Action

Action Synopsis: Bird Conservation About Actions

Provide artificial nesting sites for songbirds

How is the evidence assessed?
  • Effectiveness
    67%
  • Certainty
    85%
  • Harms
    0%

Study locations

Key messages

  • Only three studies out of 66 from across the world found low rates of nest box occupancy, although this may be partially the result of publishing biases. Thrushes, crows, swallows and New World warblers were the target species with low rates of use. Thrushes, crows, finches, swallows, wrens, tits, Old World and tyrant flycatchers, New World blackbirds, sparrows, waxbills, starlings and ovenbirds all used nest boxes. One study from the USA found that wrens used nest boxes more frequently than natural cavities.
  • Five studies from across the world found higher population densities or population growth rates in areas with nest boxes, whilst one study from the USA found higher species richness in areas with nest boxes. One study from Chile found that breeding populations (but not non-breeding populations) were higher for two species when next boxes were provided.
  • Twelve studies from across the world found that productivity of birds in nest boxes was higher or similar to those in natural nests. One study found there were more nesting attempts in areas with more nest boxes, although a study from Canada found no differences in behaviour or productivity between areas with high or low densities of nest boxes. Two studies from Europe found lower predation of some species using nest boxes. However, three studies from the USA found low production in nest boxes, either in absolute terms or relative to natural nests.
  • Thirteen studies from across the world founds that use, productivity or usurpation varied with nest box design, whilst seven found no difference in occupation rates or success with different designs.
  • Similarly, fourteen studies found different occupation or success rates depending on the position or orientation of artificial nest sites. Two studies found no difference in success with different positions.

 

About key messages

Key messages provide a descriptive index to studies we have found that test this intervention.

Studies are not directly comparable or of equal value. When making decisions based on this evidence, you should consider factors such as study size, study design, reported metrics and relevance of the study to your situation, rather than simply counting the number of studies that support a particular interpretation.

Supporting evidence from individual studies

  1. A replicated study in 1945-6 in garden habitats in Ohio, USA (Calhoun 1948), found that American robins Turdus migratorius nesting in artificial nests had lower success rates than those natural nests (33% success for 24 nesting attempts in artificial nests vs. 50% of 48 in natural nests). Fourteen pairs of robins used the nests, but only seven successfully raised chicks. Nests consisted of cones of black or green roofing paper 17.8 cm at the widest and 5.1 cm deep. This study also examines nest use by mourning doves Zenaida macroura (formerly Zenaidura macroura), and the effect of different coloured nests in ‘Use differently-coloured artificial nests’.

    Study and other actions tested
  2. A controlled study in mixed farmland in north-east Scotland in 1971 (Yom-Tov 1974) found that carrion crows Corvus corone did not nest in artificial trees, irrespective of whether they were provided with supplementary food or not. In one experiment, a line of 15 artificial trees (3-6 m branches tied to fence posts and provided with an old crow’s nest) were set up, approximately 70 m apart. Two pairs of crows established territories, but neither attempted to breed. A second experiment provided a single artificial tree in two occupied territories, 70 m from the tree used by the resident pair. Neither artificial tree was used, as the resident pairs successfully defended their territories. This study also investigated the effects of supplementary feeding on crow reproduction, discussed in ‘Provide supplementary food to increase reproductive success’.

    Study and other actions tested
  3. A small study in 1976-9 in three scrub and grassland habitats in Idaho, USA (Howard & Hilliard 1980), found that common ravens Corvus corax nested on nesting platforms provided, with four pairs using them in 1976, but only a single attempt in 1979. An average of 2.8 chicks/nest were produced. Twenty four platforms were provided in shaded/un-shaded pairs, with 23 out of 29 young fledged from shaded platforms. This study also discusses platform use by ferruginous hawks Buteo regalis, discussed in ‘Provide artificial nest sites for raptors’.

    Study and other actions tested
  4. A replicated study in 1972-8 in Texas, USA (Brown 1981), found that purple martin Progne subis reproductive output did not differ between aluminium and wooden bird ‘houses’ (82% of chicks fledging from 275 pairs in aluminium houses vs. 86% from 116 pairs in wooden houses; average of 28 day nestling period for 82 pairs in aluminium houses and 67 pairs in wooden houses). Aluminium houses had six, 12 or 24 compartments and were commercially available, wooden houses were homemade but of similar designs, with 15 x 15 x 15 cm compartments.

    Study and other actions tested
  5. A small replicated study in 1976-8 in a mixed forest in Washington State, USA (Herlugson 1981), found that mountain bluebird Sialia currucoides reproductive success (clutch size, eggs number hatched and chicks number) did not differ between two nest box designs. Of 22 marked females that successfully bred, 68% returned to the same territory and nest box type and 71% of those that changed territory selected the same type of nest box. Unsuccessful breeders, however, were more likely to change territory and box type in subsequent years (60% five unsuccessful breeders changed territories, 40% also changed nest box type). The majority of boxes had a 30 x 15 cm base and a 4.4 cm diameter entrance hole 10 cm above the floor; the rest had a 12.7 x 12.7 cm based and a 3.8 cm hole 14 cm above the floor. All were erected 1.5 m above the ground on fence posts and 400 m apart.

    Study and other actions tested
  6. A controlled study 1977-9 in riverine forests in Louisiana, USA (McComb & Noble 1981), found that Carolina wrens Thryothorus ludovicianus nested more frequently in nest boxes than in natural cavities (nests found in 0.50% of 5,374 nest box inspections vs. 0.07% of 3,993 natural cavities). Two hundred and thirty five nest boxes were erected, of three sizes: large (60 x 30 x 30 cm with a 13 cm diameter entrance); medium (45 x 20 x 20 cm with a 7.5 cm diameter entrance) and small (30 x 15 x 15 cm with a 5.0 x 7.0 cm entrance). Large boxes had a 15 x 13 x 2.5 cm shelf 8 cm below the entrance. All boxes had 5-10 cm of pine shavings in the bottom. Other songbirds also used nest boxes and cavities, were too infrequent to be analysed. This study also examined nest box use by other birds (wildfowl, woodpeckers and owls).

