Population regulation in theory and practice. Ecology, 75, 271287.
  • Newton, I. (1980) The role of food in limiting bird numbers. Ardea, 68, 1130.
  • Newton, I. (1994) The role of nest sites in limiting the number of hole-nesting birds: a review. Biological Conservation, 70, 265276.
  • Newton, I. (1998) Population Limitation in Birds. Academic Press, London, UK.
  • Nummi, P. & Pöysä, H. (1995) Habitat use by different-aged duck broods and juvenile ducks. Wildlife Biology, 1, 181187.
  • Ormerod, S.J. & Watkinson, A.R. (2000) The age of applied ecology. Journal of Applied Ecology, 37, 12.
  • Petty, S.J. (1985) A negative response of kestrels Falco tinnunculus to nestboxes in upland forests. Bird Study, 32, 194195.
  • Petty, S.J., Shaw, G. & Anderson, D.I.K. (1994) Value of nest boxes for population studies and conservation of owls in coniferous forests in Britain. Journal of Raptor Research, 28, 134142.
  • Pierre, J.P., Bears, H. & Paszkowski, C.A. (2001) Effects of forest harvesting on nest predation in cavity-nesting waterfowl. Auk, 118, 224230.
  • Pirkola, M.K. & Högmander, J. (1974) Sorsanpoikueiden iänmääritys. Suomen Riista, 25, 5055.
  • Pöysä, H. (1996) Population estimates and the timing of waterfowl censuses. Ornis Fennica, 73, 6068.
  • Pöysä, H. (1999) Conspecific nest parasitism is associated with inequality in nest predation risk in the common goldeneye (Bucephala clangula). Behavioral Ecology, 10, 533540.
  • Pöysä, H. & Virtanen, J. (1994) Habitat selection and survival of common goldeneye (Bucephala clangula) broods – preliminary results. Hydrobiologia, 279/280, 289296.
  • Pöysä, H., Milonoff, M., Ruusila, V. & Virtanen, J. (1999) Nest-site selection in relation to habitat edge: experiments in the common goldeneye. Journal of Avian Biology, 30, 7984.
  • Pöysä, H., Milonoff, M. & Virtanen, J. (1997) Nest predation in hole-nesting birds in relation to habitat edge: an experiment. Ecography, 20, 329335.
  • Pöysä, H., Rask, M. & Nummi, P. (1994) Acidification and ecological interactions at higher trophic levels in a small forest lake: the perch and the common goldeneye. Annales Zoologici Fennici, 31, 397404.
  • Pöysä, H., Virtanen, J. & Milonoff, M. (1997) Common goldeneyes adjust maternal effort in relation to prior brood success and not current brood size. Behavioral Ecology and Sociobiology, 40, 101106.
  • Price, P.W. (1984) Alternative paradigms in community ecology. A New Ecology. Novel Approaches to Interactive Systems (eds P. W.Price, C. N.Slobodchikoff & W. S.Gaud), pp. 353383. John Wiley & Sons, New York, NY.
  • Ruusila, V. & Pöysä, H. (1998) Shared and unshared parental investment in the precocial goldeneye (Aves: Anatidae). Animal Behaviour, 55, 307312.
  • Sanders, M.D. (2000) Enhancing food supplies for waders: inconsistent effects of substratum manipulations on aquatic invertebrate biomass. Journal of Applied Ecology, 37, 6676.
  • Savard, J.-P.L. (1982) Intra- and inter-specific competition between Barrow’s goldeneye (Bucephala islandica) and bufflehead (Bucephala albeola). Canadian Journal of Zoology, 60, 34393446.
  • Savard, J.-P.L. (1984) Territorial behaviour of common goldeneye, Barrow’s goldeneye and bufflehead in areas of sympatry. Ornis Scandinavica, 15, 211216.
  • Savard, J.-P.L. (1988a) Use of nest boxes by Barrow’s goldeneyes: nesting success and effect on the breeding population. Wildlife Society Bulletin, 16, 125132.
  • Savard, J.-P.L. (1988b) Winter, spring and summer territoriality in Barrow’s goldenye: characteristics and benefits. Ornis Scandinavica, 19, 119128.
  • Savard, J.-P.L., Smith, G.E.J. & Smith, J.N.M. (1991) Duckling mortality in Barrow’s goldeneye and bufflehead broods. Auk, 108, 568577.
  • Sedgeley, J.A. (2001) Quality of cavity microclimate as a factor influencing selection of maternity roosts by a tree-dwelling bat, Chalinolobus tuberculatus, in New Zealand. Journal of Applied Ecology, 38, 425438.
  • Sirén, M. (1951) Telkkäkannan lisääminen pesäpönttöjen avulla. Suomen Riista, 6, 83101.
  • Sokal, R.R. & Rohlf, F.J. (1981) Biometry, 2nd edn. Freeman, New York, NY.
  • Török, J. & Tóth, L. (1988) Density dependence in reproduction of the collared flycatcher (Ficedula albicollis) at high population levels. Journal of Animal Ecology, 57, 251258.
  • Turchin, P. (1995) Population regulation: old arguments and a new synthesis. Population Dynamics: New Approaches and Synthesis (eds N.Cappuccino & P. W.Price), pp. 1940. Academic Press, San Diego, CA.
  • Wayland, M. & McNicol, D.K. (1994) Movements and survival of common goldeneye broods near Sudbury, Ontario, Canada. Canadian Journal of Zoology, 72, 12521259.
  • Wiens, J.A. (1984) Resource systems, populations, and communities. A New Ecology. Novel Approaches to Interactive Systems (eds P. W.Price, C. N.Slobodchikoff & W. S.Gaud), pp. 397436. John Wiley & Sons, New York, NY.
  • Wiens, J.A. (1989a) The Ecology of Bird Communities, Vol. 1.Foundations and Patterns. Cambridge University Press, London, UK.
  • Wiens, J.A. (1989b) The Ecology of Bird Communities, Vol. 2.Processes and Variations. Cambridge University Press, London, UK.
  • Wilkinson, L. (1992) systat: The System for Statistics. systat Inc., Evanston, IL.
  • Williams, M.R., Abbott, I., Liddelow, G.L., Vellios, C., Wheeler, I.B. & Mellican, A.E. (2001) Recovery of bird populations after clearfelling of tall open eucalypt forest in Western Australia. Journal of Applied Ecology, 38, 910920.
  • Zar, J.H. (1996) Biostatistical Analysis, 3rd edn. Prentice Hall, Upper Saddle River, NJ.