Printer Friendly

Effects of microhabitat on nest box selection and annual productivity of eastern bluebirds (Sialia sialis) in southeastern Georgia.

ABSTRACT. -- We assessed the effects of 15 microhabitat variables (that is, habitat variables measured within a 10-meter radius of the nest box) on artificial nest site selection and on annual productivity (total number of eggs per nest per year = TOTEGGS) of eastern bluebirds (Sialia sialis) in southern Georgia. Sixty-nine (1990) and 74 (1991) artificial nesting boxes were monitored near Statesboro, Georgia, from March through July during both years of the project.

A Multivariate Analysis of Variance (MANOVA) was used to compare the raw microhabitat variables surrounding selected boxes as opposed to unselected boxes. In both years of the study, microhabitat surrounding selected and unselected boxes did not differ significantly. The raw variables surrounding boxes that were occupied were further recombined using Principal Components Analysis (PCA). Annual productivity (TOTEGGS) was regressed against the resulting first six principal component scores. During the 1990 breeding season, TOTEGGS was significantly correlated with PRIN1 (openness) and PRIN4 (orientation of nest box). In 1991 however, none of the first six principal components were significantly correlated with TOTEGGS.

Our results suggest that nest box selection and annual productivity in eastern bluebirds do not appear to be based on microhabitat variables. Key words: eastern bluebird; nesting; microhabitat; Georgia.


Many aspects of the breeding biology of eastern bluebirds (Sialia sialis) have been studied extensively (Peakall, 1970; Pitts, 1976; White and Woolfenden, 1973; Thomas, 1946; Hartshorne, 1962; Pinkowski, 1975a, 1975b, 1979; Gowaty, 1981, 1983; Gowaty and Karlin, 1984; and Gowaty et al., 1989). However, little quantitative information exists regarding the effects of microhabitat characteristics on annual productivity (total eggs per nest per breeding season). For other species, studies have demonstrated that microhabitat variables could influence nesting success in several ways, including the likelihood of predation (Martin and Roper, 1988) or availability of adequate food supplies (Martin, 1987). Rendell and Robertson (1989) reported that cavity height and floor area influenced the reproductive success of tree swallows (Tachycineta bicolor) nesting in natural cavities.

The relationship between habitat characteristics and nesting success in secondary cavity-nesting species using artificial nesting sites has been studied in few species (Nilsson, 1984). Finch (1989) reported that characteristics of habitat surrounding artificial nest boxes influenced reproductive outcome in house wrens, Troglodytes aedon (a secondary cavity nester) in Wyoming, as did Belles-Isles and Picman (1986) in Ontario. Muldal et al. (1985) found that tree swallows (also secondary cavity nesters) preferred nest sites as far from conspecifics as possible presumably because of their territorial nature.

With this in mind, the goals of our study were to assess, quantitatively, the influence of microhabitat characteristics on artificial nest site selection and annual productivity in eastern bluebirds in southern Georgia.

Specifically, we tested the following two hypotheses: 1) no differences exist in microhabitat variables surrounding selected versus unselected nest boxes; 2) microhabitat did not influence annual productivity.


Our study was conducted in Bulloch County, near Statesboro in southeastern Georgia during the 1990 and 1991 breeding seasons. For a thorough description of the study site see Scott (1992).

In 1990, 59 artificial nest boxes were set out at Meadow Lakes Golf Course, approximately 4.8 kilometers by road southeast of Statesboro. In 1991, five additional nest boxes were placed at this site. In both years, 10 additional boxes were monitored along an established bluebird trail west of Statesboro. This bluebird trail is located approximately 8.1 kilometers from the Meadow Lakes Golf Course.

In 1990, 69 artificial nest boxes (59 at Meadow Lake Golf Course plus an additional 10 along the established bluebird trail west of Statesboro) were monitored throughout the nesting season. In 1991, we monitored a total of 74 nest boxes (64 at Meadow Lakes Golf Course and the 10 along the bluebird trail).