    Study and other actions tested
  7. A replicated trial in 1979-80 in a deciduous forest in Ohio, USA (Peterson & Grubb 1983), found that two pairs of Carolina chickadees Parus carolinensis and 28 pairs of house wrens Troglodytes aedon nested in cavities excavated from polystyrene cylinders by downy woodpeckers Picoides pubsecens. This study is discussed in more detail in ‘Provide artificial nesting sites for woodpeckers’.

    Study and other actions tested
  8. A replicated study in 1972-80 (Nilsson 1984) found that predation of great tit Parus major and pied flycatcher Ficedula hypoleuca clutches in a mixed forest in Kronoberg, Sweden, was significantly lower in nest boxes than in natural cavities (5% predation for 112 great tit clutches and 4% for 112 flycatcher clutches in boxes vs. 17% of 76 clutches and 23% for 31 clutches in natural nests). There were no differences in predation of blue tit P. caeruleus (alternatively Cyanistes caeruleus) or marsh tit P. palustris (alternatively Poecile palustris) clutches. Three wood nuthatches Sitta europaea nested in boxes, none of which were predated (compared to 6% predation of 113 clutches in natural nests). For the three tit species, woodpeckers were responsible for 48% of predation in boxes, compared with 17% in natural cavities.

    Study and other actions tested
  9. A small single-site study from March-May in 1981-1982 in an island (1.2 ha) within Kentucky Lake in Tennessee, USA (Spero & Pitts 1984) found that common grackles Quiscalus quiscula nested in 20-21% of nest boxes provided, fledging 71% of nestlings, which was higher than in other areas (average of 27%). This study also discusses the use of nest boxes by wood ducks Aix sponsa.

    Study and other actions tested
  10. Three replicated trial in 1982-3 in woodlands in Sweden (Gustafsson & Nilsson 1985), found that pied flycatchers Ficedula hypoleuca and collared flycatchers F. albicollis had significantly larger clutches in larger nest boxes, whereas the effect on fledging success was less clear. A study with four nest box designs in a mixed wood in Kronobergs found that pied flycatchers laid larger clutches in larger nest boxes (6.4 eggs/clutch for 36 clutches in boxes with 57 cm2 basal area; 6.5 eggs/clutch for 62 in 87 cm2; 6.6 eggs/clutch for 16 in 104 cm2 and 7.0 for eight in 125 cm2). Fledging success was lowest in the smallest boxes (51%), highest (79%) in the second smallest and intermediate in the other designs (57% and 62%). Two trials in deciduous woods on Gotland found that collared flycatchers had smaller clutches in small boxes in 1982 only (5.6-5.7 eggs/clutch vs. 6.1-6.2 eggs/clutch). Fledging success was higher in normal boxes in one wood, but higher in smaller boxes in the other wood.

    Study and other actions tested
  11. A replicated study in May-August 1984 in flooded riparian habitats in Tennessee, USA (Petit et al. 1987), found that prothonotary warblers Protonotaria citrea nested in 39% of 301 nest boxes provided, with tree swallows Tachycineta bicolor using three boxes and tufted titmice Parus bicolor one. Three box types were used and prothonotary warblers showed a significant preference for milk cartons (55% of 145 used, 68% of all warbler nests) and an avoidance of wooden boxes (7% of 84 used, 5% of all nests). PVC pipes were used in approximate proportion to their availability (45% of 71 used, 27% of all nests). The authors suggest that the lower volume of milk cartons and PVC tubes might explain these differences. All boxes were erected 1.5-2.0 m above the water surface on trees.

    Study and other actions tested
  12. A replicated study from 1980-1984 in 5 sites of spruce, alder or mixed deciduous woodland in Trondheim, Norway (Slagsvold 1987), found that pied flycatchers (Ficedula hypleuca) prefer to nest in larger, upright nest boxes. Pied flycatchers lay larger clutches in larger (base area 180 cm2) rather than smaller (base area 70 cm2) nestboxes (6.4 compared to 6.7 eggs / nest respectively). When nestboxes were of the same small size within a plot, females laid larger clutches in nestboxes placed in normal, upright position than in tilted ones (6.9 compared to 6.3 eggs / nest). Normal, upright nest boxes were used preferentially across sites (80% of nestboxes occupied). There was no difference in fledgling success between nestboxes. Between 15 and 60 artificial nest boxes (diameter of entrance was 32 mm and 15 cm from base of entrance to bottom) were installed in each site each year.

    Study and other actions tested
  13. A controlled before-and-after study in pine-oak woodlands in Arizona, USA (Brawn & Balda 1988), found the population density six cavity-nesting songbirds more than doubled from on two out of three experimental plots (one thinned, one with 75% of the oak and pine foliage removed), following the installation of 60 nest boxes on each plot in 1979 (21-46 pairs/40 ha in 1973-5 and 1979 vs. 64-108 pairs/40 ha in 1980-3). A third plot with no management showed a small but non-significant increase in population density. Violet-green swallows Tachycineta thalassina, pygmy nuthatches Sitta pygmaea and western bluebirds Sialia mexicana all showed significant increases in population density. Nest box use by five of the species was significantly higher in open and thinned forest plots, with more than 85% of nests of violet-green swallow, pygmy nuthatch and mountain chickadee Parus gambeli in nest boxes, compared to 30-35% in dense forests. White-breasted nuthatches S. carolinensis used a lower proportion of boxes in all habitats (approximately 63% of nests in open and thinned forests in nest boxes and almost none in dense forest). Bluebirds nested almost exclusively in nest boxes in all habitats. Boxes were 1,900 cm3, made of woodcrete with entrance holes 3.2 or 3.8 cm in diameter and placed 5-11 m above ground.

