By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Calcium is an essential nutrient for avian reproduction. Calcium-rich foods are consumed by breeding birds for production of eggshells and for provisioning chicks that are mineralizing skeletal tissues. A number of studies have documented calcium-limited reproduction, and calcium supplementation has been employed over the last decade to demonstrate degrees, causes and consequences of calcium limitation. However, supplementation studies have produced equivocal findings resulting from an absence of calcium limitation in the study species, a poorly designed supplementation procedure or both. Prior to effective calcium supplementation, many factors need to be considered. Calcium-limited breeding in birds can only be detected by monitoring breeding attempts for more than one year and by ensuring that the measured breeding parameters are sensitive to calcium availability. Natural calcium availability needs to be estimated, and daily calcium budgets for the appropriate reproductive stages determined for the study species. Most crucially, if calcium limitation of breeding is caused by secondary calcium limitation (e.g. through heavy metal toxicity), calcium supplementation will probably be ineffective. Effective calcium supplementation will then be achieved through careful planning – a study over several years using appropriate supplements (i.e. naturally occurring ones used by breeding birds), applied at the appropriate time of year (i.e. prelaying and/or chick-rearing phases) and using a response variable that is highly sensitive to calcium availability. If properly planned and performed, calcium supplementation is a cost-effective and potent tool for the study of bird breeding biology.
Calcium is probably the most important micronutrient necessary for successful breeding in birds (Reynolds & Perrins 2004). Up to 98% of the dry mass of the avian eggshell consists of a crystalline form of calcium carbonate called calcium hydroxyapatite. Among its numerous functions, the eggshell prevents the incubating adult bird from crushing the egg contents, resists the entry of pathogens into the egg, controls the exchange of gases between the egg contents and the nest microenvironment, and provides a source of calcium to the embryo for early skeletal mineralization. For many species, the requirements for dietary calcium remain high during postnatal development of chicks, when mineralization of their skeletons continues (Starck 1998). Precocial chicks forage extensively for calcium-rich foods, whereas altricial chicks are fed calcareous food in the nest by their parents (see table 3 in Graveland 1996).
Apart from a few isolated reports on trace elements and their importance in the diet of egg-laying poultry (e.g. copper, Baumgartner et al. 1978), micronutrient limitation on breeding performance of birds has been largely neglected until recently. The focus on calcium was sharpened, however, with the publication of a paper by Drent and Woldendorp (1989), who reported a sharp decline in the breeding performance of Great Tits Parus major resident in the Buunderkamp forest in The Netherlands. Birds laid eggs with no shells or with such thin shells that embryos died from desiccation due to excessive and uncontrolled water loss during incubation. The authors found similar eggshell defects in resident Blue Tits Parus caeruleus, Coal Tits P. ater, Wood Nuthatches Sitta europaea and Great Spotted Woodpeckers Dendrocopos major, but not in migratory Pied Flycatchers Ficedula hypoleuca. Drent and Woldendorp (1989) attributed the declines in eggshell quality to low calcium availability on poor sandy soils in their study area, further compounded by the adverse effects of acid rain on calcium retention in base-poor soils with low buffering capacities. Hydrogen ions leach calcium ions from topsoil and the latter are replaced by toxic cations such as aluminium and lead. As pH declines, toxic cation concentration increases and uptake of calcium is impaired. Drent and Woldendorp (1989) offered compelling evidence for a relationship between exogenous calcium availability and breeding performance of birds. For example, they found that eggshell defects were absent in Great Tits breeding on soils of better quality (e.g. loam, clay).
Detection of calcium-limited reproduction of birds is possible in habitats where the decline in calcium availability has been rapid and the impact on breeding performance dramatic and easily monitored. For example, the percentage of territorial Great Tits breeding in calcium-poor areas of the Buunderkamp forest that produced inferior eggshells increased from 12% in 1983 to 57% in 1988 (Drent & Woldendorp 1989), because calcium was lost through severe anthropogenic acidification.
Calcium-limited reproduction can be much more difficult to identify in habitats where the decline in calcium availability has been protracted and gradual. Under these circumstances, ‘snapshots’ of breeding performance over a few years provide no grounds for concern. Longitudinal studies of this duration provide a detailed description of interannual variation for a given species, but ‘saw-tooth’ patterns in reproductive parameters (e.g. laying date) often mask a persistent decline of overall breeding performance that is only detected after many decades of sustained monitoring (e.g. Crick & Sparks 1999). A case in point is the secondary calcium limitation caused by the widespread use of the biocide dichlorodiphenyltrichloroethane (DDT) and the consequent decline of raptor populations in Europe (Ratcliffe 1967). The best-documented case of DDT-related decline is that of the Peregrine Falcon Falco peregrinus in Britain (Mellanby 1992). DDT caused the thinning of eggshells, which subsequently broke during incubation. The incidence of shell breakage was c. 4% in 1939, but rose to c. 39% in 1951. Measurements of Peregrine eggs revealed that eggshell thickness was constant from 1850 to 1947, but then declined rapidly in 1947, coinciding with the extensive application of DDT.
