A Window into the Breeding Ecology and Molt of the Eastern Black Rail (Laterallus jamaicensis jamaicensis)

Abstract. Knowledge of the ecology of the Eastern Black Rail (Laterallus jamaicensis jamaicensis) has remained nearly as elusive as the rail itself. Camera trapping methods facilitated the first study of breeding phenology and chick development, courtship and brood rearing behaviors, and flightless molt phenology and duration. Broods (n = 33) were observed between August 2015 - September 2019. Chicks were capable of flight at approximately 40 days after hatching. Nesting was initiated as early as 17 April 2019 ( x̄ = 5 June ± 30.0 SD) and fledging occurred as late as 30 September 2019 (x̄ = 10 August). Behavioral observations combined with phenology data provided evidence of pairs raising two or more broods during a breeding season. Flightless molt (n = 10 adults in molt), which was initiated as early as 15 August 2019 and completed as late as 11 October 2019, was completed within approximately 21 days of initiation. Conservation and management strategies should take into consideration periods of vulnerability, which coincide with increasingly severe and frequent coastal flooding events and hurricanes. It is necessary to understand factors key to fecundity and survival to effectively develop conservation strategies to ensure the persistence of the subspecies.

Breeding phenology includes the timing of breeding behaviors such as arrival on breeding grounds, courtship and copulation, egg laying and incubation, chick rearing, and fledging (Dunn and Moller 2014). However, in less studied species, the timing of these fundamental biological events often remains unknown. Rails (Rallidae) are one of the most poorly represented groups in ornithological research (Brambilla and Jenkins 2009;Stermin et al. 2012). Ripley (1977) remarked that very little is known about the habits of the Black Rail (Laterallus jamaicensis) and of species within the genus Laterallus, collectively, due to their tiny size and proclivity for secretiveness. Two subspecies of Black Rail are known to occur in the United States; California Black Rail (L. j. coturniculus) in California and Arizona, USA, and Eastern Black Rail (L. j. jamaicensis) in states east of the Rocky Mountains in the USA (Eddleman et al. 2020).
The Eastern Black Rail was listed as Threatened under the Endangered Species Act on 8 October 2020 (U.S. Fish and Wildlife Service 2020), and the subspecies is likely to be extirpated within the next 50 years under current conditions (U.S. Fish and Wildlife Service 2019). Conservation efforts to prevent extirpation and promote the reproductive success of the Eastern Black Rail (hereafter Black Rail) will require an understanding of their phenology, fecundity, and survival; information which is scarce throughout the range.
Our objectives were to document the phenology of nesting, brood rearing, and definitive prebasic molt, during which adult Black Rails lose all remiges and retrices simultaneously, becoming temporarily flightless (hereafter flightless molt; Pyle 2008). Nesting and molting are two periods during which Black Rails have high energy requirements, are less able to move among wetlands, and are likely to be particularly vulnerable (U.S. Fish and Wildlife Service 2019). In addition to achieving the primary objective, this study describes previously undocumented aspects of Black Rail ecology and behavior, which significantly advances our limited understanding of Black Rail biology.

