Space Use and Movements of Southeastern Breeding Double-Crested Cormorants ( Nannopterum auritum ) in the United States

: Seasonal movements of Double-crested Cormorants ( Nannopterum auritum ) have been studied at breeding and wintering sites in the southeastern United States, but little information exists on the movements of these birds within and from their southern breeding sites in lacustrine systems. Since 2001, cormorants have established nesting colonies on islands in Guntersville Reservoir in Alabama, USA. Following the movements of tagged cormorants using satellite telemetry, we found that the mean home range during the 2017 breeding season (May–August) was 41.76 km 2 , with a core use area of 6.36 km 2 . The mean home range used by these birds was largest during the period coinciding with incubation: 9–30 May: (98.86 ± 80.64 km 2 ) compared with the chick-rearing 31 May–4 July: 18.30 ± 22.56 km 2 ), and the post-ﬂedge periods (5 July–15 August: 42.04 ± 30.95 km 2 ). There was no signiﬁcant difference in the metrics of movement and space use between male and female cormorants assessed in this study. Differences in space used by cormorants breeding in Alabama relative to their northern breeding grounds may be explained by landscape characteristics and availability of prey.


Introduction
A conservation success story, the abundance of Double-crested Cormorants (Nannopterum auritum, cormorants) has been increasing since the late 1970's across North America after a federal ban on DDT and sustainable management [1].From around 32,000 pairs of breeding pairs in the late 1970's to 1-2 million individuals in 2005, the number of breeding cormorants has increased, particularly in the interior portion of North America [1].Additionally, the number of cormorants wintering in the southeastern United States has also increased from 10,000 birds in 1996 to 31,000 birds in 2005 [2].Increased cormorant abundance has heightened conflict between cormorants and humans, specifically manifested through concerns over the impacts of these birds on sport fishing, economic loss from depredation on aquaculture ponds, and the degradation of natural resources on breeding sites [3].Sustainable management and thorough research are critical for the continued conservation of cormorants and the habitats they use.
Various studies have assessed cormorant movements during migration, wintering, and breeding seasons.Understanding these movements provides insight into resources used by these birds during different periods within a full annual cycle.However, with the advent of cormorant populations steadily increasing and utilizing new areas for breeding sites in the southeastern United States [1,4,5], we know less about how these birds are using these relatively novel ecosystems.Therefore, it is important to evaluate how cormorants Diversity 2023, 15, 453 2 of 11 that may not leave the southeastern United States use these ecosystems and how the use of these resources may lead to heightened conflicts among stakeholders.
Water depth, distance from shore and proximity to loafing/staging areas are all factors known to affect how cormorants use aquatic systems during winter and breeding seasons [6,7].Water depth can significantly affect cormorant movements, with birds typically capturing prey between 1-10 m (m) deep [8].Loafing or staging areas are also important for movement, with birds using small islands or channel markers to rest and dive for prey.Cormorants require suitable daytime loafing areas as well as nighttime roosts during all seasons with channel markers, floating debris, and low-lying wires and trees being used throughout their seasonal ranges [7].Islands or inlets 3 km (km) from the mainland are often used as winter roosts and breeding sites as a response for predator avoidance [9], with foraging distance typically within 20 km from these roosting or breeding sites [10,11].
During winter, cormorants that nest in North America's Great Lakes region migrate south to coastal inlets or to aquaculture-producing areas throughout the southeastern United States [12,13].Geographically, the Lower Mississippi Valley and the northern Gulf of Mexico are wintering locations for cormorants within the southeastern United States [14].Yet movements between these locations differ, with cormorants overwintering on lakes and coasts using an average home range of 81 km 2 while those in aquaculture ponds average a home range of 2760 km 2 [15].King et al. (2012) also calculated home ranges and core use areas for wintering cormorants throughout the Lower Mississippi Alluvial Valley and estimated average home ranges of 17,490 km 2 and a mean core use area of 1550 km 2 .Conversely, cormorants breeding in Lake Ontario and wintering in this same region used a mean home range and core use area of only 4609 km 2 and 566 km 2 , respectively, less than half of what was reported for the Lower Mississippi Alluvial Valley [16].
During the breeding season, variable movement patterns have been observed among cormorants throughout the Great Lakes.Cormorants nesting in Lake Ontario had home ranges and core use areas of 4646 and 820 km 2 , respectively [17].Conversely, cormorants tagged in the Mississippi Alluvial Valley that bred throughout the Great Lakes had home ranges and core use areas almost triple the size seen in Lake Ontario [16,18].This variability in both wintering and breeding seasons has been explained through differences in the management of these birds in both their breeding and wintering areas with measures to control populations implemented in the forms of egg-oiling, roost dispersal, and lethal management in some locations [15,17].Simple depletion of suitable prey species could also account for these differing mean home ranges and core use areas plus individual variability between breeding and wintering years [17,19].Regardless of intra-seasonal variability, differences between wintering and breeding cormorants are also apparent.
Our objective for this study was to measure home ranges and core use areas for individual cormorants for a breeding season (May-August 2017) and explore relationships in cormorant movements and space use.Additionally, we wanted to evaluate existing theories derived from populations of these birds breeding on North America's Great Lakes relative to recent breeding areas on a lacustrine system in the southeastern United States (Figure 1).Our prediction was that cormorants breeding on this lacustrine system in the southeastern United States will use a smaller area around nesting sites due to the smaller geographic footprint, shallower bathymetry, and distribution of resources in proximity to nesting sites within this system as compared to birds that breed on North America's Great Lakes.