    Study and other actions tested
  14. A replicated, controlled study over two years in deciduous woodland in Oxfordshire, England (East & Perrins 1988), did not find consistent effects of nest box provision on tit Parus spp. reproduction. Great tits P. major and blue tits P. caeruleus (alternatively Cyanistes caeruleus) had significantly higher fledging success in nest boxes in 1983, but not 1984 (77% and 74% of eggs in nest boxes producing fledglings in 1983 vs. 35% and 46% of those in natural nests; 73% and 45% for nest boxes in 9184 vs. 56% and 51% for natural nests). In 1984, the percentage of nests fledging at least one chick was significantly higher for great tits in boxes (82% for boxes vs. 27% for natural nests), but not blue tits (62% and 69%). Differences were smaller in 1983 (91% for both species in nest boxes vs. 75% and 81% for great tits and blue tits in natural nests).

    Study and other actions tested
  15. A replicated paired study in the summers of 1977-84 in farmland in Tennessee, USA (Pitts 1988), found that eastern bluebirds Sialis sialis did not show any preference for, or have higher reproductive success in, large nest boxes, compared to small ones (50 nests built in large boxes, 4.6 eggs/clutch and 2.9 fledglings/nest vs. 44 nests built in small boxes, 4.7 eggs/clutch and 3.0 fledglings/ nest). There was some evidence that more chicks from large boxes returned to the study area, compared to those from small boxes (9% vs. 3% respectively). Ten pairs of boxes were erected 75 cm apart on metal supports. Small boxes had a 71.5 cm2 basal area, large had 143 cm2 and both had a 3.8 cm diameter entrance hole.

    Study and other actions tested
  16. A replicated study in May-June 1983-5 in three riparian woodland sites in Wyoming, USA (Finch 1989), found that 37% of 65 nest boxes erected in 1982 were used by house wrens Troglodytes aedon, with 20% contained unused nests. Wrens successfully fledged chicks from 73% of the 73 nests, were more likely to use boxes in open habitats and more likely to fledge chicks from these boxes. Boxes were 14 x 14 x 28 cm, with a 3.8 cm entrance hole.

    Study and other actions tested
  17. A replicated trial in the summers of 1983-6 at three streamside sites in Wyoming, USA (Finch 1990), found that house wrens Troglodytes aedon used nest boxes erected at all three sites, whilst tree swallows Tachycineta bicolor used boxes on one site. Clutch mortality was higher in swallows (87% of 29 clutches) than wrens (37-67% for 99 clutches), with interference by wrens occurring in 45% of 29 swallow nests. Wren clutch loss was largely due to predation, although the rate of clutch loss was lower on the site where swallows did not use nest boxes.

    Study and other actions tested
  18. A replicated trial in woodlands, hedges and around villages on Nakano-shima island, Kyshu, Japan (Higuchi et al. 1990), found that Ryukyu robins Erithacus komadori occupied approximately 5% of nest boxes in 1989, but 35% in 1990, possibly because more boxes in 1990 had entrance holes larger than 4 cm (in 1989 only 1% of 175 boxes with entrances <3 cm in diameter were occupied vs. 44% of 25 with entrances over 3.5 cm). Occupation rates varied across habitats, from 17% of 24 boxes in village vegetation to 42% in some bamboo stands.

    Study and other actions tested
  19. A replicated trial in a tidal swamp forest in Virginia, USA (Blem & Blem 1991), found that prothonotary warblers Protonotaria citrea nested in 27-34% of nest boxes erected in 1987-8 (140-214 boxes available). Warblers appeared to favour boxes on trees near open water that were surrounded by relatively large trees, but orientation and height above ground did not appear to have an effect. Wasps occupied 29-42% of nest boxes, presumably preventing colonisation by warblers. Boxes were 28 × 9 × 6 cm, with a 3.8 cm diameter entrance hole and erected on trees 20-280 cm above ground.

    Study and other actions tested
  20. A replicated, controlled, paired sites study in 1985-6 in slash pine Pinus elliotti plantations in Florida, USA (Caine & Marion 1991), found that bird species richness and abundance were significantly higher on three sites with snags and nest boxes installed, compared to control plots without them (6.8 species and 9.5 pairs of birds on experimental plots vs. 3.5 species and 4.5 pairs on control plots). Eastern bluebirds Sialia sialis and great crested flycatchers Myiarchus crinitus were both significantly more abundant on experimental plots, with flycatchers using more small boxes (15 x 15 x 20 cm, nine boxes used) than large (25 x 25 x 50 cm, three used). Sixteen snags (8 m tall pine trunks) were erected evenly across plots, each with a small or large nest box attached.

    Study and other actions tested
  21. A replicated, controlled study from May-August in 1985-7 in two woodland ranches consisting sites in Wyoming, USA (Johnson & Kermott 1991) found that house wren Troglodytes aedon males nesting in nest boxes exhibited higher frequencies of polygyny than those males in natural cavities polygyny was greater amongst males in nest boxes (controlling > 1 nest boxes/territory) than those in natural cavities (53 compared to 10% of males respectively). Furthermore, 83% of nest box males and 47% of cavity males made consistent attempts at attracting second mates.  Song output (songs sung / hr) was significantly higher for males attempting polygyny but were similar between polygynous nestbox and cavity males. The authors conclude that, although nestboxes do no cause polygyny in wrens, access to nest boxes appears to facilitate higher rates of polygyny. In 15 experimental sites two or three nest boxes (25-40 m apart) were erected, 31 control sites were also studied.

    Study and other actions tested
  22. A replicated study from February-September in 146 (in 1990) and 180 (in 1991) artificial nest boxes in the USA (Pochop & Johnson 1993) found that 74% of nest boxes were used for egg laying by house sparrows Passer domesticus. sparrows laid an average of 4.5 eggs per clutch. Average hatching success (ratio of eggs laid to nestlings hatched) was 51%. Average breeding success (ratio of eggs hatched to nestlings fledged) was 32%. In 1991, eight of the nest boxes were used by house wrens Troglodytes aedon. By 1991, only 8 of the 120 second-year boxes (but none of the 60 first-year boxes) required replacement. Nestboxes withstood severe fluctuations in temperature, humidity, wind and rain. Entrance holes were 3.8 cm in diameter, roofs and floors had 0.8 cm diameter holes (3 and 5 respectively) for ventilation and drainage.