Such studies are valuable, but they are scarce in the ornithological literature. Although studies performed over a number of decades are perhaps the best way to detect subtle fluctuations in exogenous calcium availability, such an approach requires a substantial temporal and financial investment (Green 2002). The detection of calcium-limited reproduction in birds resulting from rapid and steep declines in calcium availability (e.g. through acidification – see review by Graveland 1998) is facilitated by calcium supplementation of birds during the breeding season. A number of such investigations have been performed in the last decade (Table 2), but equivocal findings from such studies have prompted us to write this review. The cost effectiveness and ease of employment of calcium supplementation make it an attractive technique that is effective if care is taken in its use (Fig. 1). Such care has not always been apparent in published accounts of calcium-supplementation experiments. We hope that ornithologists in future will consider the issues that we raise, prior to performing calcium supplementation in the field.
In this review, we present the main findings from calcium-supplementation experiments on birds, and suggest possible reasons for the lack of significant results in some investigations. We then suggest a number of issues that should be considered in the planning of investigations. Readers should note that throughout this review we refer to the methods used by Graveland (1995) in his studies on small passerines (especially Great Tits) in The Netherlands. His study remains the model for calcium-supplementation and, as such, exemplifies the use of many of our suggestions in the planning of an effective calcium-supplementation exercise. Finally, we discuss briefly how findings from calcium-supplementation studies can be applied in ornithology.
CALCIUM SUPPLEMENTATION EXPERIMENTS: HOW EFFECTIVE HAVE THEY BEEN?
Table 2 lists the effects of calcium supplementation on the breeding performance of birds. Although most investigations have reported some measurable improvement in breeding parameters, this has not always been so. Effects of supplementation have been measured in terms of egg traits (clutch and egg size, eggshell thickness, eggshell defects, hatching success), chick traits (brood size, body mass, growth, fledging success) or female body condition.
Since Drent and Woldendorp (1989) first observed the defective eggshell structure of forest passerines, egg properties have been the favoured response variables for many calcium-supplementation studies (Table 2). Most have been successful in producing improvement in at least one egg parameter. Interestingly, relatively few of the food-supplementation studies of birds have reported a significant elevation in clutch sizes, and fewer still have reported an accompanying advancement in laying dates (reviewed in Christians 2002). Nevertheless, the savings in energetic expenditure, and in time, accrued by breeding birds as a result of calcium supplementation might be larger than those from macronutrient (e.g. proteins, fats, etc.) supplementation, which might explain why calcium supplements advance laying and increase clutch size more often than do macronutrients (Turner 1982).
In stark contrast to the studies of Dutch (e.g. Graveland 1995) and Estonian (e.g. Tilgar 2002) researchers, some calcium-supplementation studies have been completely ineffective. Most notable of these was a study of breeding Blue Tits in Scotland by Ramsay and Houston (1999). Their study area in west-central Scotland is among the most adversely affected by acid precipitation in the entire UK and, as such, snail abundance and exchangeable calcium are more depressed than those in the study area of Graveland in The Netherlands. However, provision of supplements resulted in no significant increase in egg mass, egg size, eggshell mass, eggshell thickness, clutch size or hatching success, and in no significant advancement of laying. Ramsay and Houston (1999) found that snail density in leaf litter and the top 2 cm of mineral soil was extremely low (i.e. 0.36 snails/m2), but when they examined the gizzard contents of a female collected during egg-laying, they found large numbers of small snails and calcareous fragments including bones and teeth. These results indicate that calcium availability is difficult to measure reliably, especially when some calcium-rich foods are found by birds in areas that are inaccessible to researchers (e.g. marshes). Calcium supplementation of breeding birds will be ineffective in most cases if natural foods are providing sufficient dietary calcium for the birds’ breeding requirements (see also Johnson & Barclay 1996).
