TY - JOUR
AU - Strong, Allan, M.
AB - Abstract In hay fields in the northeastern United States, Savannah Sparrows (Passerculus sandwichensis) whose first nests fail as a result of the harvesting of hay renest in the same field in a dramatically altered habitat. We evaluated availability of invertebrate prey in two treatments (harvested and unharvested hay fields) in Vermont's Champlain Valley and assessed potential effects of food resources on clutch size, food provisioning by adults, and growth of nestling Savannah Sparrows. A relative measure of invertebrate biomass (sweep-net samples) showed a 36–82% decline after harvesting, compared with continual increases throughout the nesting season on unharvested fields. Clutch sizes were not significantly different between treatments. Nestling provisioning rates obtained through video observations differed between treatments, with birds on unharvested fields delivering 73% more food than those on harvested fields (P = 0.02). However, food provisioning differences did not translate into differences between treatments in either average nestling mass or the mass of the lightest nestlings within nests. A measure of total biomass (vacuum samples) on harvested fields showed a 28–56% decline after harvesting; this reduction was insufficient to induce food limitation based on the energetic requirements of Savannah Sparrows. Our results suggest that adult Savannah Sparrows must compensate for reduced food availability on harvested hay fields, possibly by increasing the total time spent foraging. Resumen En los campos de heno del noreste de Estados Unidos, los individuos de Passerculus sandwichensis cuyos primeros nidos fracasaron como resultado de la cosecha de heno anidaron nuevamente en el mismo campo en un ambiente severamente alterado. Evaluamos la disponibilidad de presas de invertebrados en dos tratamientos (campos de heno cosechados y no cosechados) en el Valle Champlain de Vermont y estimamos los efectos potenciales de los recursos alimenticios sobre el tamaño de la nidada, el alimento provisto por los adultos y el crecimiento de los pichones de P. sandwichensis. Una medida relativa de la biomasa de invertebrados (muestras de redes de barrido) mostró una disminución de entre 36% y 82% luego de la cosecha, en comparación con los aumentos continuos a lo largo de la estación de anidación en los campos no cosechados. Los tamaños de las nidadas no fueron significativamente diferentes entre los tratamientos. Las tasas de aprovisionamiento de los pichones obtenidas de observaciones de video fueron diferentes entre los tratamientos, ya que las aves de los campos no cosechados entregaron un 73% más de alimentos que las de los campos cosechados (P = 0.02). Sin embargo, las diferencias en el aprovisionamiento no se reflejaron entre tratamientos en el peso promedio de los pichones ni en el peso del pichón más liviano de la nidada. Una medida de la biomasa total (muestras por aspiración) en los campos cosechados mostró una disminución de entre 28% y 56% luego de la cosecha. Sin embargo, esta reducción fue insuficiente para inducir una restricción en la obtención de alimentos según los requerimientos energéticos de P. sandwichensis. Nuestros resultados sugieren que los adultos de P. sandwichensis deben compensar la reducción en la disponibilidad de alimentos en los campos de heno, posiblemente aumentando el tiempo total que gastan forrajeando. Introduction The availability of food resources influences bird distribution, survival, and reproductive success (reviewed in Newton 1998, Nagy and Holmes 2005). During the breeding season, food shortages directly influence the survival and body condition of young (Rodenhouse and Holmes 1992, Siikamäki 1998). Additionally, according to food-limitation theory, scarce food resources should lead to smaller clutch sizes, a greater proportion of time spent foraging by adults, decreased feeding rates, and decreased nestling growth rates (reviewed in Martin 1987). Most temperate-breeding songbirds rely heavily on invertebrate prey to feed their young; this food source provides the protein necessary for growth. Various biotic and abiotic factors influence prey abundance, including predation, parasitism, disease, temperature, and precipitation (Ritchie 2000). However, human activities can also exert a dramatic influence on invertebrate prey through the use of herbicides, insecticides, and mechanical manipulation of invertebrate habitat (Moreby and Southway 1999, Summerville and Crist 2002, Schmidt et al. 2005, Britschgi et al. 2006, Hart et al. 2006, Taylor et al. 2006). This last influence is the focus of the present study. Agricultural lands such as hay fields provide substantial nesting habitat for Savannah Sparrows (Passerculus sandwichensis) in the northeastern United States (Askins 1999). Hay fields in this region are typically cut one to three times during the summer (Troy et al. 2005). During each of these cutting events, nearly all active nests and recently fledged young are killed by the machinery used to harvest the fields or by postharvest predation (Perlut et al. 2006). Cutting also significantly alters vegetation structure from relatively tall grass of varying heights before cutting to short, homogeneous stubble after cutting. Savannah Sparrows are unique among the common grassland birds of this region, in that most individuals maintain breeding territories on a field after it is cut and renest soon after cutting. Thus, for birds whose first nests fail as a result of harvesting, subsequent nesting attempts occur in a dramatically altered habitat that may contain decreased food resources, making the site less suitable for raising young. Food reduction experiments can provide powerful tests of the effects of decreased food resources on breeding birds. Relatively few studies, however, have used this approach, likely because of the difficulty of reducing food resources on a scale large enough to affect highly mobile insect predators such as birds (Boutin 1990). Effects on birds have varied considerably in the few food reduction experiments that have been conducted. As an example of strong effects, (Rodenhouse and Holmes 1992) found that Black-throated Blue Warblers (Dendroica caerulescens) made fewer nesting attempts and nestling diets included fewer caterpillars on a plot with experimentally reduced caterpillar abundance. Likewise, (Rands 1985) found decreased brood size in Gray Partridge (Perdix perdix) on plots that were completely sprayed with herbicides and fungicides compared with plots where small areas were left unsprayed. Other studies have found little to no effect on birds. When caterpillar populations were experimentally reduced in a West Virginia forest, Red-eyed Vireos (Vireo olivaceus) delayed nesting for three to five days, but no other reproductive variables were affected (Marshall et al. 2002). In shrubsteppe habitat, (Howe et al. 1996) found only marginal effects of insecticide-induced arthropod reductions on nestling growth and survival of Brewer's Sparrow (Spizella breweri) and Sage Thrasher (Oreoscoptes montanus). The cutting of hay fields provided a unique opportunity to conduct a large-scale food manipulation experiment in which we could test whether food reduction affected the ability of Savannah Sparrows to raise young. We assessed whether reduction in the vegetative substrate during the harvesting process reduced the standing biomass of invertebrate prey. We examined invertebrate prey in harvested and unharvested hay fields using two methods: sweep-net sampling, which provided a relative index of invertebrate biomass, and vacuum sampling, which provided an estimate of energetic content of invertebrates at the field level. Clutch size, food provisioning by adults, and growth of nestling Savannah Sparrows were examined as potential effects of reduced food availability. Specifically, we predicted that adults on harvested fields would lay smaller clutches and provide less food biomass to nestlings, resulting in reduced nestling mass. Methods Study sites.