In 1990, all nest boxes were in place by 10 February and monitoring began on 1 March. Prior to the 1991 breeding season, five boxes were moved (these boxes were judged to be too close to other boxes) and five new boxes were placed at Meadow Lakes. All boxes were in place by 15 January, and monitoring began on 20 February. Prior to the actual nesting season, each nest box was checked at least once a week for signs of bluebird activity. Once nesting began (late March or early April) the boxes were inspected at least every other day, to determine when the nesting cycle began, as well as the duration of all phases of the cycle (that is, nest building, egg laying, incubation, and brooding).

In principle, a number of microhabitat characteristics could influence the outcome at a particular box. In selecting the suite of microhabitat characteristics, the goal was to include as much information as possible regarding the habitat in the immediate vicinity (arbitrarily defined as 10 meters) of the nest box. With this in mind we selected 15 variables (see Table 1).

To obtain measurements of the microhabitat variables the area around each nest box was divided into four quadrants following north-south and east-west compass points. Each quadrant was further subdivided into segments within five meters and 10 meters of the box. All vegetative measurements were taken within the 10-meter radius in each quadrant. Analysis of ground cover was determined by placing one 0.25-square-meter quadrat at the five-meter and 10-meter marks within each quadrant. Percent basal cover by bare ground, grasses, and forbs (herbaceous dicotyledonous plants) was estimated in each .025-square-meter quadrat. To obtain percent cover by the substrate variables at five meters and 10 meters, the two values were averaged for the four quadrats at each distance (five meters and 10 meters). A PVC pole marked at 25-centimeter intervals was used to determine the height of the surrounding shrub cover. Percent canopy cover within five meters of the nest box was measured using an ocular tube (Finch, 1989).

Distances to the nearest occupied nest box and nearest human dwelling were measured with a (6-1000 meter) rangefinder, and orientation of the nest box opening was determined with a compass.

We organized boxes into two categories for statistical analysis: 1) unselected boxes showed no sign of bluebird activity; 2) selected boxes showed some evidence of bluebird activity (that is, from the mere placement of straw in a box to a completed nest); 3) occupied boxes were those which produced at least one egg; and 4) successful boxes were those from which at least one young fledged.

To test our hypotheses we used a Multivariate Analysis of Variance (MANOVA) (Tatsuoka 1971) on the 15 microhabitat variables to compare: 1) selected versus unselected boxes; and 2) successful boxes versus the combination of unsuccessful and unoccupied boxes.

We further tested the effects of microhabitat on annual productivity, using multiple regression. We recombined the raw microhabitat variables surrounding occupied boxes with a Principal Component Analysis (Harman, 1976).

We factored the correlation matrix of raw microhabitat variables to obtain our principal components, and retained only those components with eigenvalues greater than 1.0 (Wiens and Rotenberry, 1980).

We then tested the relationship between annual productivity and microhabitat variables by regressing total eggs produced per box per season (TOTEGGS) against the principal component scores.


Because the microhabitat measurements in the two years were taken by different individuals, we did not pool the data. Our results are presented below.

In the 1990 breeding season, 31 boxes were selected; of these 24 were occupied, and young were fledged successfully in 18. Of the 24 occupied boxes, 10 produced one clutch, 13 produced two clutches and one produced three clutches. One hundred sixty-eight eggs and 85 young were produced resulting in a nesting success (number fledged per number of eggs) of 50.6%.

The mean clutch size was 4.3 eggs (N = 39 clutches), with clutch size ranging from two to five eggs. Four clutches contained two eggs, 15 contained four eggs, and 20 contained five eggs. Of the 168 eggs produced, 99 (58.9 percent) hatched and of these, 85 (85.9 percent) fledged young. The first clutch of the season was completed on 30 March 1990, and the last clutch of the season was initiated on 13 July 1990. Mean values and ranges for each of the 15 microhabitat variables for all 69 boxes in 1990 are listed in Table 1.