Even studies with captive birds in which calcium availability can be tightly controlled have been of limited value. Reynolds (2001) attempted to identify a dietary calcium threshold below which captive Zebra Finches Taeniopygia guttata would struggle to produce eggs. Removal of all calcium-rich foods (i.e. calcareous grit, cuttlefish bone) for 72 h during egg-laying, leaving only chick flint grit (2.2% calcium by mass) as a mechanical grinding agent, resulted in females showing some calcium stress (i.e. reduced calcium content of shells in successively laid eggs), but no declines in other egg (volume, eggshell mass, thickness and surface area) or laying (clutch size, laying interval) parameters. On the basis of published estimates for calcium content of millet seed and digestive efficiency for calcium, Reynolds (2001) calculated that an average female Zebra Finch would have had to consume more than 1 g of flint grit per day to produce a fully calcified egg. Such grit consumption seems highly unlikely given that the average body mass of this species is only 12.5 g (Zann 1996). Instead, this estimate casts doubt on the accuracy of the digestive efficiency for calcium (calculated as the average provided by El-Wailly 1966 and Lemon 1993, respectively). Further research is required on digestive efficiency if the assimilation efficiency of nutrients is to be calculated accurately.
Calcium supplementation of chicks has been achieved either directly through supplement administration to chicks by experimenters or indirectly via provisioning adults. Direct supplementation of Black Tern Chlidonias niger chicks by Beintema et al. (1997) succeeded because calcium deficiency was identified prior to chick-feeding trials as the major cause of poor chick development and so depressing breeding performance. Post-mortem examinations of tern chicks from acid bogs in The Netherlands, which had died prior to 35 days of age, revealed the occurrence of severe rickets and spontaneous fractures in wing and leg bones. Force-feeding of calcium pills three times a week dramatically improved fledging success and mass gain, and confirmed the researchers’ suspicions that adults were feeding chicks with prey containing insufficient calcium to sustain normal skeletal development.
By contrast, Poulin and Brigham (2001) force-fed calcium to Purple Martin Progne subis chicks and, from a lack of improvement in their growth rates, concluded that growth was not limited by a calcium-deficient diet. Unlike the study of Beintema et al. (1997), Poulin and Brigham (2001) only hypothesized that the growth of martin chicks would be calcium-limited by virtue of their calcium-poor insect diet and an inability of provisioning adults to forage effectively on the ground, where calcareous material was predominantly found. Their failure to detect calcium-limited growth may indeed indicate a genuine absence of calcium limitation at the nestling stage; however, calcium was provided as a solution of Liquid Calcium®, whereas nestlings are usually supplemented naturally with solid forms of calcium (see Graveland 1996). Although nestlings presumably assimilated some of the administered calcium, much of it was probably excreted soon after force-feeding. In most altricial nestlings, the loss of calcium in solution could be high, with both the mean retention time of digesta and the digestive efficiency increasing with nestling age (e.g. Caviedes-Vidal & Karasov 2001). Furthermore, unlike the study of Beintema et al. (1997), in which the efficacy of calcium supplementation was easily measured through the survival of chicks, calcium limitations on development of Purple Martins may be too subtle to detect from the morphometrics taken during the formative days of the nestling phase.
It is possible to supplement chicks successfully by manipulating the diets of provisioning adults. Richardson et al. (1986) found that 6.8% of Cape Gyps coprotheres and 1.0% of African White-backed Vulture Gyps africanus chicks were suffering from osteodystrophy (metabolic bone disease) in ranched areas of South Africa where hyaenas were absent. Hyaenas were the only carnivores that regularly chewed the bones of ungulate carcasses rendering sufficiently small bone fragments for adult vultures to feed their chicks. In wild areas, where hyaena-produced bones were present, Richardson et al. (1986) found no chicks suffering from osteodystrophy. The establishment of ‘bone restaurants’, where crushed skeletons were provided for birds in ranched areas, resulted in a dramatic decline in the incidence of osteodystrophy in Cape Griffon chicks.
Few studies, besides those discussed above, have provided supplements at the chick stage in isolation from other stages of reproduction. However, a number have supplemented prelaying adults and then investigated whether supplementation has protracted benefits to nestlings. Results have been equivocal. Johnson and Barclay (1996) found that supplemented House Wrens Troglodytes aedon raised fledglings with slightly longer feathers than unsupplemented birds, but they did not produce more, or heavier, fledglings. Eeva (1996) found that calcium-supplemented Pied Flycatchers produced nestlings with longer wings than did unsupplemented birds. Supplementation of Great Tit and Pied Flycatcher females in Estonia resulted in significant increases in chick growth in both species and in fledging success of the former (see Table 2). Furthermore, Tilgar et al. (2004) examined the activity of alkaline phosphatase (ALP) in the plasma of 15-day-old Great Tit nestlings that had been raised by calcium-supplemented females and found that enzymic activity was reduced compared with the control (unsupplemented) group, possibly indicating lower osteoblastic activity and more complete ossification. At the same time, morphological parameters (tarsus length and body mass) did not differ significantly between these two treatment groups, perhaps due to compensatory growth in the prefledging stage. Thus, bone-ALP activity might provide a more informative measure of skeletal development, and therefore of overall chick development, than morphometrics.