—We studied Savannah Sparrows during two breeding seasons (2004–2005) on four hay fields located within the Champlain Valley of Vermont. Two fields, Windmill Hill (WH; 15 ha) and Thibault (GT; 19 ha), were harvested during the Savannah Sparrow breeding season. These two fields are hereafter referred to as "treatment fields." In 2004 and 2005, the northern half of GT was cut 7 (2005) to 25 (2004) days earlier than the southern half. Therefore, we treated GT as two independent fields, GT North (GTN; 10 ha) and GT South (GTS; 9 ha). Each treatment field was cut once in 2004 (WH, 5 June; GTN, 7 June; GTS, 2 July). In 2005, WH was cut twice (27 May and 12 July), whereas GTN and GTS were cut once (GTN, 26 June–1 July; GTS, 2 July–10 July). The remaining fields, Elm Marsh (EM; 17 ha) and Taproot (TR; 9 ha), were chosen as control fields, as they were not to be harvested until the breeding season had concluded. However, in 2004, TR's landowner cut the field early in the season (17 May), invalidating it as a true control field for that year. Therefore, we did not include data obtained from TR in 2004 as part of the present study. Windmill Hill and EM were located in Shelburne, Vermont (44.4°N, 73.3°W), and GT and TR were located in Hinesburg, Vermont (44.3°N, 73.2°W). Although we did not conduct a formal analysis of vegetative cover, qualitative statements can be made regarding the vegetative composition of the sites. The vegetation at EM was primarily grass-dominated. The vegetation at WH was a mixture of Alfalfa (Medicago sativa), clover (Trifolium spp.), Common Dandelion (Taraxacum officinale), and grasses. Grasses and clover interspersed with Bird Vetch (Vicia cracca) constituted most of GT's vegetation. Finally, TR's vegetation consisted primarily of grasses and Bird Vetch. Grasses on the four fields were dominated by Orchardgrass (Dactylis glomerata), Timothy (Phleum pratense), bluegrass (Poa spp.), and Reed Canarygrass (Phalaris arundinacea). Nest monitoring.—We marked adult Savannah Sparrows with a unique combination of one federal leg band and three plastic color leg bands. We searched for nests from mid-May through early August of each year, locating 145 nests in 2004 and 111 in 2005. Although most nests (64%) were located during the incubation stage, we also found nests during the laying (21%), nestling (9%), and building (6%) stages. Once a nest was found, we monitored its status and contents every one or two days until nestlings fledged or the nest failed. If nestlings hatched between successive nest checks and the exact hatch date could not be determined, we assigned the median date between the two checks as the hatch date for that nest. For nests that were found during the incubation or nestling stage, we assumed that clutch size was equal to the number of eggs (incubation stage) or nestlings (nestling stage) present when the nest was discovered. Food availability.—To obtain an index of standing invertebrate biomass present on the study sites, we collected sweep-net samples approximately every two weeks from late May through early August. Additional samples were taken shortly before and after fields were cut. We collected sweep-net samples each year at all sites (except TR in 2004; see above), at 10 fixed points per site. These points were located 50 m apart along transects that were established along one of the four cardinal directions. At each point, a canvas net with a 0.5-m-diameter opening was swept through the vegetation 10 times while walking slowly over a distance of ∼ 10 m. Sweep-net sampling was generally conducted between 1000 and 1500 hours EST. We did not sample invertebrates if the vegetation was wet or on very windy days. All sweep-net samples were conducted by N.J.Z., and a conscious effort was made to keep the sweeping velocity constant over all sampling occasions. To allow time for the vegetation to recover between sampling occasions, we rotated the sampling direction from each sample point through the four cardinal directions (e.g., N, S, E, W for sampling occasions one through four, respectively). The procedure for processing the sweep-net contents varied between the two years of the study. In 2004, all invertebrates >3 mm in length that were captured by the sweep net were counted in the field and identified to 1 of 12 prey categories: Formicidae, Coleoptera (adults), Diptera (adults), Hemiptera, Homoptera, Hymenoptera (adults; excluding Formicidae), holometabolous larvae (primarily Lepidoptera and Hymenoptera: Symphyta, but including vermiform Diptera larvae; this category also included some pupal-stage insects), Lepidoptera (adults), Orthoptera, Gastropoda (snails), Araneae, and Opiliones (harvestmen). Because very small invertebrates constitute only a small proportion of total biomass (A. M. Strong unpubl. data) and because nestling diets of Savannah Sparrows do not commonly include invertebrates <3 mm in length (<2% of prey items in diet samples; N. J. Zalik unpubl. data), we did not include invertebrates <3 mm in length. Invertebrates whose lengths were between 3 and 10 mm were placed in 1-mm-interval size classes (3–4 mm, 4–5 mm, 5–6 mm, etc.). Those invertebrates >10 mm in length were measured to the nearest millimeter. We then used length–mass regression equations to convert length to an estimate of dry biomass, using the midpoint of each size class for those invertebrates <10 mm in length (see below). To reduce the amount of time in the field devoted to sweep-net sampling and to better synchronize sampling among fields, in 2005 we collected the contents of the sweep nets and froze them for later processing in the laboratory. In the lab, samples were thawed, vegetation was discarded, and invertebrates were placed in 70% isopropyl alcohol. We classified these preserved invertebrates into one of the 12 prey categories and dried them in a forced-air drying oven at 100°C for 24 h. Once removed from the drying oven, invertebrates were allowed to cool to room temperature and were weighed on an electronic balance (precision ± 0.01 mg). Although sweep-net samples provided an index of standing invertebrate biomass, the amount of area sampled with a sweep net is difficult to define and, therefore, abundance and density cannot be easily estimated (Cooper and Whitmore 1990). To estimate the density of invertebrates, and the total amount of energy available on a site from invertebrate prey, we used a vacuum sampler. These samples were conducted on WH and GTS only, because of the time-intensive nature of this sampling method. We conducted vacuum samples at 10 fixed points per field that were placed along the same transect lines as the sweep-net sampling points (the sampling points were spaced 25 m apart, rather than the 50-m spacing used in sweep-net sampling). A gas-powered leaf blower–vacuum (Homelite Vac Attack II model UT08934D) was used in its vacuum mode, with a fine-mesh net placed inside the end of the inlet tube, similar to that used by (Stewart and Wright 1995). Upon arrival at a sampling point, we placed a 0.5 × 0.5 × 0.9 m plastic frame over the vegetation. Once the frame was in place, the vacuum was started and the inlet tube was passed evenly through the vegetation inside the plastic frame for 20 s. We then stored the contents of the net for later processing in the laboratory. This procedure was repeated three times at each sampling point, and the second and third samples removed invertebrates that were not captured in previous samples (White et al. 1982). We measured vegetation height (cm) at each sample point 1 m from each of the four corners of the plastic frame. These four measurements were averaged to obtain an estimate of vegetation height at each sampling point. In the laboratory, we sorted the invertebrates within each sample to 1 of 12 prey categories and measured the length of each invertebrate to the nearest 0.01 mm using digital calipers. Mass (mg) was estimated using length–mass regression equations (see below). Incomplete detection of animals is a problem that must be accounted for when estimating animal abundance and density. To address this problem, the three successive samples at each vacuum-sampling point were used to estimate detection probability of invertebrates using Huggins