Materials and Methods
This study was conducted on Guntersville Reservoir in Marshall and Jackson counties in northeast Alabama, USA (Figure 1).Guntersville Reservoir was created by damming the Tennessee River in 1935 with reservoir hydrology controlled by the Guntersville Dam and has a water surface area of 27,478 hectares and a mean water depth of 18.3 m (m) within the main channel but only 4.6 m outside of the main channel [20].Guntersville Reservoir is situated in a temperate zone and is more eutrophic and productive than northern lakes and reservoirs year-round due to high nutrient inputs from natural and agricultural sources [21,22].Temperature and climate are temperate with summers averaging 27 °C and winters around 15 °C [23].Deciduous hardwoods, such as oak (Quercus spp.) and hickory (Carya spp.), and coniferous trees such as red cedar (Juniperus virginiana) and loblolly pine (Pinus taeda) are all prevalent within the reservoir system [23].Fish communities within the reservoir include a multitude of baitfish species such as gizzard (Dorosoma cepedianum) and threadfin shad (D. petenense), skipjack herring (Alosa chrysochloris) and sportfish such as bass species (Micropterus salmoides and M. dolomieu), sauger (Sander canadensis), bluegill (Lepomis macrochirus), crappie (Pomoxis spp.), and catfish (Siluriformes) [24], with bluegill, crappie, and shad the most common prey during a diet study of cormorants feeding within Guntersville Reservoir (Dorr et.al, unpublished data).
The number of cormorants using Guntersville Reservoir (Figure 1) during the breeding season (March-August: Dorr et al., 2014) has increased since signs of nesting were first documented in 2001 [4].A systemic survey conducted in 2017 by USDA Wildlife Services estimated a total of 1620 cormorants nesting on islands within the reservoir.Little is currently known about how cormorants move within Guntersville Reservoir during their