Female body condition
The condition of the breeding female in calcium-supplementation studies has rarely been studied. Mänd and Tilgar (2003) captured Pied Flycatcher and Great Tit females during the second half of the nestling period. They found that, in female flycatchers laying large clutches, calcium supplementation improved body condition compared with unsupplemented females. Flycatchers breeding in Estonian forests appear to invest in the current breeding attempt at the expense of their own body condition. However, calcium supplementation did not improve the body condition of tits breeding in the same habitat. Results concerning the significance of female body condition in influencing the timing of reproductive events in birds remain equivocal. Whereas female body condition drives reproductive ‘decisions’ in some species (e.g. Ruddy Duck Oxyura jamaicensis, Alisauskas & Ankney 1994), Hõrak et al. (1999) found that female Great Tits that deserted broods were in better nutritional condition (higher body mass, higher levels of free fatty acids) on day 8 of the nestling period than were non-deserters. Declines in body mass of reproductive female birds may indicate physiological stress (e.g. Drent & Daan 1980), but, alternatively, they may reflect an adaptive response allowing energetic savings to be made during breeding (e.g. Freed 1981). We encourage researchers to make contemporaneous measures of female body condition and other reproductive parameters during calcium supplementation.
PLANNING OF CALCIUM-SUPPLEMENTATION STUDIES
For calcium supplementation during the breeding season to be informative, calcium must be constraining the breeding performance of the study species. Although this might seem obvious, all too often calcium-supplementation studies have been based on equivocal findings (see above). In some instances, calcium limitation is obvious prior to concerted investigation (e.g. Graveland et al. 1994) and supplementation indicates the extent of calcium limitation on breeding output. However, sometimes calcium supplementation has itself been used as an exploratory tool to identify whether calcium was the constraint on breeding performance out of a number of potential constraints. Below we discuss important aspects of experimental design that should be addressed prior to undertaking a calcium-supplementation study.
Duration of study
We recommend that the breeding performance of individuals be monitored for a number of successive years. Although within a given year an improvement in breeding performance with calcium supplementation may not be statistically significant (e.g. see table 1 in Tilgar et al. 2002), dietary calcium availability may be detected as the constraint on avian reproduction when several years of a study are considered together. The degree of improvement in reproductive success as a result of nutrient supplementation is highly sensitive to environmental conditions, including climatic events and natural food availability in any given year. For example, Tilgar et al. (1999a) found that calcium supplementation only advanced laying dates of tits and flycatchers significantly in years when the onset of breeding was relatively late. Presumably, in such years, natural food availability was scarce and unsupplemented females had to invest significantly more time and energy in calcium-specific foraging than they did in years when natural food availability was higher or if they had been supplemented (Graveland & Berends 1997).
The study site
Rarely are study sites established for the sole purpose of carrying out calcium-supplementation work. Therefore, such work must be designed thoughtfully within the constraints of an established study area. For example, the breeding performance of White-throated Dippers Cinclus cinclus varies markedly with stream acidity in upland Wales (Ormerod et al. 1991), and to test fully the merits of calcium supplementation at least ten acidic streams (five treatment, five control) and ten circumneutral streams (five treatment, five control) would be required (S.J. Ormerod, pers. comm.). In practice, identifying 20 such sites is difficult and experimental designs require modification if calcium supplementation is to be feasible within the limitations imposed by a study site.
Nevertheless, another example from Estonia demonstrates that with careful planning it is possible to perform effective calcium supplementation within a heterogeneous study area. Figure 2 illustrates the study area in southwestern Estonia where calcium supplementation has been performed successfully for a number of years by two of the authors (R.M. and V.T.). Full details of the study area and the supplementation procedure are given in Tilgar et al. (2002) but, briefly, Great Tits were supplemented by providing calcium-rich material at their nestboxes. Nestboxes were arranged in lines, with each line usually consisting of tens of nestboxes, confined within a homogeneous habitat type (i.e. coniferous or deciduous; Fig. 2). Each nestbox line was divided into alternating supplemented and control blocks, each block consisting of roughly five consecutive nestboxes (see lower panel in Fig. 2). For a multiyear supplementation study, each year the first block of each nestbox line was randomly assigned as supplemented or control. Although nestboxes were 50–60 m apart, birds from control nestboxes might have taken supplemental calcium from feeders at adjacent supplemented nestboxes, and therefore data from control birds nesting within 100 m of a supplemented block were disregarded.