closed-capture models in MARK (Huggins 1989, 1991; White and Burnham 1999). This analysis allowed estimation of biomass that was missed by the vacuuming. Ten Huggins closed-capture models were run in MARK to assess capture probability (p) for each of the three sampling occasions. The two most basic models, which did not incorporate individual covariates, demonstrate the basic structure of the remaining eight models, all of which included individual covariates. Model 1 assumed constant capture probability across all three sampling occasions and all sample points, regardless of vegetation height or mass of individual invertebrates (Table 1). Model 2 also disregarded effects of vegetation height or invertebrate mass but allowed capture probability for the first sampling occasion to differ from those of the second and third sampling occasions, which were constrained to equal each other. The second and third sampling occasions cannot be estimated separately, because estimating capture probability for the third occasion would require a fourth capture occasion. The remaining eight models included at least one of the following individual covariates: vegetation height (cm) at the sample point and mass (mg) of the captured invertebrates. Table 1. Descriptions of Huggins closed-capture models used to estimate detection probability for invertebrates by vacuum sampling. Models were run in MARK. Open in new tab Table 1. Descriptions of Huggins closed-capture models used to estimate detection probability for invertebrates by vacuum sampling. Models were run in MARK. Open in new tab To account for model selection uncertainty, we averaged the results from all models weighted by their Akaike weights (Burnham and Anderson 2002). To obtain biomass density estimates (mg dry mass per square meter) for each sample point, we divided the mass of each captured invertebrate by its model-averaged overall capture probability (p*) and divided this result by the area sampled (0.25 m2). The densities at each sample point were then averaged to obtain an estimate for each site and date. In converting biomass density estimates into energetic estimates, we assumed that 1 g of dry biomass yielded 23.2 kJ of energy (Golley 1961, Williams 1987). Invertebrate length–mass regression equations.—Length–mass regression equations were constructed to convert invertebrate length into estimates of dry biomass for the 12 prey categories described above. For invertebrates that did not belong to any of these groups, we used an equation obtained by averaging the equations of the 12 categories. To construct the equations, we measured the lengths (to the nearest 0.01 mm) of selected invertebrates obtained from sweep-net samples. We then dried the invertebrates in a forced-air drying oven at 100°C for 24 h. After allowing time for the samples to cool to room temperature, they were weighed on an electronic balance (precision ± 0.01 mg). Mass was then regressed against length using a power function of the form: mass = a × lengthb. Feeding behavior.—We videotaped nests with nestlings between four and eight days of age to examine adult feeding rates. In 2004, we taped nests using a Sony TRV-460 Digital8 camcorder mounted on a tripod. Because nests were always well concealed by vegetation, it was necessary to tie back vegetation with string to have a clear view of the nest. We placed an artificial perch between the nest and the camcorder on which adults could land. We taped nests for 1.5 h in the morning, beginning between 0530 and 0930 hours. To allow time for the birds to become accustomed to the camera, the first 15 min of each tape were not used in the analysis. Furthermore, if there was any indication that either of the adults was disturbed by the camera beyond the initial 15 min, we discarded the entire tape from analysis. In 2005, we placed a wide-angle lens contained in a small (approximately 10 × 4 × 4 cm) metal housing slightly above ground level, approximately 20–30 cm from the nest. This lens was then remotely attached to the Sony camcorder used for the 2004 videos. This system provided much clearer views of the feeding trips and did not cause as much disturbance to the surrounding vegetation as in 2004. We quantified the number of feeding trips made during each recording session and converted this to an hourly rate. In addition, we recorded the prey category (see above) and size of each prey item delivered. Prey size was estimated by comparing the size of the prey with the exposed portion of the adult's bill as a reference. We placed prey into size classes of 0–1, 1–2, 2–3, and >3 bill lengths. Using an average bill length of 11.2 mm (Yunick 1990, Pyle 1997) and the median value from each size class, we estimated lengths of 5.6 mm, 16.8 mm, 28 mm, and 39.2 mm, respectively, for the four size classes. The dry biomass of each prey item was then estimated using length–mass regression equations. For prey items that were not identifiable, we used the equation obtained by averaging across all prey categories (see above). Nestling mass.—We measured nestling mass between four and seven days posthatch (mode = 6 days). A linear regression equation was developed to allow comparison between nestlings weighed at different ages. We obtained this equation by regressing nestling mass against age for all nestlings weighed during the study. The resulting equation was mass (g) = 3.60 + 1.41 × age (days), (r2 = 0.37, P < 0.0001, n = 85). To correct to day-6 mass, we subtracted the mass predicted by the regression equation for the day of measurement from the observed mass. We then added this difference to the predicted day-6 mass to obtain the day-6 estimate. This method assumed that all nestlings grow at the same rate and that any deviation from the mean mass remains constant over the nestling period. It is likely that nestlings that were, for instance, below average mass on day 4 would be even further below average on day 6, because of their continued slow growth rate. However, because the mass measurements were always taken within one to two days of day 6, we assumed that the results obtained from this method would not differ substantially from the true day-6 values. Statistical analyses.—Clutch-size comparisons were made using Wilcoxon rank-sum tests. To assess whether differences in clutch size existed between treatments before cutting, we compared first nests between treatment and control fields. Because hay cropping typically occurred at least two weeks after Savannah Sparrows began nesting, a breeding pair's first nest of the season was nearly always constructed on an uncut field, regardless of whether that field was a treatment or a control. Because we predicted that food resources would decrease after haying, we expected that any differences in clutch size would be manifested in second nests (second nest after failure of the first). We therefore examined differences between second nests constructed on cut fields (postcut nests) and uncut fields (precut nests). We also compared clutch size of second broods (second nest after a successful first nest) between precut and postcut nests. Because sample sizes were small when examining each year separately, clutch-size data were pooled over the two years of the study (t-test, t = 1.65, P = 0.17, df = 292). Nests that were depredated before clutch completion were not included in analyses of clutch size. Ideally, parental feeding behavior and nestling mass would be analyzed in a manner similar to that for clutch size, with postcut second nests being compared with precut second nests. However, two primary factors made obtaining data for such an analysis difficult. First, predation rates were high during the two years of the study, especially for renests after cutting on treatment fields. Therefore, although clutch size data were often obtainable, many nests were depredated before videotaping and the weighing of nestlings. Second, some female Savannah Sparrows apparently immigrated into the fields after cutting, given that nests for these females were not discovered before haying. These females were often unbanded, which further suggests that they were