Materials and Methods
This study was conducted on Guntersville Reservoir in Marshall and Jackson counties in northeast Alabama, USA (Figure 1).Guntersville Reservoir was created by damming the Tennessee River in 1935 with reservoir hydrology controlled by the Guntersville Dam and has a water surface area of 27,478 hectares and a mean water depth of 18.3 m (m) within the main channel but only 4.6 m outside of the main channel [20].Guntersville Reservoir is situated in a temperate zone and is more eutrophic and productive than northern lakes and reservoirs year-round due to high nutrient inputs from natural and agricultural sources [21,22].Temperature and climate are temperate with summers averaging 27 • C and winters around 15 • C [23].Deciduous hardwoods, such as oak (Quercus spp.) and hickory (Carya spp.), and coniferous trees such as red cedar (Juniperus virginiana) and loblolly pine (Pinus taeda) are all prevalent within the reservoir system [23].Fish communities within the reservoir include a multitude of baitfish species such as gizzard (Dorosoma cepedianum) and threadfin shad (D. petenense), skipjack herring (Alosa chrysochloris) and sportfish such as bass species (Micropterus salmoides and M. dolomieu), sauger (Sander canadensis), bluegill (Lepomis macrochirus), crappie (Pomoxis spp.), and catfish (Siluriformes) [24], with bluegill, crappie, and shad the most common prey during a diet study of cormorants feeding within Guntersville Reservoir (Dorr et al., unpublished data).
The number of cormorants using Guntersville Reservoir (Figure 1) during the breeding season (March-August: Dorr et al., 2014) has increased since signs of nesting were first documented in 2001 [4].A systemic survey conducted in 2017 by USDA Wildlife Services estimated a total of 1620 cormorants nesting on islands within the reservoir.Little is currently known about how cormorants move within Guntersville Reservoir during their breeding season.It is theorized that cormorants that breed in the southeastern United States may be year-round residents with seasonal immigration augmenting overwintering populations [25,26].Cormorants breeding on Guntersville is currently limited to island colonies [4,5].Islands are present throughout the reservoir (Figure 1).The southernmost part of Guntersville has the highest number of breeding colonies of cormorants which currently include Connors Islands and South Sauty with one island within mid-Guntersville, North Sauty (Figure 1).Some islands around current breeding colonies were once colonized but have subsequently been abandoned after trees and vegetation died off [27] and a few islands (Control) have not been colonized by cormorants as of the completion of this study.
The management of cormorants on Guntersville Reservoir has been conducted since evidence of breeding was detected in 2001 [4].Both lethal and non-lethal methods have been employed to disperse birds during the breeding season, with Connors and South Sauty and surrounding waterways the site of the greatest management effort (Figure 1).After the Public Resource Depredation Order was rescinded in 2016 [28], most management was ceased for the foreseeable future including the year all cormorants were tagged for this study.
To determine the cormorant home range and core use areas we targeted Connor Island and South Sauty island due to the highest density of breeding birds in these colonies.North Sauty has recorded nests but logistically trapping in the area was not possible due to water depths < 0.2 m. Adult cormorants were targeted for telemetry deployment, but all birds were banded and processed if captured.Spring-loaded leg hold traps were modified and used to capture cormorants from May-July 2017.A rubber buffer was installed inside the clamps of each leg-hold trap to decrease the chance of injury as well as replacing factory coil springs with weaker springs to reduce the closing force on the leg [29].The factory chain was replaced with an aircraft cable and elastic shock cord to minimize injury when the captured bird lunged to escape [29].Leg hold traps were placed on known loafing areas such as downed logs or debris sticking out of the water so that when the target cormorant stepped on the trap, a spring switched to "hold" the captured leg of the cormorant without injury or escape.
All captured birds were banded with an individually numbered aluminum leg band (USGS permit # 23835) as well as a uniquely coded plastic leg band.We collected a small sample of blood from each adult cormorant to genetically determine the sex of that bird.A 25-gauge syringe needle was used to pierce the right, medial metatarsal vein of each bird and a blood sample (~20-50 uL) was collected and applied to a PermaCode Card for genetic sexing (Animal Genetics, Inc., Tallahassee, FL USA).
A backpack satellite transmitter (GeoTrak Inc., Apex, NC, USA, model number GT-22GS-GPS) was placed on each cormorant using methods described by King and Tobin (2000) and Dunstan (1972).Transmitters were positioned on each bird using a backpack harness [30] with the transmitter affixed externally and secured using Teflon ribbon, with small metal bird bands crimped to secure the harness to the bird (Figure 2).So that any harsh edges from the metal band would not adversely affect the cormorant, shrink tubing covered these attachment sites using a standard soldering iron.