In our MANOVA, comparing the 15 microhabitat characteristics between the 31 selected and 38 unselected boxes, we found no significant differences between groups (Pillai's Trace = 0.164, P = 0.78). Moreover, significant differences did not exist when the 13 unsuccessful and unoccupied boxes were combined and compared with the 18 successful boxes (Pillai's Trace = 0.551, P = 0.348).

In our analysis of annual productivity we included only the occupied boxes. In our principal component analysis, only the first six principal components had eigenvalues greater than 1.0, so we excluded the remaining principal components (Wiens and Rotenberry, 1980). Correlations between the first six principal components and the raw variables are listed in Table 2. Eigenvalues and percent of total variance accounted for by each of the first six principal components are listed in Table 3. All six of the principal components were significantly correlated with at least one raw variable and all the raw variables were significantly correlated with at least two of the principal components (see Table 2).

PC1 was positively correlated with DTNHS, BG5, and CAN, and negatively correlated with GR5 and GR10. PC2 was positively correlated with CON5, CONM5, and DEC5. PC3 was positively correlated with FOR5, FOR10, and MXSHR. PC4 was positively correlated with DECM5 and can be interpeted as a measure of the density of large deciduous trees; and PC6 was negatively correlated with DTNOB and is a measure of the distance to the nearest occupied box.

We regressed the total number of eggs (TOTEGGS) against the first six principal components to test the hypothesis that a relationship existed between annual productivity (TOTEGGS) and microhabitat (the principal components). However, only PC1 and PC4 explained significant amounts of variation in TOTEGGS.

In 1991, more boxes were available than in 1990 and more were used. Thus, of the 74 boxes monitored, 35 were selected, 32 were occupied, and young were successfully fledged in 27. Of the 32 occupied boxes, 11 produced one clutch, 16 produced two clutches and five produced three clutches. Two hundred forty-two eggs and 123 young were produced during the season. Nesting success (number of young fledged per total number of eggs) was 50.8 percent. Fifty-eight clutches were completed, ranging in size from one to five eggs. Of those, one clutch contained one egg, two contained two eggs, five contained three eggs, 28 contained four eggs, and 22 contained five eggs. The mean clutch size was 4.17 eggs.

The first clutch was completed on 5 April and the last clutch was initiated on 28 July 1991. The incubation period for 27 clutches during this season ranged from 11 to 16 days with an average incubation period of 13.92 days.

In our MANOVAS no significant differences in microhabitat characteristics were found for 35 selected versus the 39 unselected boxes (Pillai's Trace = 0.249, P = 0.243) or for 27 successful boxes as opposed to the combined total of unoccupied and unsuccessful boxes (Pillai's Trace = 0.177, P = 0.993).

In our PCA, we included the microhabitat variables for the 32 occupied boxes. We again retained just the first six principal components. Correlations and eigenvalues are listed in Table 3. A glance at Tables 2 and 3 reveals that the PCA results differed from those derived from the 1990 breeding season.

In 1991, PC1 was positively correlated with GR5, GR10, DECM5, and NOR and negatively correlated with FOR5 and FOR10; PC2 was positively correlated with BG10, MXSHR, CON5, and DEC5; PC3 was negatively correlated with BG5; PC4 was negatively correlated with CONM5 and CAN; PC6 was positively correlated with DTNHS and negatively correlated with DTNOB.

In the regression of TOTEGGS against the six principal components, we found no significant amounts of variation in annual productivity explained by any of the first six principal components (F = 1.31, P > F = 0.29).

Three patterns emerge when the two breeding seasons were compared. First, the selection rate of boxes did not differ significantly between years, but there was an increase in the rates of occupancy and success in 1991. Although there were more successful boxes in 1991, nesting success for the two seasons was approximately equal 50.6 percent (1990) and 50. 8 percent (1991).

Second, the MANOVA's for selected boxes as opposed to unselected boxes showed no significant differences in either year. Likewise, the MANOVA's comparing successful boxes to the combined unsuccessful and unoccupied boxes showed no significant differences for either season.