not present before haying. For these birds, it was uncertain whether they had attempted a first nest on another field. Therefore, even if a nest survived until videotaping and weighing, it was not always clear whether the nest was a first or a second nesting attempt. Consequently, we classified nests solely on the basis of the status of the field (precut–postcut) on which they were constructed. In other words, nests were classified as pre- or postcut nests, regardless of whether they were constructed on a treatment or control field. Although this classification method ignored differences among individuals regarding previous reproductive efforts, all nests were consistently classified on the basis of field type (cut or uncut) in which parents obtained food. We used analysis of covariance (ANCOVA) to analyze feeding rate (number of feeding trips per hour per nest and number of feeding trips per hour per nestling), mass-delivery rate (food mass delivered per hour per nest and food mass delivered per hour per nestling), and average load size (mass of food brought to the nest per feeding trip). In these analyses, we included year and cutting status (precut or postcut) as main effects and nestling age as a continuous covariate. We also included an interaction between year and cutting status to examine whether the effect of cutting status varied between years. For ANCOVA models, we used normal probability plots of model residuals to assess normality of the data. Model residuals were plotted against predicted values to assess homogeneity of variances (Montgomery 2001). The average load-size and mass-delivery variables were transformed using a natural log transformation to satisfy the ANCOVA's normality assumption. For nestling mass, two variables were examined: average nestling mass and lightest nestling mass. Because nestlings within a nest were not independent, one value from each nest was obtained by averaging the mass of all nestlings within a nest (average nestling mass). Sample sizes presented are, therefore, the number of nests in which nestlings were weighed, rather than the total number of nestlings. Lightest nestling mass refers to the lightest nestling within each nest. Both average and lightest nestling mass were analyzed using ANCOVA, with cutting status (precut or postcut) as the main effect and number of nestlings at the time of weighing as a covariate. Data presented from all ANCOVA analyses are least-square means. All statistical analyses were performed using the SAS system for Windows (SAS Institute, Cary, North Carolina). Results Invertebrate length–mass regression equations.—Length–mass regression equations were developed for 12 prey categories (Table 2). Sample sizes ranged from 10 for Opiliones to 55 for holometabolous larvae; r2 values ranged from 0.61 for Coleoptera to 0.93 for Orthoptera (all P < 0.0001). Table 2. Length–mass regression equations of 12 prey categories collected in sweep-net samples on study sites in Hinesburg and Shelburne, Vermont, in 2004 and 2005. Equations were used to predict dry biomass (mg) of prey from body length (mm). Open in new tab Table 2. Length–mass regression equations of 12 prey categories collected in sweep-net samples on study sites in Hinesburg and Shelburne, Vermont, in 2004 and 2005. Equations were used to predict dry biomass (mg) of prey from body length (mm). Open in new tab Food availability: Sweep-net samples.—In both 2004 and 2005, invertebrate biomass collected per sweep-net sample generally increased throughout the nesting season on control fields, though late-season declines beginning in mid-July were noted on EM and TR in 2005. Across sites, peak invertebrate biomass was greatest at EM in both years (238.65 ± 52.82 mg dry biomass per sample point on 29 July 2004; 229.21 ± 36.11 mg per sample point on 3 July 2005). On treatment fields, postcut biomass estimates were 36–82% lower than precut estimates (n = 7 cutting events). As vegetation grew after cutting, invertebrate biomass estimates continually increased on cut fields (Fig. 1). Fig. 1. Open in new tabDownload slide Seasonal changes in average dry biomass (mg) collected per sweep-net sample point on study sites in Shelburne and Hinesburg, Vermont, in 2004 and 2005. Cutting events are marked with a vertical arrow. Error bars show ±1 SE. (A) EM and WH, 2004. WH was cut on 5 June. (B) EM and WH, 2005. WH was cut on 27 May and 12 July. (C) GTN and GTS, 2004. GTN and GTS were cut on 7 June and 2 July, respectively. GTN and GTS were sampled as a single site on 29 May; hence, the single value for this date. The dashed lines assume that the first sample is representative of both GTN and GTS. (D) GTN and GTS, 2005. Cutting occurred from 26 June to 1 July at GTN, and from 2 July through 10 July at GTS. (E) TR, 2005. Fig. 1. Open in new tabDownload slide Seasonal changes in average dry biomass (mg) collected per sweep-net sample point on study sites in Shelburne and Hinesburg, Vermont, in 2004 and 2005. Cutting events are marked with a vertical arrow. Error bars show ±1 SE. (A) EM and WH, 2004. WH was cut on 5 June. (B) EM and WH, 2005. WH was cut on 27 May and 12 July. (C) GTN and GTS, 2004. GTN and GTS were cut on 7 June and 2 July, respectively. GTN and GTS were sampled as a single site on 29 May; hence, the single value for this date. The dashed lines assume that the first sample is representative of both GTN and GTS. (D) GTN and GTS, 2005. Cutting occurred from 26 June to 1 July at GTN, and from 2 July through 10 July at GTS. (E) TR, 2005. Food availability: Vacuum samples.—Model 8 (p(1,2=3, mass + veg ht)) was the top-ranked model used to estimate capture probability, though the top three ranked models all received substantial support (Table 3). Only those models that estimated capture probability separately for the first capture occasion received support. Vegetation height appeared to be an important covariate in influencing detection probability, given that the top three models all contained this variable. Examination of the betas from the models indicated that increasing vegetation height had a positive effect on capture probability, whereas increasing mass had a slight negative impact. The model-averaged capture probabilities for the three capture occasions were 0.46, 0.33, and 0.33 for capture occasions one, two, and three, respectively. These values yielded an overall capture probability (p*) of 0.76, meaning that, on average, 76% of the invertebrates were captured at each vacuum-sample point. Applying this capture-probability estimate yielded biomass-density estimates for each of the sampling occasions. These results revealed that biomass density declined by 28–56% after cutting, compared with precutting estimates. However, the late-season sample on WH indicated that the invertebrate community could recover relatively quickly from disturbance, because biomass on this occasion was greater than the precutting estimate (Table 4). Table 3. Huggins closed-capture model results, listed in order of lowest to highest AICc score. All models were model-averaged to estimate detection probability of invertebrates by vacuum sampling. Open in new tab Table 3. Huggins closed-capture model results, listed in order of lowest to highest AICc score. All models were model-averaged to estimate detection probability of invertebrates by vacuum sampling. Open in new tab Table 4. Mean biomass (± SE) and energetic density estimates from vacuum sampling conducted on two treatment fields in Hinesburg and Shelburne, Vermont, pre- and post-cutting, 2005. Estimates are corrected for incomplete detection of invertebrates. Open in new tab Table 4. Mean biomass (± SE) and energetic density estimates from vacuum sampling conducted on two treatment fields in Hinesburg and Shelburne, Vermont, pre- and post-cutting, 2005. Estimates are corrected for incomplete detection of invertebrates. Open in new tab Clutch size and number of fledglings.