Study Area
We studied the breeding ecology of the Black Rail at two sites in coastal South Carolina, USA: the Bear Island Wildlife Management Area and adjacent tidal marshes (BIWMA; 4,800 ha; 32° 35ʹ N, 80° 27ʹ W) and the Tom Yawkey Wildlife Center (TYWC; 9,700 ha; 33° 14ʹ N, 79° 16ʹ W). BIWMA is located in Colleton County approximately 50 km southwest of Charleston, South Carolina in the core of the Ashepoo, Combahee, and Edisto rivers (ACE) Basin. TYWC is located in Georgetown County approximately 80 km northeast of Charleston on islands at the confluence of the South Santee and North Santee Rivers, Winyah Bay, and the Atlantic Ocean. BI-WMA and TYWC are owned and managed by the South Carolina Department of Natural Resources. We selected three wetlands at BIWMA and two wetlands at TYWC, and designated between one and six areas within each wetland as monitoring equipment arrays (arrays; Table  1). Sampling within these wetlands occurred between 6 August 2015 and 31 December 2019.
To select array locations, which encompassed entire wetlands during 2015 -2016 and encompassed specific areas within each wetland during 2017 -2019, we identified potential Black Rail territories based on microhabitat characteristics (Spautz et al. 2005;Eddleman et al. 2020), results from call-response surveys (Legare et al. 1999;Conway et al. 2004) conducted in South Carolina (Cely et al. 1993;Roach and Barrett 2015;Hand 2016) and opportunistic auditory and visual observations of Black Rails. Black Rails frequented the impounded brackish marsh wetlands (hereafter impoundments; Wetlands 1, 3, 4, and 5) and salt and brackish tidal marshes (hereafter tidal marshes; Wetland 2) at BI-WMA and TYWC. Impoundments at the study sites are managed primarily for the purpose of providing wintering habitat for waterfowl using water level manipulation and prescribed burning.
Occupied microhabitat within the array sites was characterized by very shallow water (typically 0.5-5 cm) and dense emergent vegetation. Dominant species included clump cordgrass (Spartina bakeri), saltmarsh bulrush (Schoenoplectus americanus and S. robustus), saltmeadow cordgrass (S. patens), saltgrass (Distichlis spicata), and narrowleaf cattail (Typha angustifolia; Wetlands 1 and 3), or smooth cordgrass (S. alterniflora), saltgrass, saltmeadow cordgrass, and black needlerush (Juncus roemerianus; Wetlands 4 and 5) in impoundments or marsh fimbry (Fimbristylis castanea) and black needlerush (Wetland 2) in tidal marshes. The lowest-elevation areas within occupied territories in all wetlands flooded to depths above 20 cm numerous times during the study resulting from heavy rainfall (both wetland types), high tides (tidal marsh), and water level management (impoundments). Patches of clump cordgrass and/or black needlerush were present in all wetlands and were adequately tall (>1 m) and dense enough to provide temporary refugia during flooding events. Each wetland was separated from other potentially suitable habitat by persistently-flooded perimeter ditches and often also by upland habitat. Although flighted adult Black Rails likely move among wetlands in response to changing habitat conditions, it is unlikely that chicks or adults in flightless molt would be able to move among wetlands.
Fires occurred in areas of Wetland 1 during the study. The area of Wetland 1 containing Array 1c (ca. 5 ha high marsh habitat) was separated from the majority of potential Black Rail habitat within the wetland by persistently flooded ditches, and was intentionally burned on 24 February 2017 to reduce dense thatch within the vegetation. The largest continuous area of Wetland 1 (ca. 50 ha high marsh habitat), which contained Arrays 1a, 1b, and 1d, was burned on 28 February 2018 and again on 26 February 2019 when prescribed fires in adjacent upland areas escaped into the impoundment. We estimate over 70% of the emergent vegetation was completely burned during each fire, however; charred bases of clump cordgrass remained throughout the wetland. The substrate remained saturated with water during the fires, leaving the soil, roots, and rhizomes unburned. The vegetation burned in all areas of Wetland 1 during the study was dominated by clump cordgrass interspersed with patches of narrowleaf cattail and saltmarsh bulrush. These species began to reemerge within seven days after each fire, and by June grew to > 1-m in height.