For some species, the calcium demands of reproduction can be substantial. For example, a typical clutch of Blue Tit eggs can contain more calcium than the laying female's entire skeleton (Perrins & Birkhead 1983). Birds that regularly consume calcium-rich diets, such as raptors that eat skeletons of vertebrate prey, do not shift diet as they enter the breeding season. However, many species that routinely consume calcium-poor foods (e.g. granivores, insectivores, frugivores), and especially small passerines that are incapable of carrying substantial reserves of calcium, change their foraging strategies to consume calcium-rich foods during egg production and chick-rearing.
The calcium-rich supplements consumed by birds when breeding are numerous (see table 1 in Reynolds & Perrins 2004). Therefore, the detection of calcium limitation in one species is not evidence that other members of the avian community are also calcium-limited during reproduction. For example, Graveland et al. (1994) found that Great Tits breeding on calcium-poor soils in Dutch forests exhibited severe calcium-limited reproduction, whereas Pied Flycatchers occupying the same habitat showed no signs of calcium deficiency during breeding attempts. Analysis of stomach contents and droppings, and observations of chick-provisioning by adults, revealed that tits and flycatchers took different calcium-rich materials during their breeding seasons (Graveland 1995). Great Tits responded to a scarcity of snail shells, their principal source of dietary calcium, by taking anthropogenic materials such as Domestic Chicken eggshell and grit. By contrast, Pied Flycatchers consumed many more millipedes Diplopoda spp. and woodlice Isopoda spp. and far fewer snail shell fragments. Woodlice and millipedes contain significantly more calcium than do other forest arthropods (see table 10 in Graveland & van Gijzen 1994). Furthermore, Great Tits and Pied Flycatchers exhibit different responses to calcium supplementation. Calcium supplements (fragments of snail shell and chicken eggshell) were more scarce in the stomachs of egg-laying female flycatchers, in the droppings of nestlings and in nestbox contents than they were in tits. They were also taken less frequently from feeders attached to nestboxes and were fed less frequently by parents to nestlings (Graveland 1995).
Differences between the life histories of the two species might explain the different responses to calcium deficiency and therefore to calcium supplementation. Mänd and Tilgar (2003) suggest that, in response to low calcium availability in a given season, it may be advantageous for Great Tits to restrain reproductive effort and thereby reduce breeding success, whereas Pied Flycatchers may benefit from investing more heavily despite reproduction being more costly than in years when calcium-rich foods are more abundant.
Multiple breeding attempts
We recommend that breeding attempts other than just the first are monitored in multi-brooding species. Although they studied second broods in only one year, Tilgar et al. (2002) found that calcium-supplemented Great Tit pairs produced significantly larger clutches and fledged more young from second broods than did unsupplemented pairs. They concluded that calcium limitation of Great Tits breeding in Estonian deciduous and coniferous forests is not a transient phenomenon, resulting from the source of dietary calcium being ephemeral, but instead that it persists from early spring through to mid-summer. This is supported by studies in Wytham Woods, near Oxford in the UK, which have found that snail abundance declined significantly towards mid-summer (e.g. Phillipson & Abel 1983).
Reproductive biology of individuals
We recommend that researchers carefully consider the variation in reproductive biology of individuals within and between study populations. Differences between individuals might explain why calcium supplementation has sometimes been ineffective when results are considered at the population level. For example, Sandberg and Moore (1996) suggested that migrant American Redstarts Setophaga ruticilla that arrived on the breeding grounds with greater fat loads experienced greater reproductive benefits, both direct and indirect, than those with smaller fat loads. They further suggested that fat buffered the time budgets of individuals, potentially allowing them more time to forage for specific nutrients (e.g. calcium) whose deficiency could limit reproduction. The findings of Sandberg and Moore (1996) highlight a poorly considered manifestation of calcium supplementation under conditions at which dietary calcium availability does not appear to be limiting. Any savings in time or energy that a bird can make in calcium-specific foraging, which occupies a non-trivial amount of time (e.g. Turner 1982), can be invested in other behaviours (e.g. O’Halloran et al. 1990) and foraging for foods that are important sources of other nutrients (e.g. amino acids, see Ramsay & Houston 2003). In years when such foods are limiting, calcium supplementation will have a dramatic effect on the breeding performance of birds, despite natural calcium availability seemingly remaining stable between years.
As Graveland (1995) acknowledges, we have limited knowledge of the specific calcium sources favoured by breeding birds. We urge researchers to survey the study area thoroughly to determine the variety of calcareous materials available to birds. This is particularly important in species that range widely during the breeding season to meet the nutritional needs of egg-laying and chick-rearing (e.g. hirundines, Turner 1982).