—Average clutch size of first nests on treatment fields did not differ significantly from that of first nests on control fields (Wilcoxon rank-sum exact test, P = 0.30; Table 5). Neither average clutch size of second nests nor average size of second broods differed between cut and uncut fields (second nests: Wilcoxon rank-sum exact test, P = 0.47; second broods: Wilcoxon rank-sum exact test, P = 0.63). However, we found that the number of young fledged per successful nest was lower late in the season (nests with clutch completion dates after 24 June) on cut fields (2.2 fledglings nest−1; n = 12) than on uncut fields (3.2 fledglings nest−1; n = 6; Wilcoxon rank-sum exact test, P = 0.06), despite similar clutch sizes. Early-season nests, however, were not statistically different between cut (3.1 fledglings nest−1; n = 7) and uncut fields (3.3 fledglings nest−1; n = 28; Wilcoxon rank-sum exact test, P = 0.57) in the number of young fledged per successful nest. Table 5. Mean clutch size of Savannah Sparrows on study sites in Hinesburg and Shelburne, Vermont, in 2004 and 2005. "Treatment fields" refers to fields that were cut at least once during the breeding season. "Control fields" were not cut until after the breeding season. "Precut nests" included any second nests constructed on uncut fields, regardless of whether the field was classified as a treatment or control field. "Postcut nests" were second nests constructed on cut treatment fields. Open in new tab Table 5. Mean clutch size of Savannah Sparrows on study sites in Hinesburg and Shelburne, Vermont, in 2004 and 2005. "Treatment fields" refers to fields that were cut at least once during the breeding season. "Control fields" were not cut until after the breeding season. "Precut nests" included any second nests constructed on uncut fields, regardless of whether the field was classified as a treatment or control field. "Postcut nests" were second nests constructed on cut treatment fields. Open in new tab Feeding behavior.—All food-delivery variables (feeding rate, feeding rate per nestling, average load size, mass-delivery rate, and mass-delivery rate per nestling) significantly increased with nestling age (all P ≤ 0.05). When controlling for nestling age, parents delivered 73% more food mass per hour to nests on uncut fields (least-squares [LS] mean = 229.4 mg, 95% CI: 170.9–307.7 mg, n = 23) than on cut fields (LS mean = 132.8 mg, 95% CI: 92.8–190.1 mg, F = 5.75, df = 1 and 32, P = 0.02, n = 15). Because of larger average brood size of videotaped nests on uncut fields (3.5 nestlings) than on cut fields (3.2), mass-delivery rates on uncut fields were only 50% greater than on cut fields on a per-nestling basis (uncut fields: LS mean = 66.2 mg, 95% CI: 49.3–88.8 mg; cut fields: LS mean = 44.2, 95% CI: 30.9–63.3 mg; F = 3.14, df = 1 and 32, P = 0.085). Average load size appeared to be the primary factor driving the difference in mass-delivery rate, rather than the number of trips made per hour. Although this was not quite statistically significant, parents on uncut fields brought 48% more food mass per trip (LS mean = 29.0 mg, 95% CI: 21.9–38.4 mg) than parents on cut fields (LS mean = 19.6 mg, 95% CI: 13.9–27.6 mg, F = 3.26, df = 1 and 32, P = 0.08), again controlling for nestling age. The number of feeding trips per hour, accounting for nestling age, was similar between cut (LS mean = 8.98 trips h−1, n = 16) and uncut fields (LS mean = 8.49 trips h−1, F = 0.12, df = 1 and 33, P = 0.73, n = 23). Nestling mass.—When controlling for brood size, average nestling mass did not differ between nests on cut and uncut fields either in 2004 (F = 0.13, df = 1 and 33, P = 0.72) or 2005 (F = 0.33, df = 1 and 24, P = 0.57) or when data from both years were combined (F = 0.02, df = 1 and 60, P = 0.88; Table 6). Brood size had a significant negative effect on the mass of the lightest nestlings when data from both years were pooled (F = 5.73, df = 1 and 60, P = 0.02). However, lightest-nestling mass, though greater on uncut fields, did not differ statistically between cut and uncut fields, again accounting for brood size (2004: F = 0.25, df = 1 and 33, P = 0.62; 2005: F = 0.38, df = 1 and 24, P = 0.54; both years combined: F = 1.30, df = 1 and 60, P = 0.26; Table 6). Table 6. Least-square means (± SE) of all nestlings and lightest nestlings within a nest, controlling for brood size, in Hinesburg and Shelburne, Vermont, in 2004 and 2005. "Precut" refers to nests that were constructed before cutting, regardless of field type (treatment or control). "Postcut" refers to nests constructed on treatment fields after cutting. Open in new tab Table 6. Least-square means (± SE) of all nestlings and lightest nestlings within a nest, controlling for brood size, in Hinesburg and Shelburne, Vermont, in 2004 and 2005. "Precut" refers to nests that were constructed before cutting, regardless of field type (treatment or control). "Postcut" refers to nests constructed on treatment fields after cutting. Open in new tab Discussion Using two different sampling methods, we found that the biomass of invertebrate prey on hay fields declined after cutting. Sweep-net samples showed a 36–82% decline in biomass after cutting, whereas vacuum samples showed a 28–56% decline. This effect is likely attributable to both the removal of vegetation on which invertebrates live and feed and the physical destruction and collection of invertebrates during the harvesting process. Confidence in the sweep-net samples as reliable estimators of food availability for Savannah Sparrows is based on comparisons with vacuum samples as well as foraging observations of adults feeding young. Vacuum sampling, which is a more thorough method of examining prey biomass (mean individual capture probability of invertebrates in the present study was 0.76), showed patterns similar to those obtained from the sweep-net samples (a decrease in food abundance after cutting). Although the two sampling methods were not designed to be quantitatively compared, qualitative patterns are similar. Nevertheless, both sampling methods are representative of food availability only if they sample the foraging areas used by birds. Our study sites were bordered by woody vegetation, including shrubs and mature deciduous trees. Daily observations indicated that males often used these field edges as singing perches and females often retreated to these areas after being flushed from the field. A concern after the first year of the study was that if birds used these edges for foraging, the sweep-net and vacuum samples would not fully represent the available food resources. Therefore, we performed 0.5-h observations of provisioning adults (n = 12 nests) to determine whether they were using field edges or feeding exclusively within the study fields. In only one observation period did adults obtain food from outside the study field, in this case foraging in an adjacent field. Although adults sometimes left field borders, we never observed them bringing to nestlings food that was obtained in the woody vegetation bordering the study fields. Another concern with sweep-net sampling is that its effectiveness decreases with increasing vegetation height, because a smaller proportion (the upper portion) of the vegetation is effectively sampled (Cooper and Whitmore 1990). Although this is a valid concern, it only suggests that our estimates of biomass reduction in cut fields were conservative (uncut-field estimates may have been low, whereas cut-field estimates should have been more accurate). The effect of reduced food availability on Savannah Sparrows was examined by comparing clutch size, food provisioning of nestlings, and nestling mass between cut and uncut fields. We found no difference between cut and uncut fields in clutch size or nestling mass. However, adults on cut fields delivered less food biomass per hour and per feeding trip to their nests. The difference in mass delivery was nearly significant on a per-nestling basis. Given lower food availability on cut fields, these results are expected (sensuBlancher and Robertson 1987, Naef-Daenzer and Keller 1999). However, birds on cut fields were able to raise nestlings whose mass was equal to that of nestlings on uncut fields. The lack of difference in nestling mass between treatments suggested that adult Savannah Sparrows compensated for reduced food