    Study and other actions tested
  23. A replicated trial in 1988-90 in wetlands in Michigan, USA (Lombardo 1994), found that adult and two year-old female tree swallows Tachycineta bicolor used large and small nest boxes with equal frequency (adults used 27% of large boxes available and 19% of small; two year-olds used 19% and 17% respectively), and that there were no significant differences in reproductive success between nest box types. Boxes were all 19.5 cm tall with a 12.4 cm diameter entrance whole and erected on metal poles (greased to deter mammalian predators). Large boxes had a 14 x 14 cm base, small were 10 x 14 cm.

    Study and other actions tested
  24. A randomized, replicated and controlled before-and-after study at two pine forest sites in Colorado, USA (Bock & Fleck 1995), found that population densities of cavity-nesting birds on 27 experimental plots increased by 500-550% in 1992-3 following the installation of four nest boxes on each plot (7.5-10.0 birds/plot in 1990-1 vs. 41.5-52.5 birds/plot in 1992-3). There were significantly smaller increases on 27 control plots, which had had similar densities before nest box addition (8.5-16.0 birds/plot in 1990-1 vs. 15-20 birds/plot in 1992-3). Birds used 33-55% of the 108 boxes, with the most common species being pygmy nuthatch Sitta pygmaea (25% of box uses), house wren Troglodytes aedon (21%), mountain chickadee Parus gambeli (18%), white-breasted nuthatch Sitta carolinensis (14%) and western bluebird Sialia mexicana (12%). Open-nesting birds also increased in both experimental and control plots. All plots were circles with a 50 m radius.

    Study and other actions tested
  25. Two replicated trials in 1992-4 in woodlots in Ohio, USA (Grubb & Bronson 1995), found that Carolina chickadees Parus carolinensis showed no preference for artificial snags filled with sawdust over those that weren’t filled, but nested more frequently in snags located high above ground (nine snags occupied) and in non-shrub habitat (18% of 38 snags occupied), compared to low snags (two occupied) and in shrub cover (8% of 38 snags occupied). Snags were placed in groups of four (two filled with sawdust and two empty; two in shrub habitat and two in other habitats) or in pairs (one with the entrance 1.2 m above ground, one 3 m above ground). Snags consisted of 1.2-3.08 m sections of 7.8 cm PVC tubing with a nest chamber attached at the top, ‘planted’ in the ground.

    Study and other actions tested
  26. A randomised, replicated trial in 1985 in South Carolina, USA (Plissner & Gowaty 1995), found that a 59% of 94 eastern bluebird Sialia sialis nesting attempts were in 23 territories provided with two nest boxes, compared to 41% in 20 territories supplied with a single box. There were no significant differences in the number of eggs, nestlings or fledglings between territory types. When two boxes were used, they were erected less than 24 m apart, allowing a single male to defend both.

    Study and other actions tested
  27. A replicated, controlled study on mixed farmland and suburban habitats in Punjab, India, in 1992 (Dhanda & Dhindsa 1996), found that Indian mynahs Acridotheres tristis nesting in nest boxes had larger clutches and lower nestling mortality than those nesting in natural sites (4.8 eggs/clutch for 22 clutches and 49% mortality for 68 chicks in boxes vs. 3.9 eggs/clutch for 16 clutches and 68% mortality for 41 chicks in natural sites). However, whilst nesting success and overall productivity were higher in nest boxes, these differences were not significant (64% success and 1.6 fledglings/pair for nest boxes vs. 50% and 0.8 chicks/pair for natural nests) and hatching success was similar (approximately 65%) Mynahs also appeared more likely to lay second and third broods in boxes than natural sites. Thirty boxes were erected, half wooden (22 × 22 × 34 cm) and half PVC tubes (16 cm diameter with a wooden base).

    Study and other actions tested
  28. A replicated study in marshland in 1991-2 in British Columbia, Canada (Rendell & Verbeek 1996), found that tree swallows Tachycineta bicolor occupied 90-100% of 125 nest boxes provided. The effect of cleaning boxes is described in ‘Clean nest boxes to increase occupancy or reproductive success’.

    Study and other actions tested
  29. Three replicated trials in pine forests in north and east Scotland (Summers & Taylor 1996) found that tits Parus spp., with the exception of crested tits P. cristatus preferentially nested in deep nest boxes over shallow ones and empty boxes over those filled with wood shavings. In 1991, trials at two sites found that, of 50 pairs of nest boxes erected (one ‘deep’: 12 x 8 x 25 cm; one ‘shallow’: 11.5 x 10.5 x 15 cm), 15 of the deep boxes were occupied (eight by great tits P. major, five by blue tits P. caeruleus and two by coal tits P. ater) with only a single shallow box occupied by a pair of crested tits. In 1993-4, a second trial at one site found that, of 83 pairs of nest boxes erected (one empty, one with wood shavings, all of the ‘deep’ design), 23 empty boxes were occupied (16 by great tits, four by blue tits and three by coal tits), compared to 12 filled boxes (eleven by crested tits and one by great tits).

    Study and other actions tested
  30. A replicated study in tropical forest in Kanungu District, Uganda (Perrins 1997), found that two or three nest boxes provided in March-April 1996 were used by stripe-breasted tits Parus fasciiventer for nesting, hatching one or two chicks each. Fledging success was unknown and no boxes were used in 1997. Boxes were wooden, 30 cm deep and attached 3-10 m up in trees.

    Study and other actions tested
  31. A replicated, controlled study in 1989-94 in grazed and ungrazed pine-oak woodlands in California, USA (Purcell et a. 1997), found that the benefits of nesting in nest boxes, compared to natural cavities, varied between songbird species. Western bluebirds Sialia mexicana gained the most advantage, with higher nesting success, lower predation rates and marginally more young fledged in boxes. Plain titmice Parus inornatus and house wrens Troglodytes aedon marginally benefited from nesting in boxes, with marginally lower predation rates, more eggs hatched and more young fledged (titmice); or lower predation rates, larger clutches, more eggs hatched, more young fledged and marginally higher nesting success (wrens). Ash-throated flycatchers Myiarchus cinerascens experienced no apparent benefits from nesting in boxes. Nest boxes had a basal area of 137 cm2 with 3.2 or 3.8 cm entrance holes, and placed 2 m above ground on trees. Between 44 (1989-91) and 92 (1992-4) boxes were monitored annually.