Motion-activated Camera Traps
To determine the phenology of breeding and flightless molt, we deployed and maintained motionactivated camera traps within each array. Breeding by Eastern Black Rails generally occurs during the spring and summer and flightless molt follows the breeding season (Eddleman et al. 2020). Sampling was typically initiated during February or March and concluded during August or September due to recurring severe storms during each year of the study. Camera traps were redeployed within one month following the storms in a subset of arrays maintained year-round. To maximize Black Rail detections, and due to the descriptive nature of the study, camera trap locations within arrays were not randomized or stratified. Instead, the placement of camera traps was deliberately biased (Meek et al. 2014), targeting what appeared to be ideal microhabitat based upon previous observations of Black Rails.
During 2015-2016, arrays consisted of a variable number of camera traps, which were moved frequently within and between wetlands. Beginning in January 2017, we implemented a more standardized sampling approach, which is described below. Arrays typically consisted of four Reconyx HyperFire Professional Covert IR camera traps (Models PC900 and HP2X), which were supplemented with additional infrared camera traps with video and audio recording capabilities (Advanced Security SSC-24C36, Bushnell Na-tureView 119439, 119440, and 119740, Bushnell Trophy Aggressor 119777C and 119877C, and Reconyx UltraFire XP9). Reconyx HyperFire camera traps (x -= 5.6 ± 2.8 SD) were deployed 5-10 m apart and any supplemental camera traps (x -= 1.9 ± 1.7 SD) were deployed 0.5-15 m from Reconyx HyperFire camera traps. All camera traps in the array were within a 100 m radius from the array center. Black Rail detections from arrays spaced < 100 m apart were combined because the territories of individual rails appeared to encompass portions of multiple arrays. Camera traps were mounted between 0.15 m and 0.75 m above the ground and aimed toward naturally occurring trails or gaps in clumps of vegetation, which Black Rails use to navigate the wetland (Weske 1969). Camera traps were occasionally relocated within and among arrays throughout the duration of the study to optimize Black Rail detection as water levels and suitable vegetation structure fluctuated. Reconyx Hyperfire camera traps were programmed to take ten photographs per trigger during 2015 -2017 and 5 photographs per trigger during 2018 -2019. Other camera models were programed to take three photographs followed by a 10 -60 sec video recording per trigger. Camera traps were programed to operate 24 hrs per day with no quiet periods, and Reconyx Hyperfire camera traps were programed to take a timelapse photo every four hrs in addition to motion-triggered photographs.
We used CPW Photo Warehouse (Newkirk 2016) to organize and review photographs and to perform data summarizations. Photographs of Black Rails were categorized by sex and age based on plumage characteristics (Pyle 2008;Eddleman et al. 2020) when possible .

results
Black Rails were photographed in 86% of arrays (n = 14, Table 1) and 80% of wetlands (n = 4) during 2017-2019. Sampling effort during this period totaled 46,033 camera trap days and resulted in a total of 2,040,536 photographs, 30,841 (1.5%) of which contained Black Rails.

Brood Size
In broods documented within 6 days of hatching, brood sizes ranged from a minimum of one chick to seven chicks (x -= 3.4 ± 1.8 SD, n = 16). For chicks > 10 days old when initially detected, we did not estimate brood size because chicks ventured farther from parents and siblings after frequent brooding was no longer required.

Chick Development
We documented the rate of chick development and characterized key physical and behavioral developmental stages ( Fig. 1; Appendix) of broods photographed repeatedly between hatching and fledgling. Age was calculated based on chicks remaining in the nest for approximately 24 hrs (Davidson 1992) and the developmental stage of the chicks photographed. Fledging occurred at approximately 40 days after hatching. Asynchronous development was documented among chicks within one brood in Array 1b during each year of the study (n = 3 broods), resulting in a difference in fledg-ing dates as great as 3 days among siblings. Developmental differences were first noted between days 14-19 after hatching and persisted through fledging in the two asynchronous broods from which multiple chicks were photographed together throughout their development. Chicks that appeared to be developing asynchronously were also observed in Wetland 2 but were not photographed together. Although most broods appeared to develop synchronously, we were unable to definitively confirm synchronous development occurred because additional chicks may have been present but not photographed.

Breeding Phenology
Black Rail chicks were detected between May and September during 2015 -2019 (Fig.  2). Estimated hatch dates ranged from 1-13 May-20 August (x -= 1 July ± 30.0 SD, n = 33; Table 2), suggesting incubation occurred between April and August based on an approximated 26-day period between the initiation of egg laying and hatching (Flores and Eddleman 1993;Legare and Eddleman 2001;U.S. Fish and Wildlife Service 2019). In arrays where vegetation growth was delayed by fire during February 2018 and February 2019 (Arrays 1a, 1b, 1d, 1e, and 1f), the mean estimated average hatch date was 18 July (± 28.1 SD, n = 9) compared to an average of 25 June (± 30.8 SD, n = 24) in arrays not burned. At least one chick survived to 15 days post-hatching in a minimum of 30 broods, and fledging was confirmed for eight broods. We did not determine if chicks from the remaining 25 broods fledged, however; based on hatch dates and a period of 40 days between hatching and fledging, we estimated potential fledging dates ranged from 23 June to 30 September (x -= 10 August). Fledglings were photographed on 27 June 2017, 16 July-7 September 2018, and 11 August-7 October 2019.