availability, though the method of compensation was not clear. In a grassland system studied by (Adams et al. 1994), experimentally reduced grasshopper populations did not affect Vesper Sparrow (Pooecetes gramineus) food-provisioning rates or nestling growth, and adults compensated by expanding their foraging area. One explanation for the discrepancy between the feeding observations and the nestling-mass results in our system is that the quality of food brought by adults may have varied between cut and uncut fields. If birds delivered higher-quality foods on cut fields, nestlings would have been able to attain mass equal to that of nestlings on uncut fields, even though they were fed less food biomass. It seems unlikely that the availability of high-quality food would increase after cutting, though an examination of late-season (>23 June) diet samples revealed that the diet of nestlings on cut fields consisted of greater proportions of larvae (26.1% on cut fields vs. 3.3% on uncut fields), which may be high-quality foods, and lesser proportions of orthopterans (20.8% on cut fields vs. 30.6% on uncut fields), which may be of lesser quality because of their decreased digestibility (Zalik 2007). Decreased chemical or structural defense of grasses following cutting (sensuMattson 1980) may have increased food quality for invertebrates, which may have improved foraging conditions for Savannah Sparrows. Although these differences in diet and food quality between treatments may have partially compensated for the reduced amount of biomass delivered per nestling on cut fields, it is unlikely that these differences could fully account for the 50% greater mass-delivery rate on uncut fields. A second explanation of compensation is that the availability of prey may have been altered by the change in vegetation structure (sensuEvans et al. 2006, Jiménez-Valverde and Lobo 2007) resulting from cutting. Because vegetation was less dense after cutting and there was less vegetation to search, the proportion of prey that was available to birds may have been greater. However, ultimately, the proportion of prey that is available is of less importance than the total amount of available prey, which depends on both prey abundance and the proportion available. Furthermore, greater overall prey availability should have been reflected in greater mass of food delivered to nests, so this explanation seems unlikely. A third explanation is that the feeding observations did not represent feeding behavior over the entire day. All feeding observations were conducted during the early and midmorning hours, at which time birds may have been feeding at a maximum rate. To avoid feeding at a maximum rate during the hot afternoon hours, birds may have focused feeding efforts toward the cooler morning hours. If, however, the amount of food gathered in the morning was insufficient for the energetic needs of nestlings, high feeding rates may have been required throughout the day to compensate for this energetic shortfall. Several studies have found that feeding rates vary according to time of day. (Knapton 1984) found that Nashville Warblers' (Vermivora ruficapilla) nestling-feeding rates were highest during the early-morning (0800–1100) and evening (1700–2000) hours. Likewise, (Best 1977) found that feeding rates of Field Sparrows (S. pusilla) peaked after dawn and before dusk, with the lowest rates recorded at midday. Other studies, of Ipswich Savannah Sparrows (P. s. princeps), Black-throated Blue Warblers, and Yellow Warblers (D. petechia), have found no relationship between feeding rates and time of day (Welsh 1975, Biermann and Sealy 1982, Goodbred and Holmes 1996). (Goodbred and Holmes 1996) suggested that time of day may influence feeding rates in open habitats more than in forested habitats, through depressed insect activity or increased physiological stress on foraging adults. Our study sites typically become hot during the afternoon, especially during late June and July (average high temperature ∼ 26.7°C; National Climatic Data Center 2007), a time in which birds whose first nests were destroyed by haying are raising young. Thus, it is reasonable that given abundant food, adult birds would normally decrease their afternoon feeding efforts. However, given that Savannah Sparrows delivered more food per hour during the morning on uncut fields than on cut fields, differential feeding rates between treatments throughout the day should be investigated as a possible mechanism of compensation for reduced food availability. Regardless of the mechanism, the decreased fledging rate on cut fields late in the season suggests that any compensatory mechanisms may have been insufficient to overcome decreased resource availability after cutting. As noted above, other changes occur to hay fields following cutting in addition to the significant decreases in invertebrate biomass. Decreased grass height likely led to changes in the thermal microclimate of nest sites. However, the influence of this change on foraging patterns of adult Savannah Sparrows likely depends on ambient environmental conditions and could increase or decrease the amount of incubation or brooding time required of females (Kluijver 1950). Exposure of nests to predators is likely greater after hay cutting; however, we documented equal feeding rates in cut and uncut fields, an unexpected result if parents were attempting to avoid detection by nest predators (Martin et al. 2000). Additionally, nesting densities of Red-winged Blackbirds (Agelaius phoeniceus; uncommon in all fields) and Bobolinks (Dolichonyx oryzivorus; common in all fields) were substantially reduced after hay cutting, potentially decreasing interspecific competition (Perlut et al. 2006). However, for the first week after cutting, the fields supported substantial numbers of foraging American Crows (Corvus brachyrhynchos) and Ring-billed Gulls (Larus delawarensis), which likely decreases the biomass of invertebrates (Perlut et al. 2006). The ability of fields to provide for the daily needs of adult and nestling Savannah Sparrows can be viewed from an energetic standpoint. On the basis of data obtained from an experiment with doubly labeled water, (Williams 1987) estimated that male Savannah Sparrows tending a brood of four nestlings expended 93 kJ of energy per day, whereas females expended 80.6 kJ. He also estimated that a brood of four nestling Savannah Sparrows requires 1,497.2 kJ over an eight-day nestling period. Thus, if energy expenditure of the nestlings is averaged over the nestling period, a brood of four young raised by two parents requires 360.8 kJ day−1. Given the lowest energetic estimate (GTS postcutting), this energy demand could be met by a foraging area of only 116 m2 (the smallest area needed would be 40 m2 for WH late in the season). Clearly, given territory sizes that are at least 1,000 m2 in size (Wheelwright and Rising 1993), the amount of daily energy required by birds is small compared with the amount present on these fields, even after cutting. Thus, even after substantial postcutting food reduction, Savannah Sparrows should not be food-limited on the basis of food abundance. This result is consistent with the hypothesis of superabundant food resources in grasslands (Wiens and Rotenberry 1979, Rotenberry 1980). However, because of cryptic coloration and escape tactics of invertebrates, certainly not all of this food is available to birds (Holmes 1990). Without knowing the foraging efficiency of Savannah Sparrows, we did not know whether such values indicate food shortage or superabundance. Foraging costs (time spent searching for, capturing, and handling prey) can be great and can greatly reduce the amount of energy that adults bring to nestlings, as indicated by the feeding observations. Thus, food limitation may result from constraints on foraging time as much as from an actual shortage of available prey (Weathers and Sullivan 1989). Additional costs of nesting in cut hay fields may be associated with the change in food resources after cutting. For instance, (Perlut et al. 2006) found that Savannah Sparrows took significantly longer to