Once local calcium availability has been determined, breeding birds can be studied to learn which calcium-rich foods are favoured during egg production and chick-rearing. Birds use olfactory and visual stimuli to forage for calcium-rich material (Hughes & Wood-Gush 1971) and, consequently, birds should be provided with supplements that occur naturally on their breeding territories. For example, Graveland et al. (1994) provided Great Tits with fragments of snail shell and eggshells of Domestic Chickens, calcareous foods that were available in the study area and were used by resident birds. Supplements were provided in open cups attached to the outside of nestboxes and cups were checked three times a week to confirm that supplements were being used. This approach allows identification of food items preferred by breeding birds, by examining the relative losses of known quantities of the candidate supplementary materials. As a last resort if feeding trials fail, vacated nests can be visited at the end of the breeding season when a search of the nest contents may reveal undigested calcareous food items.
Most supplementation studies (see Table 2) have involved species that breed in nestboxes and therefore the methods of Graveland et al. (1994) have been adopted (see above). However, Ramsay and Houston (1999) also supplemented Blue Tits by hanging cuttlefish bone close to nestboxes, an approach that has also been used successfully to supplement Red-cockaded Woodpeckers Picoides borealis (R. Bowman pers. comm.). Somewhat surprisingly, Dhondt and Hochachka (2001) found that across North America, 31 species used calcium supplements provided on the ground and/or on feeder platforms. This group included corvids, species that will readily exploit novel food sources (e.g. Florida Scrub-Jay Aphelocoma coerulescens–Reynolds et al. 2003), but Dhondt and Hochachka (2001) reported that even arboreal birds, such as parids and woodpeckers, foraged on the ground for calcium supplements.
So far our discussions have focused on the diet of female birds prior to and during egg-laying. It is more problematic to determine supplement use during the nestling period. Although altricial species feed chicks in nests, and supplements can be administered in the same way during both the laying and the chick-rearing periods, there is no guarantee that supplements taken by parents during the latter are always fed to chicks. Rarely is it possible to observe parents feeding their chicks and, even then, it is difficult to identify individual food items with any confidence. However, there are alternatives (see review by Rosenberg & Cooper 1990). Faecal analysis and neck ligatures have been used successfully in the past to study nestling diet (e.g. Bureš & Weidinger 2000, Moreby & Stoate 2000). Furthermore, regurgitation of recently ingested food items can be induced by administering emetic drugs (e.g. Johnson et al. 2002) or by oesophageal massage (R. Bowman pers. comm.). It is advisable to study the composition of nestling diet by using one of these techniques if the calcium supplementation of chicks is to be effective (see below).
A good approach for studying the use of different calcium supplements by breeding birds in both egg-laying and nestling periods was employed by Bureš and Weidinger (2003) for Collared Ficedula albicollis and Pied Flycatchers. They found that the diet of free-living birds was significantly richer in woodlice and millipedes than it was in snails. They also found that birds in aviaries bred poorly when provided with snail shell and eggshell but breeding performance was dramatically improved (two- to three-fold) when woodlice were also provided.
Timing of supplementation
Some species (e.g. Red Knot Calidris canutus, Piersma et al. 1996; early breeding Ruddy Duck, Alisauskas & Ankney 1994) deplete endogenous reserves of calcium for reproduction and therefore calcium supplementation might prove totally ineffective as a technique to investigate calcium logistics for these so-called capital breeders (Drent & Daan 1980). However, in many species, the timing of calcium-specific foraging coincides so precisely with the onset of egg-laying that it is highly likely that the ingestion of such calciferous foods provides much of the calcium required for egg production (Reynolds 1997). Indeed, Graveland and Berends (1997) removed snail-shell fragments and then reintroduced them to captive Great Tits and found that curtailment and resumption of egg-laying, respectively, were highly sensitive to dietary calcium availability. Removal of snail shell after the first egg was laid resulted in eggshell defects or interruptions in laying after 1–3 days, and many females resumed laying normal-shelled eggs within a day of snail shell being reinstated.
Although extensive information is available on the dietary intake of calcium during reproduction by poultry species (e.g. see review by Etches 1987), such information is much more limited for wild birds. However, estimates of calcium intake have been obtained for wild birds breeding in captivity where dietary intake of food can be closely monitored. For example, Graveland and Berends (1997) found that captive Great Tits consumed a constant 4 mg of calcium per day in a 2-week period preceding clutch initiation, but they increased their daily dietary intake of calcium dramatically to 65 mg when the first egg was laid. Calcium intake was then maintained until the clutch was complete.