Flightless Molt Phenology
Flightless molt (Fig. 3) was documented in Wetland 2 in 2016 and in Arrays 1a, 1b, and 2d in 2018 and Arrays 1b and 2a during 2018 and 2019. The duration of molt was approximately 21 days from the initiation of primary feather loss to the completion of primary feather regrowth. We observed ten flightless adults, five of which molted in habitat burned during the prior February. In the three instances when pairs were documented molting, the paired adults molted nearly simultaneously with the females beginning 2 to 3 days later compared to males. Molt initiation dates ranged from 15 August to 20 Sep-tember (x -= 30 August ± 14.0 SD, n = 10; Table 3), with estimated molt completion between 5 September and 11 October. Molt was initiated an average of 20 days later by individuals observed in recently-burned habitat.

Behavioral Observations
Courtship behaviors included offering food (n = 11 male to female, n = 1 female to male), allopreening (n = 7 male preening female, n = 1 female preening male), and pursuit of the female by the male (n = 25).  Throughout the study, we documented seven copulations. During one copulation (Fig.  4), adult male and female Black Rails were simultaneously photographed and filmed. Concluding copulation, the male tumbled from the female and circled her while walking with a lowered head and raised wings. As the male circled, the female bowed forward with raised tail feathers then ruffled her feathers before beginning to preen.
Throughout the study, this raised wing display by males was observed following three copulations and females were observed ruffling feathers and preening following four copulations. Males closely pursued females prior to five copulations.
In seven instances during the study, two or more broods were photographed in the same array during a single breeding season. Copulations occurred while young Table 2 Array 1a, 1b, and 1d  chicks were present in arrays during two such instances. Following copulation on 5 June 2017 in Array 2a, described above, adults departed then returned to the same location 53 min later with downy chicks. On 18 July 2017, downy chicks from a second brood were documented in the array. Estimated hatch dates for the first and sec-ond broods were 41 days apart. The estimated hatch date for a third brood in the array was 34 days after the hatch date for the second brood.  During 2019, four broods were documented in Array 1b. It is feasible that a single pair hatched the first and third broods or second and fourth broods, however, the number of days between estimated hatch dates for the first and second broods and the third and fourth broods were 23 and 14 days apart, respectively. The breeding phenology within Array 1b suggests breeding territories of two or more pairs overlapped within the camera array. Chicks from different broods were observed in close proximity to each other. On 30 August 2019, a 10-dayold chick from the fourth brood closely following an adult appeared in the same photograph as a partially feathered chick presumably from the third brood in the array. Chicks from both broods continued to be photographed in the Array throughout September.

Table 3. Flightless periods of definitive prebasic molt in adult Eastern Black Rails (Laterallus jamaicensis jamaicensis) in Colleton County, South Carolina, USA. Initiation dates were estimated based on plumage characteristics observed in photographs by camera traps, and completion dates are based on an approximately 21 day flightless period. Accuracy of estimated molt timing is likely higher in instances when individuals were photographed on multiple dates. Emergent vegetation in Array 1a and 1b was burned on 28 February 2018 and again on 26
We documented 57 interactions between adult Black Rails and their chicks (Fig. 4). During these interactions, both adults (n = 7), a male (n = 22), a female (n = 12), or an adult of unknown sex (n = 16) were present. The most frequently observed chick-adult behaviors were chick(s) following an adult (n = 49), being brooded (n = 11), being fed (n = 7), and being preened (n = 1). We observed multiple behaviors during 14 interactions. Chicks solicited brooding and feeding by adults by bowing, fluttering their wings and vocalizing. Broods of downy chicks were frequently observed in a group sitting on the ground and preening themselves, and siblings occasionally pecked each other. Young continued to be observed within natal territories beyond fledging but were not observed receiving parental care after 25 days post-hatching. Although most chicks over 25 days old were alone in photographs, we observed two chicks from one brood brooding each other at 26 days and allopreening at 37, 38, and 39 days. Their allopreening behavior closely resembled allopreening courtship behavior between adults.