Each GPS tag weighed 22 gram (g) and had a proportional mass of 1.1% relative to bird weight.Each tag was equipped with a solar panel that supplied a voltage of 3.6 to 4.2 volts DC providing continuous battery life to the transmitter.Sensor data from each GPS PTT unit were checked every other day to ensure each unit was not experiencing a significant loss of power.If a unit was found to be relaying GPS coordinates consistently from one location for more than 3 days, or a transmitter had not transmitted for more than 3 days, the unit was tracked via the VHF device attached to the tag until the device was recovered.Argos data were uploaded every 48 h through the Argos system.Each GPS tag weighed 22 gram (g) and had a proportional mass of 1.1% relative to bird weight.Each tag was equipped with a solar panel that supplied a voltage of 3.6 to 4.2 volts DC providing continuous battery life to the transmitter.Sensor data from each GPS PTT unit were checked every other day to ensure each unit was not experiencing a significant loss of power.If a unit was found to be relaying GPS coordinates consistently from one location for more than 3 days, or a transmitter had not transmitted for more than 3 days, the unit was tracked via the VHF device attached to the tag until the device was recovered.Argos data were uploaded every 48 h through the Argos system.
Telemetry data were subset into five temporal categories for analysis based on breeding periods: (1) incubation, (2) chick-rearing, (3) post-fledge, (4) breeding season, and (5) all movements.The incubation period was delineated as May 9-May 30 (21-28 days), [7].Since capture was delayed, a week was cut from this period to make up for Guntersville birds nesting earlier than their Great Lake counterparts.Chick-rearing was defined as from May 31 to July 4 (28-30 days) [7], and the post-fledge period was defined as from July 5 to August 15.Because cormorants nested in Guntersville from late March through early June, we combined incubation and chick-rearing into one time period.All movements described the complete time that the individual was tracked.Home range and core use areas were calculated for all movements of these individuals.
All analyses for this study were completed using Program R [31].Home range and core use areas were calculated by finding the utilization distribution (UD) of individuals through kernel density estimation, with no limit for the number of points used [32].Kernel method estimation of the UD does not hold estimates to parametric assumptions so can more accurately estimate densities of any shape by smoothing locations to not over or underestimate distributions [33,34].For tagged cormorants in this study, bivariate normal kernel functions were placed over each relocation point, with all functions averaged for a kernel density estimation of the UD for that individual.A smoothing parameter, h, was also applied to the estimation to control the width of the kernel function placed over the point used.The ad hoc method, or reference bandwidth, was calculated as the h for each Telemetry data were subset into five temporal categories for analysis based on breeding periods: (1) incubation, (2) chick-rearing, (3) post-fledge, (4) breeding season, and (5) all movements.The incubation period was delineated as 9-30 May (21-28 days), [7].Since capture was delayed, a week was cut from this period to make up for Guntersville birds nesting earlier than their Great Lake counterparts.Chick-rearing was defined as from 31 May to 4 July (28-30 days) [7], and the post-fledge period was defined as from 5 July to 15 August.Because cormorants nested in Guntersville from late March through early June, we combined incubation and chick-rearing into one time period.All movements described the complete time that the individual was tracked.Home range and core use areas were calculated for all movements of these individuals.
All analyses for this study were completed using Program R [31].Home range and core use areas were calculated by finding the utilization distribution (UD) of individuals through kernel density estimation, with no limit for the number of points used [32].Kernel method estimation of the UD does not hold estimates to parametric assumptions so can more accurately estimate densities of any shape by smoothing locations to not over or underestimate distributions [33,34].For tagged cormorants in this study, bivariate normal kernel functions were placed over each relocation point, with all functions averaged for a kernel density estimation of the UD for that individual.A smoothing parameter, h, was also applied to the estimation to control the width of the kernel function placed over the point used.The ad hoc method, or reference bandwidth, was calculated as the h for each point.The home range was found by using the 95% use of the individual while the core use area was found by using the 50% use area.We present results as mean ± standard deviation throughout the manuscript and attached tables.Due to a small sample size (n = 8), analyses by sex were not run but mean home range and core use areas were calculated for summary statistics to show differences in ranges between sexes and also by colony affiliation.