    Study and other actions tested
  32. A replicated study between 1968 and 1994 in orchards and fields in Wisconsin, USA (Radunzel et al. 1997), found that eastern bluebirds Sialis sialis were less likely to suffer total clutch loss in nest boxes with screen openings in the roof than in three other designs (16% of 1,506 nesting attempts lost in open-top boxes vs. 28% of 1,066 in standard boxes, 33% of 36 in one gallon tin-cans and 31% of 29 in hollowed posts). In addition, clutches in open-top boxes were larger than those in standard boxes (4.4 eggs/clutch for open-top vs. 4.3 eggs/clutch for standard boxes) and had higher hatching success (82% vs. 72%), chick survival (93% vs. 87%) and overall survival (76% vs. 62%). Open-top boxes were 10 x 13 x 29 cm, with a 9 cm screened opening in the roof; standard boxes were 10 x 13 x 18-21 cm with no roof hole.

    Study and other actions tested
  33. A replicated trial in summer 1992 in a wetland in Ontario, Canada (Stewart & Robertson 1999), found that tree swallow Tachycineta bicolor clutches were significantly smaller in small nest boxes than in large (average of 5.8 eggs/clutch for 11 small boxes and 6.5 eggs/clutch for 12 large) and fewer eggs hatched, although this difference was not significant (72% of eggs hatching in small boxes vs. 84% in large). Once brood sizes were standardised by transferring chicks between nests, there were no differences in the timing of reproduction, chick mass or size or fledging rates (91% for small boxes vs. 97% for large), although nestling flight feathers were shorter after 15 days in small boxes, possibly due to overcrowding. All nest boxes were the same when installed, but after swallows began nesting, small boxes had an insert installed, which reduced the basal area from 178 cm2 to 75 cm2.

    Study and other actions tested
  34. A replicated, randomised study from March-August in 1986-93 in five sites within a suburban park containing 25 solitary (spaced 50 m apart) nest boxes and 25 colonial nest boxes (3-5 m apart) in Budapest, Hungary (Sasvari & Hegyi 2000), found that the reproductive patterns of tree sparrows Passer montanus were affected by nest box spacing. Most fledglings were produced by females that moved from colonial to solitary nests (an average of 0.2 fledglings/brood) and fewest by females that retained colonial nests in subsequent seasons (0.07 fledglings/brood). Overall, female fledglings were significantly more likely to have been hatched from solitary nests (65% of all females) whereas males were more frequent in colonial nest broods (67% of all males). Both female and male recruits were most likely to breed for the first time in colonial nests. The distance between colonial and solitary nest boxes within each study plot was 100 m. The distance between neighbouring study plots was 500 m.

    Study and other actions tested
  35. A replicated study in 1996-8 in a longleaf pine forest in South Carolina, USA (White & Seginak 2000), found that great crested flycatchers Myiarchus crinitus nested in 20% of 330 nest boxes provided, laying eggs in 88% of these (59 boxes in total). Differences in occupation and reproduction between areas burned in different seasons are described in ‘Use prescribed burning’.

    Study and other actions tested
  36. A replicated trial in June-August 1997-8 at 12 sites with differing tree densities in Utah, USA (Lawler & Edwards 2002), found that only 2% of 120 nest boxes erected were occupied by birds: four by tree swallows Tachycineta bicolor and one by mountain bluebirds Sialia currucoides. All five occupied boxes were in sparsely-treed meadows, with none in densely-treed meadows or forests. The low uptake may have been due to large numbers of suitable natural cavities, with 15% of 271 natural cavities surveyed containing bird nests.

    Study and other actions tested
  37. A replicated, controlled study in 1996-9 in coastal forests on Puerto Rico (Lopez-Ortiz et al. 2002) found that 65% of yellow-shouldered blackbird Agelaius xanthomus nests were successful in artificial nests, compared to 48% of those in natural nests. In addition, natural nests suffered higher egg loss (74% vs. 43%) and chick loss (52% vs. 22%) than artificial nests. Nests were PVC ‘elbows’ on top of fence posts, with wire baskets inside and protected from rats by a metal cone attached to the fence post. This study occurred at the same time as a shiny cowbird Molothrus bonairensis eradication programme, discussed in ‘Remove/control brood parasites’.

    Study and other actions tested
  38. A replicated, controlled trial in 14 slash pine Pinus elliotti plantations in Florida, USA (Miller 2002), found that, over 1997-8, estimates of nesting success of great crested flycatchers Myiarchus crinitus were almost identical between nest boxes (37% survival for 32 nests) and tree cavities (38% for 27 nests). Success was higher in boxes in 1997 (53%), but lower in 1998 (26%), due to increased predation during the incubation period. The height of nest boxes and habitat variables did not appear to influence occupation rates of nest boxes, with 32 of 160 boxes across eight plots occupied over the two years.

    Study and other actions tested
  39. A replicated trial in 2000 in a suburban wetland park in Indiana, USA (Dailey 2003), found that 37% of 67 milk-carton nest boxes were occupied by house wrens Troglodytes aedon (24 boxes) and Carolina chickadees Poecile carolinensis (one box). Only 23% of wrens and no chickadees successfully fledged young, with failures due to predation, mostly by mammals that ripped open the cardboard nest boxes. Wrens preferentially nested in boxes on small trees, possibly as a defence against climbing predators. No prothonotary warblers Protonotaria citrea, the target species of the study, nested in any of the boxes.

    Study and other actions tested
  40. A replicated, controlled study in old growth forest in Podlaskie, Poland, between 1993 and 1999 (Mitrus 2003) found no differences in reproductive output of collared flycatchers Ficedula albicollis nesting in artificial nest boxes, compared to those in natural nests (average of 4.9 fledglings/nest for 122 nests in boxes and 150 nests in natural sites).  There were no differences in clutch size, laying date or the percentage of nests losing eggs or chicks, but artificial nests were less likely to be completely predated (0-28% of broods completely lost to predation), compared to natural nests (20-54%).  The authors suggest this may be due to predators not yet adapting to nest boxes.