discussion
Vocalization-based surveys are the predominant tool used to study secretive marsh birds (Taylor and van Perlo 1998;Conway 2011), but have a limited ability to determine basic biological characteristics, such as the phenology of incubation, brood rearing, and flightless molt. Camera trapping is emerging as a valuable tool for the study of rails (O'Brien and Kinnaird 2008;Colyn et al. 2017;Shirkey et al. 2017;Znidersic 2017;Colyn et al. 2019), and has the ability to capture the chronological development of these characteristics. We successfully applied camera trapping techniques to the study of the Eastern Black Rail in coastal South Carolina and conducted the longest sustained sampling effort of any kind for the subspecies. Our results provide novel information about their nesting and breeding phenology, of which very little is known. Prior to our study, only 23 observations of adults with broods and 147 egg records spanning 170 years were available for the Eastern Black Rail in the USA (Watts 2021). Most nesting records for Black Rails come from historical accounts by egg collectors (Bent 1926), which do not include lay or hatch dates (Eddleman et al. 2020). We have contributed 33 hatch date records, more than any previous study of the species, and the only Atlantic coast records collected between 2010 and 2019 (Watts 2016).
Our breeding phenology estimates were based on photographic documentation of chicks in their respective broods rather than observations of eggs or nests. Chick growth and development in the Black Rail was virtually unknown prior to our study (U.S. Fish and Wildlife Service 2019;Eddleman et al. 2020), therefore we first needed to document the duration and progression of development from hatching to fledging before we could estimate hatch dates. In the only prior study of Laterallus chick development, Franklin et al. (1979) described physical development of Galapagos Rails (L. spilonota) based on eleven chick captures, ten juvenile captures, and ten adult captures; however, in their study development was based on body weight rather than age post hatching, precluding direct comparisons of developmental timing. Chicks in our study exhibited a similar progression of development of eye and bill color compared to Franklin et al. (1979). During our study, we determined fledging occurs at approximately 40 days after hatching. This timeframe differs slightly from Eddleman et al. (2020), who noted that juvenile plumage is thought to be attained during the first 6 weeks while also acknowledging the timing of development was uncertain. While Black Rail eggs are currently thought to hatch synchronously (Walker 1941;Davidson 1992;Eddleman et al. 2020), we observed asynchronous development among siblings, demonstrating a small amount of variability in developmental rates. This asynchrony may be related to habitat conditions and food availability during the egg laying or brood rearing periods. Further research is needed to elucidate the relationship between developmental rates and food availability.
The photographs and videos we captured of breeding behaviors of the Black Rail have provided valuable insights into their fecundity. Behavioral observations in conjunction with our phenology data suggested pairs successfully raised as many as three broods during a single breeding season and corroborate previous findings of Black Rails laying two or more clutches in one breeding season. Flores and Eddleman (1993) reported a female California Black Rail with an egg in her oviduct 18 days after her previous brood hatched and a male incubating eggs in both April and August. Second broods are not unusual in Rallidae (Taylor and van Perlo 1998) and have been reported in at least two of the remaining eight Laterallus species. Franklin et al. (1979) reported a banded Galapagos Rail that was associated with two distinct broods during a breeding season, and in captivity, Everitt (1962) observed a pair of Red-and-white Crakes (L. leucopyrrhus) raise two broods successfully. The apparent ability of the Eastern Black Rails to produce large clutches of eggs, nest throughout five months in the lower latitudes of their range, and successfully raise as many as three broods suggests the subspecies has the potential for high fecundity in all or part of their range (Todd 1977;Flores and Eddleman 1993;Legare and Eddleman 2001).
Prescribed fire is a widely-used tool for vegetation management in coastal marshes throughout the range of the Eastern Black Rail (Davidson 1992;Eddleman and Legare 1995;Conway et al. 2010;Haverland 2019). In coastal South Carolina, prescribed fire and water-level manipulation (periodic flooding) are used to manage vegetation structure and species composition within impoundments (Folk et al. 2016). In contrast to prior studies examining the effects of fire within the range of the Eastern Black Rail, which suggested Black Rails did not return to burned habitat until a minimum of 27 months post-burn (Haverland 2019), we documented Black Rails beginning to breed in burned high marsh habitat approximately 2.5 months following fires during February in our study area. Nine broods of flightless chicks which hatched within 4 to 6 months