Regardless of whether researchers investigate macro- or micronutrient limitation using food supplementation to manipulate local food availability, food supplements must be provided to birds at an appropriate stage of the annual cycle. In one of the most effective food-supplementation studies to date, Reynolds et al. (2003) advanced laying and increased clutch and egg size by providing Florida Scrub-Jays with food supplements 2 months prior to clutch initiation for two consecutive years. Although Reynolds et al. (2003) used supplements rich in fats and proteins (i.e. macronutrients), the timing of calcium supplementation might be just as important in influencing reproductive parameters. Although largely untested, where calcium is the nutrient limiting breeding performance, we might expect that calcium supplementation of birds early in their breeding seasons would influence the size of individual eggs as well as perhaps clutch size (Tilgar et al. 2002).
The comments above apply to timing of calcium supplementation in relation to the onset of egg-laying, but we encourage researchers also to consider whether dietary calcium limits the reproductive success of birds through the growth and survival of their chicks. The calcium requirements of the chick increase dramatically during the first 2 weeks after hatching. Precocial young can feed themselves almost as soon as they hatch. MacLean (1974) found bones and teeth of Brown Lemmings Lemmus trimucronatus in the stomachs of juvenile Dunlins Calidris alpina and Pectoral Sandpipers C. melanotos in July, when most growth occurs. In altricial species, adults deliver calcium-rich food to chicks before they fledge. Bilby and Widdowson (1971) found that between hatching and 12 days of age, Common Blackbird Turdus merula chicks increased in body mass by 14 times and in total somatic calcium content by 100 times. They also found that the gut contents of Common Blackbird and Song Thrush T. philomelos nestlings sometimes contained half of all the calcium in their entire bodies.
Once calcium-rich food sources for chicks have been determined (see above, Bureš & Weidinger 2000) for a species in which calcium-limited skeletal development in chicks is suspected, researchers intending to feed chicks directly should initially determine the time of maximal skeletal development for their study species. Skeletal development is relatively predictable and fixed across the altricial–precocial spectrum in birds (see Starck 1998) and, at the very least, calcium supplements should be provided when daily calcium intake rates of chicks are maximal. Indirect supplementation of chicks should be effected by providing ad libitum calcium to parents throughout the nestling period when adults determine the calcium intake rates of chicks through their provisioning behaviour.
Cases of calcium-limited reproduction reported in the ornithological literature in the last 15 years (Table 1) have almost exclusively applied to birds breeding in acidified habitats, due to the dramatic reduction of exogenous calcium under such circumstances (Graveland 1998). Graveland and Drent (1997) recognized that calcium-limited reproduction of birds may not be the exclusive domain of such areas, but might also occur in non-acidified, naturally calcium-poor areas of the world (e.g. large areas of northern Europe (Fennoscandian block) and eastern North America (the Canadian Shield)).
To date, only a few studies have been conducted in naturally calcium-poor areas where acidic rocks and soils support few calcium-rich food items, but their findings should stimulate further work in this area. The default assumption for such areas has typically been that breeding birds have had sufficient time to produce an adaptive reproductive response to the prevailing low calcium availabilities (Graveland & Drent 1997). However, the most compelling evidence to refute this has emerged from Estonia, a country predominantly covered with pine forests on naturally acidic (base-poor) soils supporting relatively few snails and other calcium-rich foods for breeding birds (Mänd et al. 2000a). Calcium supplementation in Estonia's northern temperate forests advanced laying (Mänd et al. 2000a, 2000b), and increased the clutch size, fledging success and fledgling tarsus length (Tilgar et al. 2002) of Great Tits, and increased the egg volume, eggshell thickness (Tilgar et al. 1999a, 1999b) and fledgling tarsus length (Mänd & Tilgar 2003) of Pied Flycatchers. Although this study area is not adversely affected by industrial acidification, there was no difference between its snail abundance and that of the forests in acidified areas of The Netherlands (Mänd et al. 2000a).
Ormerod and Rundle (1998) warned against making assumptions about calcium availability based simply upon the pH of the substrate. They found that the calcium content of invertebrate prey did not differ significantly between artificially acidified, limed sites and control sites. Collectively, these findings demonstrate that considerations of natural calcium availability, even in study areas where it has been stable for many decades, are paramount at the planning stage of calcium-supplementation work.