Results
During May-July 2017, 10 Solar GPS PTT tags equipped with a VHF device were deployed, with 2916 locations recorded for all birds, 1775 locations from the breeding season and 1140 locations from the post-fledge season.One tag, ID 129279, was deployed twice on two separate cormorants, due to bird mortality on the first bird on 17 July 2017.Three tags were deployed on juvenile birds.Information derived from these tags was not included in this study.
The mean home range for all cormorants for the entire breeding season was 41.8 ± 33.2 km 2 .When separated into periods, the mean home range for the incubation period was 98.9 ± 80.6 km 2 , for the chick-rearing period mean home range size was 18.3 ± 22.6 km 2 and for the post-fledge period, the mean home range size was 42.0 ± 31.0 km 2 .Combining incubation and chick-rearing into one period provided a mean home range of 40.2 ± 32.3 km 2 (Table 1).Home range overlap within breeding periods was within a 3 km radius.The mean home ranges for all movements between males and females were 27.6 ± 31.3 km 2 and 55.9 ± 32.5 km 2 , respectively, and for breeding season was 30.95 ± 24.9 km 2 for males and 52.5 ± 31.6 km 2 for females.Looking at colony affiliation, South Sauty had a mean home range of 59.1 ± 26.8 km 2 and 60.1 ± 25.1 km 2 for all movements and breeding seasons, respectively, while Connors was 12.8 ± 5.5 km 2 and 13.7 ± 5.0 km 2 for the same periods (Table 2).Mean home ranges for breeding sub periods can also be found for cormorant sex and colony affiliation in Table 3.The mean core use area for cormorants for all movements was 6.4 ± 5.9 km 2 .When separated into periods, the mean core use area for the: Incubation period was 14.6 ± 11.2 km 2 ; chick-rearing was 2.8 ± 4.1 km 2 ; post-fledging was 5.5 ± 4.7 km 2 (Table 3).Combining incubation and chick-rearing into one breeding period revealed the mean core use area for this combined time period was 5.0 ± 4.0 km 2 (Table 3).Core use overlaps within breeding periods within a 2 km radius.The mean core use areas for all movements between males and females were 5.4 ± 4.8 km 2 and 7.0 ± 5.0 km 2 , respectively, and for the breeding season, it was 4.6 ± 4.6 km 2 for males and 5.5 ± 3.8 km 2 for females.Looking at colony affiliation, South Sauty had a mean home range of 7.8 ± 4.3 km 2 and 7.1 ± 3.5 km 2 for all movements and breeding seasons, respectively, while Connors was 2.2 ± 1.1 km 2 for both breeding season and all movements (Table 4).The mean core use areas for breeding sub periods can also be found for cormorant sex and colony affiliation in Table 4. Mean home ranges for female cormorants for all movements and during the breeding season, respectively, were 55.9 ± 32.5 and 52.5 ± 38.7 km 2 while for males they were 27.6 ± 30.9 and 31.3 ± 28.7 km 2 (Table 2).Core use areas for female cormorants for all movements was 7.0 ± 5.0 km 2 and for males was 5.4 ± 4.8 km 2 .For the breeding season core use areas were 5.5 ± 3.8 km 2 and 4.6 ± 4.6 km 2 , respectively (Table 4).