post-burn, including three broods which appeared to be second or third broods, were documented in the burned area of Wetland 1. We concluded these broods hatched from nests within the burned habitat based on the following evidence: adult Black Rails were documented during the egg laying and incubation periods of six of the broods, three of the broods were detected at less than 7 days old, and the burned area was separated from all unburned potential habitat by persistently-flooded ditches. Vegetation including clump cordgrass regrew rapidly in Wetland 1, reaching mature height within the same growing season as the burns (Hand et al. 2020). Conway et al. (2010) documented similarly rapid vegetation regrowth and detected California Black Rails during callresponse surveys within three months postburn in fresh-water marshes in the Lower Colorado River Basin, Arizona, USA. Soil type and moisture level can affect the rate of regrowth in wetland plant species including Spartina (McAtee et al. 1979) and should be considered when planning prescribed fires and burn intervals. Optimal burn intervals in impoundments in South Carolina may differ from optimal intervals in tidal marsh habitat in other portions of the range of Eastern Black Rails.
All North American Rallidae species are temporarily flightless at the end of their breeding season during definitive prebasic molt (Pyle 2008). Prior to our study, the only observation of flightless molt in the Black Rail was a male California Black Rail captured in Arizona during September 1987or 1988(Flores and Eddleman 1991Eddleman et al. 2020). Pyle (2008) asserts that North America Black Rail populations undergo definitive prebasic molt during July through September, however, the Eastern Black Rails observed in our study molted flight feathers during August through October. As evidenced by decreased body weights observed during late summer in California Black Rails (Flores and Eddleman 1991) and the high energetic and nutritional costs of flight feather replacement documented in species with similar molt strategies (Guillemette et al. 2007), adult Black Rails may be at increased risk of starvation and predation during this period if habitat conditions are not favorable. Avian species breeding in unpredictable habitats may delay remigial molt when breeding is protracted or delayed (Keast 1986; Hahn et al. 1992). The initiation of breeding during our study was delayed in recently-burned habitat compared to unburned habitat. Negative consequences to adult survival may occur if flightless molt is delayed into late September and October, when tropical cyclones most frequently occur along the South Carolina coast (National Oceanic and Atmospheric Administration (NOAA) 2021).
The increased frequency and severity of flooding events resulting from sea level rise and climate change are among the greatest threats to the productivity and survival in tidal marsh-dependent avian species in the Eastern USA (Atlantic Coast Joint Venture 2019). To understand and predict the responses of Black Rail populations to these threats, we must first recognize when these events coincide with periods of elevated vulnerability such as breeding and flightless molt. More than 75% of the remaining Eastern Black Rails in Atlantic and Gulf Coast states are believed to breed and molt in coastal Texas, Florida and South Carolina (Watts 2016).
During the three years of our study, all areas known to be occupied by Black Rails in these states (U.S. Fish and Wildlife Service 2019) experienced expansive flooding resulting from tidal surge and/or rainfall associated with hurricanes on one or more occasions during the timeframe when we documented flightless molt in coastal South Carolina. These areas are also experiencing increasingly frequent and severe high tide flooding due to sea level rise. Sweet et al. (2020) reported that the frequency of high tide flooding in the USA. doubled between 2000 and 2020, with Atlantic and Gulf Coast locations flooding at twice the national rate. We suggest continued and expanded monitoring of breeding and molt phenology using camera trapping techniques, which are relatively noninvasive compared to capturing and examining rails in hand and are now proven in their ability to successfully monitor Black Rails. Our study provides novel resources that will facilitate the collection of phenology data in other areas of the Eastern Black Rail's range and the development of conservation strategies that maximize potential benefits to Eastern Black Rail populations.

acknoWledgMents
We thank the South Carolina Department of Natural Resources, the U.S. Department of the Interior, Fish and Wildlife Service, and private landowners for their logistical and/or financial support. Funding was primarily provided through the U.S. Department of the Interior, Fish and Wildlife Service, Section 6 Grant Program (E-1-38, F17AP01001 and E-F19AP00004), State Wildlife Grant Program (SC-T-F17AF01208), and by the South Carolina Department of Natural Resources. We are grateful to the field technicians who spent countless hours maintaining camera traps and reviewing photographs. In particular, we thank K. Gundermann, S. Chandhok, C. Adams, C. Powell, C. Worthington, B.