Agents causing calcium loss
Many of the agents causing calcium loss (e.g. pollution, agricultural intensification) have been discussed comprehensively elsewhere and we refer the reader to Graveland (1998) and Pain and Donald (2002) for further detailed information. Increasing the availability of exogenous calcium through the provision of supplementary calcium-rich material will only improve the breeding performance of birds if dietary calcium shortage is the primary cause of reproductive limitation. Therefore, initial planning of calcium supplementation once calcium-limited reproduction has been detected must always involve the determination of the factor(s) limiting calcium availability.
The most straightforward cause of calcium-limited reproduction is a shortage of dietary calcium that directly limits calcium supply to the egg-laying female or chick-provisioning parents. This is exemplified by the constraints on the reproductive output of forest passerines in Estonia (see above) in which there are minimal anthropogenic environmental inputs that act directly and indirectly to bring about primary and secondary calcium limitation, respectively. However, most documented cases of calcium-limited reproduction (Table 1) concern birds breeding in acidified aquatic and terrestrial habitats. Although acidification, and its resulting increase in the biological activity of toxic cations (e.g. aluminium, lead), causes a direct decline in calcium availability to the breeding female (e.g. Hames et al. 2002), it also lowers calcium uptake across the gut wall. Metal cations occupy sites on transport proteins in the avian gut that are normally occupied by calcium, and thereby enter the bloodstream (Six & Goyer 1970). Toxicity of ingested metal cations is minimal when calcium intake is high, but is severe when calcium intake is low (Scheuhammer 1991). Graveland (1998) argued that it is sometimes very difficult to attribute the poor breeding performance of birds on acidified calcium-poor habitats definitively to either calcium deficiency or metal toxicity.
The other agent that is a major cause of secondary calcium limitation is DDT. Despite its use being restricted in the UK in 1964 and in the USA in 1972, DDT is still used extensively throughout southeast Asia and Central America. Its adverse effects on bird populations of North America are still detectable, and birds at higher trophic levels in sub-Saharan Africa, where DDT use persisted until a few years ago, still experience marked reductions in reproductive success compared with pre-DDT records (e.g. Hartley et al. 1995). The effects and persistence of DDT in relation to avian breeding performance are discussed in Reynolds and Perrins (2004).
APPLICATIONS OF FINDINGS FROM CALCIUM-SUPPLEMENTATION STUDIES
Although the review of Graveland (1998) addressed acidification and its effects on avian reproduction, his recommendations are applicable in a wider ornithological context. He recognized that the findings of many field studies were only correlative in nature and that future experimental work (e.g. calcium supplementation) would be relatively more informative.
Of course, calcium supplementation is not the only way to investigate calcium-limited reproduction in birds. Graveland and van der Wal (1996) employed a 4-year liming programme to demonstrate that declines in snail abundance in Dutch forests had been caused by a loss of soil calcium as a result of anthropogenic acidification. Declines in snail abundance had resulted in increases in eggshell defects of forest passerines. Clearly, however, improvements in calcium availability occur only after many years of liming compared with a few weeks in the case of calcium supplementation. Although liming results in wholesale increases in the availability of calcium at many trophic levels, temporal costs of the procedure are not commensurate with its employment as an exploratory tool to detect calcium limitation. Instead, if used intelligently in directed, planned studies, calcium supplementation could enable us to identify cases of calcium-limited reproduction relatively rapidly. Furthermore, by varying the timing of supplementation and/or the type of supplements presented, it should be possible to identify at what reproductive stage(s) calcium limitation is most acute. Supplement use should also indicate the nutritional requirements for breeding of some species for which the reproductive biology is little understood.
It is hoped that findings from calcium-supplementation studies will facilitate improvement of breeding conditions for those species that are currently exhibiting calcium-limited reproduction. Detecting calcium limitation is the starting point for such an aim, and this will only be achieved through concerted and coordinated monitoring of the breeding success of birds over many years. As Hames et al. (2002) acknowledge, however, a more complete understanding of the processes that give rise to the observed patterns of decline (e.g. reductions in calcium availability) will only be achieved through massive research effort. It was with such an approach that Hames et al. (2002) combined long-term data on breeding attempts by birds (Birds in Forested Landscapes Project and the Breeding Bird Survey) and on acidification (National Atmospheric Deposition Project and the Natural Resources Conservation Service) to show that there was a highly significant negative effect of acid rain on the likelihood of breeding for the Wood Thrush Hylocichla mustelina across North America. Currently, the size of the task ahead of us is inestimable, but so too are the rewards.
We thank Andy Radford and three anonymous referees for numerous valuable comments on the manuscript. Continued financial support for S.J.R. has been provided by the Natural Environment Research Council, the National Science Foundation, the University of Memphis and the University of Birmingham. Continued financial support for R.M and V.T. has been provided by the Estonian Science Foundation.