Third, in 1990, we found that annual productivity varied with PC1 (openness) and PC4 (orientation of nest box), but no such significant relationships were found in 1991.


There are at least four possible explanations for the lack of a significant difference in microhabitat variables between selected and unselected boxes in the two years of our study.

First, it is possible that eastern bluebirds do not select nest sites on the basis of microhabitat variables. They might select nest boxes on the basis of some larger scale habitat characteristic shared by all the boxes such as Martin and Roper (1988) reported for hermit thrushes.

Second, it is possible that too few boxes at the golf course were selected for any patterns in microhabitat preference to emerge. The selection rate of nesting boxes 44.9 percent (1990) and 47.3 percent (1991) was low. The selection rate for just those boxes at the golf course was 39.0 percent for both years, compared to 80 percent (1990) and 100.0 percent (1991) on the long established bluebird trail. If one views the bluebird trail as a control of sorts, it would seem possible that over time the selection rate on the Meadow Lakes site could increase. Because eastern bluebirds remain in Georgia during the winter, they are able to defend nest sites throughout the year. It is possible that newer nest boxes may attract younger, less experienced birds and this may account for the low selection rate. This would also explain the difference in selection rate of the new (Meadow Lakes Golf Course) as opposed to old (established bluebird trail) nest boxes. If this is the case we would anticipate a difference in microhabitat between selected and unselected boxes in the future.

Third, it is possible that another factor may be influencing nest site selection. One possibility is nest site fidelity: eastern bluebirds might simply select nest sites based on nesting success in the previous year. Some studies indicate that males and females tend to return to the same nest site year after year (Laskey, 1939, 1940). Nest site fidelity has been reported in cup nesting species (Lanyon and Thompson, 1986; Benard and LaPointe, 1984).

Fourth, we may have failed to include some unknown, yet important, microhabitat variable in our analyses. We know of no way to evaluate this possibility, and it seems unlikely to us given the diversity of the 15 variables that we included.

Results of our analyses of annual productivity and microhabitat were inconclusive. In 1990, TOTEGGS was negatively related to PRIN1 (openness) and positively related to PRIN4 (orientation of nest box), whereas in 1991, however, no significant relationships were found. We believe that this difference may exist partly because the microhabitat data were collected by different individuals. In any case, it is presently unclear if microhabitat variables influence annual productivity.
TABLE 1. Means and ranges for the 15 microhabitat variables for all
boxes by year. BG5, FOR5, GR5 are average percent cover by bare ground,
forbs and grasses based on four, 0.25-square-meter quadrats; BG10,
FOR10, and GR10 same as BG5, FOR5 and GR5 except this is the average
between five meters and 10 meters of the nest box; MXSHR = maximum
percent shrub cover in four quadrats around the nest box; CON5 and DEC5
are the numbers of coniferous and deciduous trees within five meters of
the box diameter of breast height (dbh) < less than five centimeters;
CONM5 and DECM5 are the same as CON5 and DEC5 except here these are
numbers of trees with dbh more than five centimeters; CAN = percent
canopy cover; DTNHS = distance to nearest human structure; DTNOB =
distance to nearest occupied box; NOR = degrees from North.

Variable 1990 1991
 Mean (range) Mean (range)

BG5 47.2% 42.4%
 (0.3-100%) (11.3-77.5%)
BG10 47.6% 42.4%
 (0.0-100%) (1.3-97.5%)
FOR5 21.8% 21.2%
 (0.0-79.8%) (0.0-81.3%)
FOR10 23.3% 22.7%
 (0.0-82.5%) (0.0-91.3%)
GR5 30.8% 36.7%
 (0.0-87.5%) (0.0-82.5%)
GR10 31.8% 33.7%
 (0.0-99.0%) (0.0-88.8%)
MXSHR 3.8% 2.8%
 (1.0-9.0%) (0.0-6.0%)
CON5 .28 0.1
 (0.0-10.0) (0.0-1.5)
CONM5 0.20 0.1
 (0.0-1.8) (0.0-1.8)
DEC5 1.3 0.6
 (0.0-18.8) (0.0-3.3)
DECM5 1.2 0.6
 (0.0-8.0) (0.0-4.5)
CAN 23.7% 17.7%
 (0.0-94.3%) (0.0-88.8%)
DTNHS 468.8m 192.7m
 (20.0-1500m) (25-800m)
DTNOB 255.8m 269.6m
 (20.0-1500m) (50-1500m)
NOR 156.6* 88.2*
 (0.0-350.0) (0.0-225.0)