    Study and other actions tested
  41. A replicated paired study in 1996-7in North Carolina, USA (Stanback & Rockwell 2003), found that eastern bluebirds Sialia sialis showed an overwhelming preference for nest boxes made from woodcrete over those made from wood (90% of 102 pairs picking woodcrete boxes). When given the choice of re-using a soiled woodcrete box or switching to a clean wooden box, 73% of 26 pairs remained in the woodcrete box. The effect of cleaning boxes is discussed in more detail in ‘Clean nest boxes to increase occupancy or reproductive success’. Approximately 100 pairs of boxes were mounted in pairs, on poles 1 m apart, one of woodcrete and one of wood.

    Study and other actions tested
  42. A replicated, paired site study from March-August in 2000-2003 in 20 paired nest box groups (10 placed along wetland edges and 10 in farmlands) in Rutland, England (Field & Anderson 2004) found that tree sparrows Passer montanus showed a significant preference for nest boxes in wetland habitat, compared to those in farmland sites (eight wetland nest boxes colonised vs. no farmland sites). Nest box groups consisted of 5 nest boxes placed 2-20 apart; sunflower seeds were randomly provided to one nest box group within each pair, with the results discussed in ‘Provide supplementary food to increase reproductive success’.

    Study and other actions tested
  43. A replicated study in mixed forests in eastern Ontario, Canada, between 1999 and 2002 (Mennill & Ratcliffe 2004), found that black-capped chickadees Parus atricapillus (also called Poecile atricapillus) only nested in 20-27% of 176 nest boxes provided (compared with 99 nests in natural cavities in the study area). Boxes were erected in sets of four, facing 90o from each other on the same tree, and chickadees showed no preferences for particular orientations. Boxes were made from PVC tubing and filled with sawdust.

    Study and other actions tested
  44. A replicated study from over 800 nest boxes across the eastern USA and Canada from 1998-2002 (Cooper et al. 2005) found that eastern bluebirds Sialia sialis used nest boxes across the study area (between 28° and 48° N). However, breeding pairs exhibited higher net reproductive rates in nest boxes at lower latitudes, having on average 17-33% higher likelihood of repeated egg-laying, multiple brooding and successful fledging events than boxes in the northern range (an average of 1.8 broods/box/pair in the south vs. 1.3 in the north). The average number of fledglings was also significantly greater in the south (1.7 vs. to 1.3 fledglings/box/pair).

    Study and other actions tested
  45. A replicated trial in 1995-2000 in both deciduous and coniferous forests in Pärnu County, Estonia (Mand et al. 2005), found that great tits Parus major laying in nest boxes had significantly higher breeding success in coniferous forests, compared to those in deciduous woods, with heavier fledglings and higher recruitment (8.1 fledglings/pair, 17.5 g/chick and 2% recruitment for coniferous forest vs. 7.7 fledglings/pair, 16.8 g/chick and 1.1% recruitment for deciduous). However, tits laid earlier and had larger clutches and eggs in the deciduous forest (first egg on 29th April, 10.9 eggs/clutch and 1.7 ml/egg for deciduous vs. 30th April, 10.6 eggs/clutch and 1.6 ml/egg for coniferous). They also occupied a higher proportion of nest boxes in deciduous forest, although it should be noted that only half the number of boxes were erected (approximately 20% occupancy of 500-600 boxes in deciduous vs. 9% of 1,200-1,300 in conifers).

    Study and other actions tested
  46. A replicated trial in 1998-2003 in beech, holly and oak forests on Sicily, Italy (Sara et al. 2005), found that blue tits Parus caeruleus (also Cyanistes caeruleus) showed a significant preference for small nest boxes with a 3.2 cm entrance hole, compared to large boxes with a 5 cm entrance. This study also describes the impact of hazel dormouse Muscardinus avellanarius exclusion on blue tit nesting success (see ‘Reduce inter-specific competition for nest sites by removing or excluding competitor species’).

    Study and other actions tested
  47. A replicated trial between 1997 and 2004 in a reedbed in Lancashire, England (Wilson 2005), found that bearded tits (bearded reedlings) Panurnus biarmicus used 13-66% (average of 42%) of the 40-73 nest boxes provided, using boxes more frequently if they were placed over 10-15 cm of standing water. Nest boxes consisted of bundles of reeds tied on to wooden poles and pulled apart to allow tits to enter the bundle and build nests. There were positioned approximately 10 m inside the reedbed and installed in early February.

    Study and other actions tested
  48. A replicated study in open farmland in Massachusetts, USA (Ardia et al. 2006), found that tree swallows Tachycineta bicolor occupied 55% of 153 nest boxes provided in 2004, with an average of five eggs/clutch. Swallows preferentially selected east- and south-facing nest boxes before June, but used nest boxes based on availability later in the summer. East- and south- facing boxes heated up faster in the morning, but only early in the breeding season.

    Study and other actions tested
  49. A replicated trial in arable farming landscapes in Norfolk, England, in the summers of 1997-2001 (Browne 2006) found that tits Parus spp. nested in a higher proportion of hanging woodcrete boxes (38% 48 boxes occupied), compared to tree-mounted woodcrete boxes (25% of 48) or thick and thin wooden boxes (20% and 16% of 48 boxes respectively). Patterns were the same for great tits Parus major, blue tits P. caeruleus (also Cyanistes caeruleus) and all species combined (also including coal tits P. ater (also Periparus ater) and marsh tits P. palustris (also Poecile palustris)), although a higher proportion of great tits used woodcrete boxes (91% of great tits vs. 47% of blue tits). Clutch size, brood size and number of young fledged by blue tits and great tits did not differ significantly between box types. Woodcrete boxes were either attached to a tree trunk (18 cm high, base 18 cm diameter) or free-hanging (19 cm high, base 11 cm diameter). Wooden boxes were 16.5 x 15 x 19.5 cm, and of either 1.9 cm or 2.4 cm thick wood. All designs had a 3.2 cm diameter entrance. Another trial found that a higher proportion of tit Parus spp. nests were in 50 green nest boxes (72% of 41 nests) than in 50 brown boxes (28%), and in 50 boxes with circular entrances (68%) compared to those with a wedge-shaped entrance (32%).

    Study and other actions tested
  50. A replicated trial in oak-savannah and vineyards in California, USA (Fiehler et al. 2006), in 2003-4 , found that western bluebirds Sialia mexicana occupied 56-67% of 120 nest boxes with entrance holes large enough to admit bluebirds. Clutches were larger (5.3 eggs/clutch vs. 5.0) and started earlier in vineyards than savannah. Tree swallows Tachycineta bicolor, house wrens Troglodytes aedon, ash-throated flycatchers Myiarchus cinerascens and violet-green swallows Tachycineta thalassina also occupied small numbers of boxes (all <5%).

    Study and other actions tested
  51. A study in summer 2001 in a hayfield in southeast Ontario, Canada (Male et al. 2006), found there were no differences in tree swallow Tachycineta bicolor nest-building behaviour or reproductive success between areas with a high density of nest boxes (boxes 20 m apart, average of 5.6 eggs/clutch and 3.4 chicks fledged/nest) or a low density (boxes 28 m apart, average of 5.3 eggs/clutch and 3.7 chicks fledged/nest). A total of 52 nest boxes were erected in a grid, 1.5 m above ground on metal poles, although only 22 were used for analysis.

    Study and other actions tested
  52. A small study, as part of a large study in mixed agricultural and woodland habitats in Arkansas, USA (Risch & Robinson 2006), found that, in 2003, a female blue grosbeak Passerina caerulea successfully reared four chicks in a nest box designed for eastern bluebirds Sialis sialis. The authors indicate that, to their knowledge, this represented the first record of cavity-nesting (and hence nest-box use) by blue grosbeak. A total of 200 bluebird nest boxes (with 10 x 15 cm bases) were erected 2 m above ground.

    Study and other actions tested
  53. A replicated, controlled trial in 1999 at a southern beech Nothofagus forest in Maule Region, Chile (Tomasevic & Estades 2006), found that the breeding population densities of thorn-tailed rayaditos Aphrastura spinicauda and house wrens Troglodytes aedon were significantly on higher ten experimental plots, each with five nest boxes installed, compared to ten control plots (4.2 rayaditos/ha and 4.0 wrens/ha vs. 1.0/ha and 1.4/ha for control plots). However, there were no differences in non-breeding season populations. Overall, 22% of boxes were used by rayaditos and 14% by wrens. A further 28% had nesting material in, but no active nest.

    Study and other actions tested
  54. A trial from 2003 to 2005 on a single farm, Rawcliffe Bridge, East Yorkshire, UK (Bryson et al. 2007) found that nest boxes were 54%, 50% and 68% occupied in 2003, 2004 and 2005 respectively. In 2003, all five boxes designed for tree sparrow were occupied. In 2005, 20 tree sparrow Passer montanus boxes (70% of the 28 provided) were occupied. The number of breeding tree sparrows on the farm increased from 6 to 20 pairs between 2003 and 2005. In the years 2003, 2004 and 2005, 32, 60 and 84 bird nest boxes were put up, including some designed for tree sparrows. They were inspected in February each year. Birds on the farm were monitored five times each year from 2003 to 2005, by walking the field boundaries. The number of breeding pairs/ha was estimated from clusters of sightings.

    Study and other actions tested
  55. A before-and-after trial at a wetland site in Cambridgeshire, England (Guilickx et al. 2007), found that between eight and 50 pairs of sand martins Riparia riparia (and one pair of common kingfishers Alcedo atthis) nested annually in 130 artificial burrows drilled in a limestone cliff in 1995. No martins nested in 1996 because the burrows were too small. They were enlarged in 1997 (using a high pressure water jet) to include a nest chamber. The cliff is 3 m high and 80 m long and the burrows 5 cm in diameter and 60-90 cm long.

    Study and other actions tested
  56. A replicated study in the summers of 2005-7 in an area of swamp forest in New York State, USA (Cooper & Bonter 2008), found that black-capped chickadees Parus atricapillus (also Poecile atricapillus) nested in a higher proportion of artificial snags than nest boxes filled with wood shavings (60-70% of 20 snags excavated by chickadees each year and 25-30% used for nests vs. 40-50% and 15% for filled boxes). Chickadees also used more nest boxes filled with wood shavings than unfilled boxes. Nests in artificial snags were less likely to be usurped by mice and no more likely to be usurped by house wrens Troglodytes aedon than nest boxes. Twenty sites were used, 12 with snags and filled boxes and eight with unfilled boxes as well. Snags consisted on 10.2 cm diameter PVC pipes with a 2.8 cm entrance hole and filled with wood shavings.

    Study and other actions tested
  57. A replicated trial in the summers of 2002-6 in suburban habitats in Toledo, Spain (Garcia-Navas et al. 2008), found higher tree sparrow Passer montanus occupancy rates, and higher reproductive success in woodcrete nest boxes, compared to wooden ones (average of 76.5% of 50 woodcrete boxes occupied amd 81% success for 152 clutches vs. 33.5% of 50 wooden boxes occupied and 79% success for 68 clutches). Differences in success were due to earlier clutches, a shorter incubation period and more reproductive attempts per season. Clutch size and nestling condition did not differ between box types. The authors suggest that higher temperatures in woodcrete boxes could explain the differences. One hundred boxes were placed in pairs (one of each material) less than 5 m apart and hung from trees. Wooden boxes were larger than woodcrete ones (2,057 cm3 vs. 1,869 cm3).

    Study and other actions tested
  58. A replicated study in an oak-dominated forest in Gloucestershire, England, between 1990 and 2004 (Goodenough et al. 2008) found that significantly fewer pied flycatcher Ficedula hypoleuca chicks fledged from southwest-facing nest boxes, compared to other orientations. There were no differences for great tits Parus major and blue tits Parus caeruleus in productivity, but great tits showed a preference for certain nest box orientations, with significantly fewer nests in southwest-facing nest boxes. There were no preferences in the other two species.