renest (by 5–6 days) after haying-related nest failure than after nest failure in uncut fields. This delay could be attributable to the reduced food availability immediately after a cutting event. Thus, decreased food resources may have effects on reproductive ecology that are not directly quantified through metrics associated with nest contents. Adult Savannah Sparrows must increase their foraging effort to provide sufficient food for adequate nestling growth when food is relatively scarce. Increased reproductive effort has been linked to lower probability of surviving the nonbreeding season (Linden and Møller 1989). Proposed reasons for this tradeoff include reduced immunocompetence (Gustafsson et al. 1994) and reduced feather quality in the prebasic molt (Nilsson and Svensson 1996, Dawson et al. 2000). Also, if birds delay migration, their chances of facing unfavorable weather conditions associated with the onset of winter may increase. Interestingly, survival rates of adult Savannah Sparrows in this system vary with the intensity of management on nesting fields, with lower apparent survival rates found on more intensively managed fields (i.e., earlier and more frequent cuts per year; Perlut 2008). Tradeoffs between current reproductive effort and future survival have rarely been empirically demonstrated in migratory passerines, undoubtedly because of the difficulty of monitoring individuals of species that migrate great distances between seasons. Such costs in migratory species, however, may be greater than those in resident species, given the added energetic cost and dangers associated with migration. Even without the added difficulties of reduced food resources on postcut fields, Savannah Sparrows that nest in these fields raise nestlings that may be at a disadvantage because they have less time to mature before the fall migration. Studies have shown that birds hatched early in the breeding season are more likely to survive to breeding age than those hatched later in the year (Naef-Daenzer et al. 2001). Reduced food availability may exacerbate this effect, given that recently fledged Savannah Sparrows have poorly developed foraging skills and are highly dependent on adults (Wheelwright and Templeton 2003). Thus, to provide food for their young, increased adult foraging efforts likely extend for several weeks beyond the nestling period, placing an additional energetic burden on the adults. Further research should examine carry-over costs of reduced food resources to adult and nestling birds beyond the breeding season. Acknowledgments This research was supported by the National Research Initiative of the U.S. Department of Agriculture, Cooperative State Research, Education and Extension Service, grant no. 03-35101-13817, and by the Natural Resource Conservation Service's Wildlife Habitat Management Institute. We are grateful to T. Donovan, R. Mickey, two anonymous reviewers, and members of the Strong lab group for reviewing earlier drafts of this manuscript. We thank Shelburne Farms and the Ross and Thibault families for allowing us to conduct this study on their properties. Valuable field assistance was provided by N. Perlut, C. Lang, T. Lawrence, D. Leblanc, C. Lucas, L. MacDade, H. Murray, and K. Willard. Lab assistance was provided by A. Mallon, K. Proudman, and A. Vardo. Literature Cited Adams , J. S. , R. L. Knight , L. C. Mcewan , and T. L. George . 1994 . Survival and growth of nestling Vesper Sparrows exposed to experimental food reductions. Condor 96 : 739 – 748 . Google Scholar Crossref Search ADS WorldCat Askins , R. A. 1999 . History of grassland birds in eastern North America. 60 – 71 . Ecology and Conservation of Grassland Birds of the Western Hemisphere ( Vickery P. D. , and J. R. Herkert , editors. Eds.) . Studies in Avian Biology, no. 19 . Best , L. B. 1977 . Nestling biology of the Field Sparrow. Auk 94 : 308 – 319 . WorldCat Biermann , G. C. , and S. G. Sealy . 1982 . Parental feeding of nestling Yellow Warblers in relation to brood size and prey availability. Auk 99 : 332 – 341 . WorldCat Blancher , P. J. , and R. J. Robertson . 1987 . Effect of food supply on the breeding biology of Western Kingbirds. Ecology 68 : 723 – 732 . Google Scholar Crossref Search ADS WorldCat Boutin , S. 1990 . Food supplementation experiments with terrestrial vertebrates: Patterns, problems, and the future. Canadian Journal of Zoology 68 : 203 – 220 . Google Scholar Crossref Search ADS WorldCat Britschgi , A. , R. Spaar , and R. Arlettaz . 2006 . Impact of grassland farming intensification on the breeding ecology of an indicator insectivorous passerine, the Whinchat Saxicola rubetra: Lessons for overall Alpine meadow management. Biological Conservation 130 : 193 – 205 . Google Scholar Crossref Search ADS WorldCat Burnham , K. P. , and D. R. Anderson . 2002 . Model Selection and Multimodel Inference: A Practical Information-theoretic Approach . 2nd ed. Springer-Verlag . New York . Cooper , R. J. , and R. C. Whitmore . 1990 . Arthropod sampling methods in ornithology. 29 – 37 . Avian Foraging: Theory, Methodology, and Applications ( Morrison M. L. , C. J. Ralph , J. Verner , and Jehl J. R. Jr. , editors. Eds.) . Studies in Avian Biology, no. 13 . Dawson , A. , S. A. Hinsley , P. N. Ferns , R. H C. Bonser , and L. Eccleston . 2000 . Rate of moult affects feather quality: A mechanism linking current reproductive effort to future survival. Proceedings of the Royal Society of London, Series B 267 : 2093 – 2098 . Google Scholar Crossref Search ADS WorldCat Evans , D. M. , S. M. Redpath , S. A. Evans , D. A. Elston , C. J. Gardner , P. Dennis , and R. J. Pakeman . 2006 . Low intensity, mixed livestock grazing improves the breeding abundance of a common insectivorous passerine. Biology Letters 2 : 636 – 638 . Google Scholar Crossref Search ADS PubMed WorldCat Golley , F. B. 1961 . Energy values of ecological materials. Ecology 42 : 581 – 584 . Google Scholar Crossref Search ADS WorldCat Goodbred , C. O'N. , and R. T. Holmes . 1996 . Factors affecting food provisioning of nestling Black-throated Blue Warblers. Wilson Bulletin 108 : 467 – 479 . WorldCat Gustafsson , L. , D. Nordling , M. S. Andersson , B. C. Sheldon , and A. Qvarnström . 1994 . Infectious diseases, reproductive effort and the cost of reproduction in birds. Philosophical Transactions of the Royal Society of London, Series B 346 : 323 – 331 . Google Scholar Crossref Search ADS WorldCat Hart , J. D. , T. P. Milsom , G. Fisher , V. Wilkins , S. J. Moreby , A. W A. Murray , and P. A. Robertson . 2006 . The relationship between Yellowhammer breeding performance, arthropod abundance and insecticide applications on arable farmland. Journal of Applied Ecology 43 : 81 – 91 . Google Scholar Crossref Search ADS WorldCat Holmes , R. T. 1990 . Ecological and evolutionary impacts of bird predation on forest insects: An overview. 6 – 13 . Avian Foraging: Theory, Methodology, and Applications. ( Morrison M. L. , C. J. Ralph , J. Verner , and Jehl J. R. Jr. , editors. Eds.) . Studies in Avian Biology, no. 13 . Howe , F. P. , R. L. Knight , L. C. Mcewen , and T. L. George . 1996 . Direct and indirect effects of insecticide applications on growth and survival of nestling passerines. Ecological Applications 6 : 1314 – 1324 . Google Scholar Crossref Search ADS WorldCat Huggins , R. M. 1989 . On the statistical analysis of capture experiments. Biometrika 76 : 133 – 140 . Google Scholar Crossref Search ADS WorldCat Huggins , R. M. 1991 . Some practical aspects of a conditional likelihood approach to capture experiments. Biometrics 47 : 725 – 732 . Google Scholar Crossref Search ADS WorldCat Jiménez-Valverde A. , and J. M. Lobo . 2007 . Determinants of local spider (Araneidae and Thomisidae) species richness on a regional scale: Climate and altitude vs. habitat structure. Ecological Entomology 32 : 113 – 122 . Google Scholar Crossref Search ADS WorldCat Kluijver , H. N. 1950 . Daily routines of the Great Tit, Parus m. major L. Ardea 38 : 99 – 135 . WorldCat Knapton , R. W. 1984 . Parental feeding of nestling Nashville Warblers: The effects of food type, brood-size, nestling age, and time of day. Wilson Bulletin 96 : 594 – 602 . WorldCat Lindén M. , and A. P. Møller . 1989 . Cost of reproduction and covariation of life history traits in birds. Trends in Ecology and Evolution 4 : 367 – 371 . Google Scholar Crossref Search ADS PubMed WorldCat Marshall , M. R. , R. J. Cooper , J. A. Dececco , J. Strazanac , and L. Butler . 2002 . Effects of experimentally reduced prey abundance on the breeding ecology of the Red-eyed Vireo. Ecological Applications 12 : 261 – 280 . Google Scholar Crossref Search ADS WorldCat Martin , T. E. 1987 . Food as a limit on breeding birds: A life-history perspective. Annual Review of Ecology and Systematics 18 : 453 – 487 . Google Scholar Crossref Search ADS WorldCat Martin , T. E. , P. R. Martin , C. R. Olson , B. J. Heidinger , and J. J. Fontaine . 2000 . Parental care and clutch sizes in North and South American birds. Science 287 : 1482 – 1485 . Google Scholar Crossref Search ADS PubMed WorldCat Mattson , W. J. Jr 1980 . Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics 11 : 119 – 161 . Google Scholar Crossref Search ADS WorldCat Montgomery , D. C. 2001 . Design and Analysis of Experiments . 5th ed. Wiley . New York . Moreby , S. J. , and S. E. Southway . 1999 . Influence of autumn applied herbicides on summer and autumn food available to birds in winter wheat fields in southern England. Agriculture, Ecosystems & Environment 72 : 285 – 297 . Google Scholar Crossref Search ADS WorldCat Naef-Daenzer B. , and L. F. Keller . 1999 . The foraging performance of Great and Blue tits (Parus major and P. caeruleus) in relation to caterpillar development, and its consequences for nestling growth and fledging weight. Journal of Animal Ecology 68 : 708 – 718 . Google Scholar Crossref Search ADS WorldCat Naef-Daenzer B. , F. Widmer , and M. Nuber . 2001 . Differential post-fledging survival of Great and Coal tits in relation to their condition and fledging date. Journal of Animal Ecology 70 : 730 – 738 . Google Scholar Crossref Search ADS WorldCat Nagy , L. R. , and R. T. Holmes . 2005 . Food limits annual fecundity of a migratory songbird: An experimental study. Ecology 86 : 675 – 681 . Google Scholar Crossref Search ADS WorldCat National Climatic Data Center. 2007 . Daily normals of temperature, precipitation, and heating and cooling degree days. Climatography of the United States No. 84, 1971–2000 U.S. Department of Commerce . Google Scholar Newton , I. 1998 . Population Limitation in Birds . Academic Press . San Diego, California . Nilsson , J-Å , and E. Svensson . 1996 . The cost of reproduction: A new link between current reproductive effort and future reproductive success. Proceedings of the Royal Society of London, Series B 263 : 711 – 714 . Google Scholar Crossref Search ADS WorldCat Perlut , N. G. , A. M. Strong , T. M. Donovan , and N. J. Buckley . 2006 . Grassland songbirds in a dynamic management landscape: Behavioral responses and management strategies. Ecological Applications 16 : 2235 – 2247 . Google Scholar Crossref Search ADS PubMed WorldCat Perlut , N. G. , A. M. Strong , T. M. Donovan , and N. J. Buckley . 2008 . Grassland songbird survival and recruitment in heterogeneous agricultural landscapes: Implications for source–sink demography. Ecology 89 . in press. Pyle , P. 1997 . Identification Guide to North American Birds, part 1: Columbidae through Ploceidae . Slate Creek Press . Bolinas, California . Rands , M. R W. 1985 . Pesticide use on cereals and the survival of Grey Partridge chicks: A field experiment. Journal of Applied Ecology 22 : 49 – 54 . Google Scholar Crossref Search ADS WorldCat Ritchie , M. E. 2000 . Nitrogen limitation and trophic vs. abiotic influences on insect herbivores in a temperate grassland. Ecology 81 : 1601 – 1612 . WorldCat Rodenhouse , N. L. , and R. T. Holmes . 1992 . Results of experimental and natural food reductions for breeding Black-throated Blue Warblers. Ecology 73 : 357 – 372 . Google Scholar Crossref Search ADS WorldCat Rotenberry , J. T. 1980 . Bioenergetics and diet in a simple community of shrubsteppe birds. Oecologia 46 : 7 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat Schmidt , M. H. , G. Lefebvre , B. Poulin , and T. Tscharntke . 2005 . Reed cutting affects arthropod communities, potentially reducing food for passerine birds. Biological Conservation 121 : 157 – 166 . Google Scholar Crossref Search ADS WorldCat Siikamäki P. 1998 . Limitation of reproductive success by food availability and breeding time in Pied Flycatchers. Ecology 79 : 1789 – 1796 . Google Scholar Crossref Search ADS WorldCat Stewart , A. J A. , and A. F. Wright . 1995 . A new inexpensive suction apparatus for sampling arthropods in grassland. Ecological Entomology 20 : 98 – 102 . Google Scholar Crossref Search ADS WorldCat Summerville , K. S. , and T. O. Crist . 2002 . Effects of timber harvest on forest Lepidoptera: Community, guild, and species responses. Ecological Applications 12 : 820 – 835 . Google Scholar Crossref Search ADS WorldCat Taylor , R. L. , B. D. Maxwell , and R. J. Boik . 2006 . Indirect effects of herbicides on bird food resources and beneficial arthropods. Agriculture, Ecosystems & Environment 116 : 157 – 164 . Google Scholar Crossref Search ADS WorldCat Troy , A. R. , A. M. Strong , S. C. Bosworth , T. M. Donovan , N. J. Buckley , and M. L. Wilson . 2005 . Attitudes of Vermont dairy farmers regarding adoption of management practices for grassland songbirds. Wildlife Society Bulletin 33 : 528 – 538 . Google Scholar Crossref Search ADS WorldCat Weathers , W. W. , and K. A. Sullivan . 1989 . Juvenile foraging proficiency, parental effort, and avian reproductive success. Ecological Monographs 59 : 223 – 246 . Google Scholar Crossref Search ADS WorldCat Welsh , D. A. 1975 . Savannah Sparrow breeding and territoriality on a Nova Scotia dune beach. Auk 92 : 235 – 251 . Google Scholar Crossref Search ADS WorldCat Wheelwright , N. T. , and J. D. Rising . 1993 . Savannah Sparrow (Passerculus sandwichensis). The Birds of North America, no. 45 . ( Poole A. , and F. Gill , editors. Eds.) . Academy of Natural Sciences . Philadelphia American Ornithologists' Union . Washington, D.C . Wheelwright , N. T. , and J. J. Templeton . 2003 . Development of foraging skills and the transition to independence in juvenile Savannah Sparrows. Condor 105 : 279 – 287 . Google Scholar Crossref Search ADS WorldCat White , G. C. , D. R. Anderson , K. P. Burnham , and D. L. Otis . 1982 . Capture–recapture and removal methods for sampling closed populations . Publication LA-8787-NERP, Los Alamos National Laboratory . Los Alamos, New Mexico . White , G. C. , and K. P. Burnham . 1999 . Program MARK: Survival estimation from populations of marked animals. Bird Study 46 : Supplement S120 – S138 . Google Scholar Crossref Search ADS WorldCat Wiens , J. A. , and J. T. Rotenberry . 1979 . Diet niche relationships among North American grassland and shrubsteppe birds. Oecologia 42 : 253 – 292 . Google Scholar Crossref Search ADS PubMed WorldCat Williams , J. B. 1987 . Field metabolism and food consumption of Savannah Sparrows during the breeding season. Auk 104 : 277 – 289 . WorldCat Yunick , R. P. 1990 . Identification criteria for some emberizid sparrows. North American Bird Bander 15 : 1 – 5 . WorldCat Zalik , N. J. 2007 . Effects of agricultural management on breeding Savannah Sparrows (Passerculus sandwichensis) mediated through food availability . M.S. thesis . University of Vermont, Burlington . Author notes Address correspondence to this author. E-mail: allan.strong@uvm.edu Associate Editor: A. D. Rodewald © The American Ornithologists' Union, 2008
TI - Effects of Hay Cropping on Invertebrate Biomass and the Breeding Ecology of Savannah Sparrows (Passerculus Sandwichensis)Efectos de la Cosecha de Heno sobre la Biomasa de Invertebrados y la Ecología Reproductiva de Passerculus sandwichensisZalik and StrongHay Cropping and Invertebrate Biomass
JF - Auk: Ornithological Advances
DO - 10.1525/auk.2008.07106
DA - 2008-07-01
UR - https://www.deepdyve.com/lp/oxford-university-press/effects-of-hay-cropping-on-invertebrate-biomass-and-the-breeding-pVm3FzADSE
SP - 700
VL - 125
IS - 3
DP - DeepDyve
ER -