TABLE 2. Pearson product moment correlation coefficients between raw
variables and the first six principal components for 1990 and 1991 (in
parentheses) nesting seasons. Only the largest correlation coefficients
per row are listed.


BG5 .89067 * * * *
 (*) (.48032) (-.71194) (*) (*)
BG10 .87772 * * * *
 (*) (.54672) (*) (*) (.47004)
GR5 -.76660 * * * *
 (.80183) (*) (*) (*) (*)
GR10 -.80430 * * * *
 (.78581) (*) (*) (*) (*)
FOR5 * -.53490 .74054 * *
 (-.71237) (*) (.50404) (*) (*)
FOR10 * -.50906 .76274 * *
 (-.66879) (*) (.63901) (*) (*)
MXSHR * * .76172 * *
 (*) (.71242) (*) (*) (*)
CON5 * .76220 .47085 * *
 (*) (.55413) (*) (*) (-.48882)
CONM5 * .77115 * * *
 (*) (.36879) (*) (-.66114) (*)
DEC5 * .68299 .49592 * *
 (.60648) (.58723) (*) (*) (*)
DECM5 .45644 * * * .62019
 (.56973) (*) (*) (*) (.42499)
CAN .64298 * * * *
 (.36076) (*) (*) (-.55013) (.40497)
DTNHS .61313 * * * .49178
 (*) (*) (.44006) (.45005) (*)
DTNOB * .49672 * * *
 (*) (.35813) (*) (.41142) (*)
NOR * * * .69816 *
 (.53730) (*) (.50044) (*) (*)

Variable PRIN6

BG5 *
BG10 *
GR5 *
GR10 *
FOR5 *
FOR10 *
CON5 *
DEC5 *
DTNOB -.60996
NOR .58987

TABLE 3. Eigenvalues and proportions of total variance accounted for by
each of the first six principal component for 1990 and 1991 (in

 Percent of
Principal Component Eigenvalue total variance

PRIN1 4.27711 28.5%
 (3.48670) (23.2%)
PRIN2 3.00222 20.0%
 (2.35072) (15.7%)
PRIN3 2.82706 18.8%
 (1.96963) (13.1%)
PRIN4 1.28196 8.5%
 (1.36525) (9.1%)
PRIN5 1.12005 7.5%
 (1.25767) (8.4%)
PRIN6 1.03136 6.9%
 (0.94580) (6.3%)


We thank Joseph Norris and C. K. Dudley for field assistance. This project was funded by grants from the Georgia Department of Natural Resources, Nongame Small Grants Program, and partly by a Faculty Research Grant to MPM from Georgia Southern University on behalf of the first author.


Belles-Isles, J. C. and J. Picman. 1986. Nesting losses and nest site preferences in house wrens. Condor, 88: 483-486.

Benard, J. and G. LaPointe. 1984. The savannah sparrow territorial system: can habitat features be related to breeding success? Canadian J. Zool., 62: 1819-1828.

Finch, D. M. 1989. Relationships of surrounding riparian habitat to nest-box use and reproductive outcome in house wrens. Condor, 91: 848-859.

Gowaty, P. A. 1981. Aggression of breeding eastern bluebirds (Sialia sialis) toward their mates and models on intra- and interspecific intruders. Anim. Behav., 29: 1013-1027.