Discussion
Compared to previous studies of home ranges of northern breeding cormorants, home ranges on Guntersville Reservoir were much smaller.Studies of cormorants tagged near aquaculture facilities in the Mississippi Delta region report a mean home range of 30,547 ± 6197 km 2 during the summer on their northern breeding grounds [16] compared to a mean 95% home range of just 41.8 ± 33.2 km 2 on Guntersville Reservoir.This was similar in core use areas where cormorants tagged in the Mississippi Delta and breeding up north used over two times the area as cormorants in Guntersville Reservoir [16].Cormorants breeding in Lake Huron foraged in areas of around 3000 km 2 , over double that of birds breeding on Guntersville, though birds breeding on Lake Huron were actively dispersed and subject to egg oiling over the course of the breeding season [15].Additionally, cormorants tagged in the greater Great Lakes area from 2004 to 2007 revealed a mean home range of 1937 ± 5297 km 2 and were also subject to harassment and management [6].Conversely, cormorants breeding in eastern Lake Ontario had 95% kernel home ranges of 42 ± 44 and 75 ± 13 km 2 in 2000 and 2001, respectively, with ongoing egg oiling management [17].Compared to the Great Lakes, Guntersville Reservoir is a small aquatic system and therefore cormorants in this reservoir do not have to travel far for prey.Moreover, since cormorants tend to forage in waters <10 m in depth [10], cormorants in the Great Lakes must find appropriate prey on shorelines or areas of shallow depth.Guntersville Reservoir is shallow, with depths no deeper than 10.6 m outside of the main channel.
For cormorants breeding in the southeastern U.S., island size and distance from the mainland may not be significant drivers of home range as water depth and available loafing areas are readily available and the size of the lentic bodies is much smaller than for other populations such as those that nest on the Great Lakes [16].Additionally, many breeding colonies in the southeast are situated in the same or similar ecosystems to those used as wintering roosting spots [2,7,35,36].It is hypothesized that cormorants that started breeding in the southeastern United States reflect birds that historically migrated through this area but changed their migratory behavior as food resources became available year-round [1,7,25].Cormorant foraging seems to coordinate in areas of possible Lepomis breeding habitats that are present at Guntersville Reservoir.Though different in size, water depth is likely just as important in the southeastern United States as it is in the Great Lakes, for optimal foraging as well as loafing areas to dry off and rest between foraging sessions [7,37].The smaller home range size may be more a function of foraging on an inland reservoir that is smaller than the geographic extent of aquaculture in the MS Delta or areas of the Great Lakes.
Within Guntersville Reservoir, the size of cormorant home ranges differed among the defined periods of incubation, chick-rearing, and post-fledge though the areas used by individual cormorants did not differ and were consistent, within 2.5 km, throughout the season.Cormorants had the largest home ranges during the incubation period while the smallest home ranges were observed during the chick-rearing period across all colony types and sex (Table 1).These differences could be accounted for by the sex of the cormorants tracked.Studies have suggested that male cormorants forage farther from colonies during incubation than females and vice versa for females during chick rearing [18,38].This was not apparent in our study, where females foraged almost twice as far as males during incubation and had similar distances for chick-rearing (Tables 3 and 4).Additional energy constraints may affect females more than males as the energy needed for building nests and laying and incubating eggs rests largely on the adult female [18,38].Parents could be foraging closer to the nest during chick rearing and make shorter trips for fish to feed chicks, accounting for the lower home ranges during this period.Overall, the mean home range and core use area of males (27.6 ± 31.3 km 2 ) for all movements on Guntersville Reservoir were smaller than that of females (55.9 ± 32.5 km 2 ; Table 3).Additionally, colony affiliation may have accounted for some of these large differences in home range size, with South Sauty birds having larger home range (59.1 ± 26.8 km 2 ) and core use areas (7.8 ± 4.3 km 2 ) than their Connors Island counterparts (12.8 ± 5.5 km 2 and 2.2 ± 1.1 km 2 ), though some individual birds varied within these areas (Figures 1 and 3 and Tables 1-4).This may be explained due to Connors having a larger breadth of water to forage on, and depths of 0.3 m compared to 4.5 m around surrounding colonies such as South Sauty.Conversely, South Sauty is situated in a much narrower area within Guntersville and birds may travel farther north or south to forage, though water depths do not differ between these two sites.
km 2 ) than their Connors Island counterparts (12.8 ± 5.5 km 2 and 2.2 ± 1.1 km 2 ), though some individual birds varied within these areas (Figures 1 and 3 and Tables 1-4).This may be explained due to Connors having a larger breadth of water to forage on, and depths of 0.3 m compared to 4.5 m around surrounding colonies such as South Sauty.Conversely, South Sauty is situated in a much narrower area within Guntersville and birds may travel farther north or south to forage, though water depths do not differ between these two sites.

Conclusions
Managers in the southeastern United States who are monitoring and controlling expanding cormorant populations on reservoirs can use data derived from telemetered birds to understand how these birds move on the landscape and what characteristics are present in their home ranges.Knowing that cormorants breeding on Guntersville reservoir use smaller areas than northern birds nesting in the Great Lakes helps managers focus on a much smaller and precise scope to manage cormorants and implement solutions for better and sustainable conservation.Additionally, our results support current data that cormorants forage in shallow water, depths ≤ 5 m, and take advantage of loafing areas scattered around colony sites, relationships similar for cormorants nesting in the Great Lakes.These contrasts between southern and northern breeding birds are pivotal in managing and understanding not only the behavior and movements of cormorants but also gives insight into the expanding range of cormorants in southeastern reservoirs.

Conclusions
Managers in the southeastern United States who are monitoring and controlling expanding cormorant populations on reservoirs can use data derived from telemetered birds to understand how these birds move on the landscape and what characteristics are present in their home ranges.Knowing that cormorants breeding on Guntersville reservoir use smaller areas than northern birds nesting in the Great Lakes helps managers focus on a much smaller and precise scope to manage cormorants and implement solutions for better and sustainable conservation.Additionally, our results support current data that cormorants forage in shallow water, depths ≤5 m, and take advantage of loafing areas scattered around colony sites, relationships similar for cormorants nesting in the Great Lakes.These contrasts between southern and northern breeding birds are pivotal in managing and understanding not only the behavior and movements of cormorants but also gives insight into the expanding range of cormorants in southeastern reservoirs.

Figure 1 .
Figure 1.Guntersville Reservoir within Alabama with map at right showing islands colonized, abandoned, and not used by nesting Double-crested Cormorants during March-August 2017.

Figure 1 .
Figure 1.Guntersville Reservoir within Alabama with map at right showing islands colonized, abandoned, and not used by nesting Double-crested Cormorants during March-August 2017.

Figure 2 .
Figure 2. Picture illustrating placement of transmitter on an adult Double-crested Cormorant secured using a Teflon ribbon.

Figure 2 .
Figure 2. Picture illustrating placement of transmitter on an adult Double-crested Cormorant secured using a Teflon ribbon.

Figure 3 .
Figure 3. Home range and core use areas of for male Double-crested Cormorants at breeding colonies on the left (n = 4) and females on the right (n = 4) for the complete period of study March-August 2017.Different color lines differentiate individual birds and line types differentiate 95% (solid line) or 50% (dotted line) home range.

Figure 3 .
Figure 3. Home range and core use areas of for male Double-crested Cormorants at breeding colonies on the left (n = 4) and females on the right (n = 4) for the complete period of study March-August 2017.Different color lines differentiate individual birds and line types differentiate 95% (solid line) or 50% (dotted line) home range.

Table 1 .
Ninety-five percent home ranges (km 2 ) measured for all movements of Double-crested Cormorants (n = 8 birds) on Guntersville Reservoir, Alabama in 2017.Breeding season (n = 7) is incubation (n = 6) and chick-rearing periods (n = 6), combined.Post-fledge (n = 4), colony affiliation refers to a nesting site.Dashes indicate no data for that time period.

Table 2 .
Means and standard deviation (SD) of 95% home range (km 2 ) of Double-crested Cormorants on Guntersville Reservoir for female, male, and colony affiliation by breeding periods (in bold first column and top rows).

Table 3 .
Fifty percent home ranges (km2) measured for all movements of Double-crested Cormorants (n = 8 birds) in Guntersville Reservoir, Alabama in 2017.Breeding season (n = 7) is incubation (n = 6) and chick-rearing periods (n = 6), combined.Post-fledge (n = 4), colony affiliation refers to a nesting site.Dashes indicate no data for that time period.

Table 4 .
Means and standard deviation (SD) of 50% core use areas (km 2 ) of Double-crested Cormorants on Guntersville Reservoir for female, male and colony affiliation by breeding periods (in bold first column and top rows).