    Study and other actions tested
  59. A replicated study in acacia stands and shrubland in New South Wales, Australia (Griffith et al. 2008), found that, after the installation of 400 nest boxes in July-September 2005, nearly all breeding attempts by zebra finches Taeniopygia guttata were in nest boxes, with 572 clutches laid in 2005-7 and over 90% of boxes being used at least once. Clutch size (average of 4.9 eggs/clutch for 559 clutches) and fledging success (58% of 522 clutches successful) in boxes was higher than in natural nests (average of 4.0 eggs/clutch for 110 natural nests and 13% success for 84 clutches). Small numbers of southern whitefaces Aphelocephala leucopsis and chestnut-rumped thornbills Acanthiza uropygialis also used nest boxes. Boxes were 14 x 9.3 x 12 cm with a 3 cm entrance hole and erected at 1-1.85 m above ground on trees or steel posts. The only unused boxes were high, far from natural cover and on poles, not trees

    Study and other actions tested
  60. A replicated study at an ancient temple site in rainforest on Java, Indonesia (Kurniandaru 2008), found that Java sparrows Padda oryzivora nested successfully in 20 wooden nest boxes provided (two pairs in 2007, three in 2008), but not in 25 made from bamboo, or 25 of coconut shell. Wooden boxes were 32 x 15 x 23 cms with a 5 cm diameter entrance; bamboo nests were 40 cm sections of bamboo, 10-12 cm diameter with a 5 cm entrance hole; coconut nests were coconut shells with an 8 cm entrance hole. All nests were placed 8-20 m above ground in trees, either on the trunk of amongst the canopy. Two pairs of Javen mynas Acridotheres javanicus also used wooden boxes in 2007.

    Study and other actions tested
  61. A study on Motuihe Island, New Zealand (Parker & Laurence 2008), found that only two of 150 nest boxes provided were used by 20 North Island saddleback Philesturnus rufusater translocated from Matangi Island in August 2005. Saddlebacks also preferred natural cavities over the 110 roosting boxes put out prior to release. The success of the translocation is discussed in ‘Translocate individuals’.

    Study and other actions tested
  62. A small replicated, controlled study from October-January in 2003-6 in two cattle ranches with predominantly Celtis tala woodlots in Buenos Aires Province, Argentina (Llambias & Fernandez 2009) found the southern house wrens Troglodytes aedon bonariae had higher reproductive rates in four nest box sites than in five natural cavity sites (72% of nest box nests produced at least one fledgling vs. 41%  of natural cavities). The overall probability that at least one chick would fledge from nests was 66% in nest boxes and 25% in cavities. Nest box breeding pairs produced more nestlings in each attempt (3 nestlings/nest for nest boxes vs. 1 for natural cavities). The main cause of nest failure was predation. There was no difference between clutch size, brood size, or fledglings produced. The social mating system was unaffected by nest boxes. Nestboxes were wood (30.5 × 16.5 × 12.7 cm) with an entrance-hole 38 mm in diameter.

    Study and other actions tested
  63. A long-term controlled paired sites study (Mand et al. 2009) in the same Estonian sites as in Mand et al. 2005 found that the breeding density of great tits Parus major was significantly higher in areas with nest boxes in, compared to areas with no boxes. This held for all 13 pairs of transects in both deciduous (5.2 pairs/km for six transects in nest box areas vs. 1.8 pairs/km for six in control areas) and coniferous forests (3.3 pairs/km for seven transects in nest box areas vs. 0.2 pairs/km for seven in control areas). Over 1,000 nest boxes were erected in the experimental areas in the 1970s, on trees 1.5-2.0 m above ground.

    Study and other actions tested
  64. A replicated study in eucalyptus stands in farmland in Lower Galilee, Israel (Charter et al. 2010), in 2008-9, found both jackdaws Corvus monedula and house sparrows Passer domesticus nested more frequently in nest boxes with small (7.5 cm diameter) entrances, compared to nest boxes with large (15 x 30 cm) entrance holes (jackdaws: nested in approximately 25% of 49 small-entrance boxes vs. 2% of 51 large-entrance boxes; sparrows: nested in 10% of small-entrance boxes and no large-entrance boxes). This may have been due to competition with barn owls Tyto alba which were able to enter the large-entrance boxes only. Boxes were attached to shaded parts of eucalyptus trees and the positions of large and small boxes were exchanged between years. This study also investigated nest box use by kestrels and owls.

    Study and other actions tested
  65. A replicated, randomised study from January-June in 2007 in 58 semi-arid rural gardens in Jezreel Valley, Israel (Charter et al. 2010) found that great tits Parus major 47% of the nest boxes and succeeded in fledging at least one young in 74% of the 20 nest boxes in which they laid eggs, with an average of five chicks/pair. Breeding density was 5.4 pairs/10 ha, with more nests built in areas of higher tree density and tree species richness.

    Study and other actions tested
Please cite as:

Williams, D.R., Child, M.F., Dicks, L.V., Ockendon, N., Pople, R.G., Showler, D.A., Walsh, J.C., zu Ermgassen, E.K.H.J. & Sutherland, W.J. (2020) Bird Conservation. Pages 137-281 in: W.J. Sutherland, L.V. Dicks, S.O. Petrovan & R.K. Smith (eds) What Works in Conservation 2020. Open Book Publishers, Cambridge, UK.

 

Where has this evidence come from?

List of journals searched by synopsis

All the journals searched for all synopses

Bird Conservation

This Action forms part of the Action Synopsis:

Bird Conservation
What Works 2021 cover

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 21

Go to the CE Journal

Discover more on our blog

Our blog contains the latest news and updates from the Conservation Evidence team, the Conservation Evidence Journal, and our global partners in evidence-based conservation.


Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape ProgrammeRed List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Mauritian Wildlife Supporting Conservation Leaders
Sustainability Dashboard National Biodiversity Network Frog Life The international journey of Conservation - Oryx Cool Farm Alliance UNEP AWFA Bat Conservation InternationalPeople trust for endangered species Vincet Wildlife Trust