_____. 1983. Male parental care and apparent monogamy among eastern bluebirds (Sialia sialis). Amer. Nat., 121: 149-157.

Gowaty, P. A., and A. A. Karlin. 1984. Multiple maternity and paternity in single broods of apparently monogamous eastern bluebirds (Sialia sialis). Behav. Ecol. Sociobiol., 15: 91-95.

Gowaty, P. A., J. H. Plissner, and T. G. Williams. 1989. Behavioural correlates of uncertain parentage: mate guarding and nest guarding by eastern bluebirds, Sialia sialis. Anim. Behav., 38: 272-284.

Harman, H. H. 1976. Modern factor analysis. The University of Chicago Press. Chicago, xx + 487 pp.

Hartshorne, J. M. 1962. Behavior of the eastern bluebird at the nest. Living Bird, 1: 130-149.

Lanyon, S. M., and C. F. Thompson. 1986. Site fidelity and habitat quality as determinants of settlement pattern in male painted buntings. Condor, 88: 206-210.

Laskey, A. R. 1939. A Study on nesting eastern bluebirds. Bird-Banding, 10: 23-32.

_____. 1940. The 1939 nesting season of bluebirds at Nashville, Tennessee. Wilson Bull., 52: 183-190.

Martin, T. E. 1987. Food as a limit on breeding birds: a life-history perspective. Ann. Rev. Ecol. Syst., 18: 453-487.

Martin, T. E., and J. J. Roper. 1988. Nest predation and nest-site selection of western populations of the hermit thrush. Condor, 90: 51-57.

Muldal, A., H. L. Gibbs, and R. J. Robertson. 1985. Preferred nest spacing of an obligate cavity-nesting bird, the tree swallow. Condor, 87: 356-363.

Nilsson, S. G. 1984. The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition. Ornis Scand., 15: 167-175.

Peak all, D. B. 1970. The eastern bluebird: its breeding season, clutch size, and nesting success. Living Bird, 9: 239-256.

Pinkowski, B. C. 1975a. Growth and development of eastern bluebirds. Bird-banding, 46: 273-289.

_____. 1975b. A summary and key for determining causes of nesting failures in eastern bluebirds using nesting boxes. Inland Bird Banding News, 47: 179-186.

_____. 1979. Annual productivity and its measurement in a multibrooded passerine, the eastern bluebird. Auk, 96: 562-572.

Pitts, T. D. 1976. Nesting habits of eastern bluebirds. Unpublished Ph.D. thesis, Univ. Tennessee, Knoxville, 146 pp.

Rendell, W. B. and R. J. Robertson. 1989. Nest-site characteristics, reproductive success and cavity availability for tree swallows breeding in natural cavities. Condor, 91: 875-885.

Scott, M. A. 1992. Effects of microhabitat variables on the breeding success of eastern bluebirds (Sialia sialis) in south Georgia. Unpublished master's thesis, Georgia Southern Univ., vii + 1-58 pp.

Tatsuoka, M. M. 1971. Multivariate analysis: techniques for educational and psychological research. John Wiley and Sons, New York, New York, xiii + 310 pp.

Thomas, R. H. 1946. A study of eastern bluebirds in Arkansas. Wilson Bull., 58: 143-183.

White, S. C., and G. E. Woolfenden. 1973. Breeding of the Eastern bluebird in central Florida. Bird-Banding, 44:110-123.

Wiens, J. A., and J. T. Rotenberry, 1980. Patterns of morphology and ecology in grassland and shrubsteppe bird populations. Ecol. Monogr., 50: 287-308.


Department of Biology, LB 8042, Georgia Southern University, Statesboro, Georgia 30460-8042
COPYRIGHT 1993 Texas Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Scott, Melissa A.; Lockwood, Julie L.; Moulton, Michael P.
Publication:The Texas Journal of Science
Geographic Code:1U5GA
Date:Feb 1, 1993
Previous Article:Effect of feed quality on growth of the Gulf of Mexico white shrimp, Penaeus setiferus, in pond pens.
Next Article:Mammals from the beach mountains of Culberson County, Trans-Pecos Texas.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters