Reproductive ecology of interior least tern and piping plover in relation to Platte River hydrology and sandbar dynamics

In a recent study, Farnsworth et al. (2017) used distributions of nest initiation dates drawn mostly from human‐created, off‐channel habitats and a model of emergent sandbar habitat to evaluate the hypothesis that least terns (Sternula antillarum) and piping plovers (Charadrius melodus) are physiologically adapted to initiate nests concurrent with the cessation of spring river flow rises on two sections of the Platte River, Nebraska. The study by Farnsworth et al. (2017) has several shortcomings which bring into question the authors’ principal assertion that interior least tern and piping plovers are not adapted to occupying and nesting on river sandbars on the Platte River system. We identify these shortcomings and provide information, which, we suggest, would change their conclusions if incorporated. Linked Article: https://doi.org/10.1002/ece3.4097


| FORMATIVE RIVER S TAG E , EMERG ENT SANDBAR HEI G HT, AND NE S TING HEI G HT
Elevation of sandbars relative to river stage is a foundational component of the authors' analysis as it determines whether habitat will be available or unavailable (i.e., emergent sandbars exposed above river flow level or sandbars that are fully inundated) for nesting. The sensitivity analysis presented by Farnsworth et al. (2017) showed that assumptions of sandbar heights (depth below peak river flow stage, hereafter referred to as a "stage gap"; see Figure 1 herein) accounted for the clear majority (>90%) of the variance in their emergent sandbar habitat nesting success window estimates. The authors' stage gap assumptions and applications are problematic because of (1) the decision to not describe sandbar height data collection and analysis methods for unpublished values, (2) the assumption of a constant stage gap for each study reach despite empirical evidence to the contrary, and (3) the assumption that most nests are placed at the mean sandbar height.
The authors used mean values for the stage gap, one published (Alexander, Schultze, & Zelt, 2013)  The authors' assumption of a constant mean value for the magnitude of the stage gap in each reach of the Platte River ignores evidence, suggesting a pattern of increasing stage gap with increasing discharge. Previous studies (Brice, 1964;Cant & Walker, 1978;Mohrig and Smith 1996;Smith 1971) indicate that sandbars submerged during low-magnitude discharges often have shallow gaps at their crests (0.10 m or less; Figure 2). Observations of sandbars during (Ashworth et al. 2000;Crowley 1983) and following (Alexander et al., 2013) moderate-to high-magnitude flow events demonstrate that the stage gap can be as much as 1 to 2 m. This concept is illustrated in figure 8 of Alexander et al. (2013), which shows This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.  Ziewitz et al. (1992), which shows that median and maximum nest elevations were safe from inundation in 40% and 90% of years, respectively (measurements were made in CPR andLPR, 1958-1988). As terns and plovers select higher sandbars and nest in higher locations on those sandbars, the number of years with successful nesting windows is certainly higher than those reported by Farnsworth et al. (2017).

| COMPARISON OF AVER AG E PL AT TE RIVER HYDROG R APH WITH NE S T INITIATI ON DATE D IS TRIBUTI ON S
In section 3.1 of Farnsworth et al., the authors use an overlay of the long-term mean daily hydrograph (long-term mean daily discharge F I G U R E 1 Illustration of the concept of a "stage gap" between the elevation of the top surface of an emergent sandbar and the elevation of the water surface (stage) during the annual peak discharge when the bar formed. Note that both nesting sites are on the high platform of the bar surface, but the slight topographic variation in the high platform results in different stage gaps and therefore different potential for flooding at each nesting site. Note also that the mean sandbar elevation may or may not be representative of the nesting elevation F I G U R E 2 River-level photograph of emergent and submerged (active) sandbars in the wide, braided, Niobrara River of northern Nebraska. The photograph was taken during baseflow conditions in August of 2014. The water depth over the top of the submerged sandbar in the foreground ranged from approximately 3-10 cm. The slipface of the submerged sandbar is marked by the vertical sticks. Note the flat surface of the emergent sandbar in the background; the high platform is the area above the top of the scalloped margin of the sandbar. The emergent sandbar is approximately 40-50 m long values for each day of the year) for segments of the Platte River with distributions of nest initiation dates for both species (figure 8 in Farnsworth et al.) to assert that the annual spring rise typically occurs after the nest initiation date for both species. The authors use this simple overlay to suggest (see abstract, section 3.1. and discussion of that paper) that the hydrology of the Platte River creates adverse physical conditions for nesting because the typical spring rise would occur after nest establishment and, due to the large stage gaps assumed by the authors, typically inundate established nests.
Although the mean daily hydrograph can be useful for understanding basic hydrologic patterns at a location in a river, such hydrographs mask variability, particularly in the timing of the annual instantaneous peak flow, which is the typical emergent sandbar habitat formative event. For example, the mean daily hydrograph The authors' distributions of nest initiation dates only include data from "all on-channel and off-channel" (Farnsworth et al.,page 2) from the CPR for the years 2001-2013. Although not stated in their paper, nearly all (more than 96%, n = 1,089) of the nests reported in the CPR during this 13-year period were found on human-created off-channel habitats (mostly sand and gravel mines; Baasch, 2014;Howlin, Strickland, & Derby, 2008), where suitable nesting habitat is always available when terns and plovers arrive in spring. Using nest initiation data from static, human-created, off-channel habitat is an incomplete representation of the species' breeding phenology and range of nest initiation dates. This can easily result in incorrect or misleading conclusions when applied to species' behavior in dynamic river systems where nesting habitat is not always available for nesting upon the birds' arrival in spring. Nest initiation in many avian species (e.g., Gilbert & Servello, 2005), including terns and plovers (Elliott-Smith & Haig, 2004;Thompson et al., 1997), is variable and occurs in response to environmental conditions. For example, least tern nest initiation on the LPR from 2008 to 2013 occurred later at river habitats (median = 16 June) compared to off-channel habitats (median = 10 June, t 1,193 = 4.97, p < .001; JGJ, MBB, pers. obs.). Least tern mean nest initiation dates on the Yellowstone River, Montana, where off-channel habitats are not available, occurred 16 June, 30 June, and 1 July in 1994, 1995, and 1996, respectively, following cessation of spring rises that occurred as late as mid-to late June (Bacon & Rotella, 1998 Farnsworth et al. suggest that meeting or exceeding reproductive rates (fledge ratios) found in a report (Lutey, 2002) are necessary to maintain "stable to growing populations" of piping plovers and least terns along the Platte River. They provide calculations that purport to show the biologically improbable reproductive rates (e.g., 7.06 fledglings/pair for piping plovers) regularly needed during the years when their hydrological analysis suggests nesting was possible on the Platte River. These calculations led the authors to their principal conclusion that the historical CPR was, and contemporary LPR is, incapable of supporting least tern and piping plover populations.

| TERN AND PLOVER P OPUL ATI ON ECO LO GY
The analytical approach used by the authors is too simple to address complex questions about metapopulation dynamics.
The authors' calculations incorrectly assume closed populations (or that immigration and emigration are equal) within the CPR and within the LPR, which is not valid because (1) (Kirsch & Sidle, 1999).
Piping plovers and least terns are capable of dispersing widely and occupying nesting habitats over broad spatial scales (Catlin et al., 2016;Elliott-Smith & Haig, 2004;Hunt et al., 2015;Roche, Gratto-Trevor, Goossen, & White, 2012;Roche et al., 2016;Thompson et al., 1997;Ziegler et al. 2017). Both species are relatively long-lived and can experience high reproductive success and high reproductive failure (Elliott- Smith & Haig, 2004;Thompson et al., 1997). These are significant aspects of both species' life history strategies that allow them to occupy and persist in dynamic environments. Both species will renest if their nests fail during early stages of incubation (Elliott- Smith & Haig, 2004;Thompson et al., 1997), and both species can maintain viable populations without annual breeding, breeding successfully, or achieving a certain reproductive rate at all sites or in arbitrarily defined river segments (Catlin et al., 2016;Lott, Wiley, Fischer, Hartfield, & Scott, 2013;McGowan et al., 2014). Piping plovers are known to successfully breed in one area, disperse long distances, and breed again within the same nesting season (Hunt et al., 2015). Birds occupying new or replenished habitats may experience reproductive success followed by declines in local populations and reproduction as habitat quality declines (Catlin et al., 2016;Cohen, Houghton, & Fraser, 2010). A more germane question about the terns and plovers that nested on the historical, and which continue to nest on the contemporary Platte River, is how those birds interacted, and interact, with other regional populations of their species' metapopulation. Successful nesting occurred, and until recently (late 20th century) still occurred, on in-channel habitats in the historical CPR and still occurs on in-channel habitats in the contemporary LPR. These habitats contributed to, and still do contribute, to the overall metapopulation of both species in the midcontinent of North America.

| HIS TORIC AL RECORD
The authors expressed doubts about the historical occurrence of least terns and piping plovers nesting on in-channel (sandbars) habitat of the Platte River and suggest human-created off-channel habitats were both species' primary nesting habitat which allowed them to "expand into and persist in a basin where hydrology is not ideally suited to their reproductive ecology (Farnsworth et al.,." To support their contentions, the authors refer only to 20th-century nesting on sandbars and human-created habitats along the CPR and off-channel nesting by least terns during 2 years at a single playa wetland in the Rainwater Basin of south-central Nebraska and along lake shorelines. A more rigorous review of the historical record shows that least terns and piping plovers were found along the Platte and other regional rivers since the earliest recorded ornithological observations. Lewis and Clark observed least terns and piping plovers along the Missouri River in 1803-1804, as did numerous others during the late 1800s and early 1900s (Catlin et al., 2010).
Least terns were observed at the Platte-Missouri River confluence in 1823 (Ducey, 2000). The earliest observation of piping plovers on the Platte River occurred on 8 July 1857 when members of the Warren Expedition collected five piping plover specimens and observed least terns at the confluence of the Loup and Platte rivers, a location 160 km upstream from the Platte-Missouri river confluence and between the two river sections considered by the authors (Ducey, 2000). Least terns were observed upstream of the historical CPR on the Platte River near the Colorado border in 1859 (Ducey, 2000).
In the first major review of Nebraska avifauna, Bruner, Wolcott, and Swenk (1904) concluded piping plovers were fairly common migrants that bred along the Platte, Loup, and Niobrara rivers and at lakes in the Sandhills of north-central Nebraska. Bruner et al. (1904) described the least tern as a common migrant and "not a rare breeder" in Nebraska, citing nesting records along the Missouri and Niobrara rivers and at a Rainwater Basin playa wetland in 1896 and 1897 (Tout, 1902). Both species have been widely observed breeding on the Platte and other Great Plains rivers, as well as other habitats, and historically, both species were widespread and numerous. Various authors (Currier, Lingle, & VanDerwalker, 1985;National Research Council, 2005;USFWS, 2006) have concluded the Platte and other Great Plains rivers were areas of regular breeding prior to major anthropogenic modifications of the rivers. Contemporary nesting by piping plovers and/or least tern populations on other Great Plains rivers, such as the Niobrara (Adolf, Higgins, Kruse, & Pavelka, 2001), which possess similar hydrographs, and which lack off-channel habitats, provides additional evidence contradicting the notion that adjacent off-channel habitats are a prerequisite for these species to colonize and breed within a river segment.

| MANAG EMENT AND P OLIC Y IMPLIC ATIONS
The authors state that a shift in the Platte River Recovery Implementation Program's (PRRIP) activities directed toward least tern and piping plover recovery away from in-channel habitat restoration to off-channel habitat maintenance represents a success of adaptive management that is "unique among riverine restoration programs" (Farnsworth et al.,page 10). We believe conclusions about threatened and endangered species management and recovery, as well as stewardship of natural resources, must be made considering the full spectrum of tradeoffs and consequences. Loss of habitat due to human alterations of natural systems is the principal reason regional populations of least terns and piping plovers declined, remain small compared to historical levels, and why they were listed under the Endangered Species Act and remain on the federal Endangered Species List (USFWS, 1988(USFWS, , 1990. It should be noted the least tern has been proposed for federal delisting based on a number of factors, including, but not limited to, conservation efforts and increasing populations in some areas (see USFWS, 2013). Industry (i.e., sand and gravel mining) in the Platte River basin has created sequences of short-lived patches of off-channel nesting habitat incidental to their business activities which have played a role in the population dynamics of these two species for many decades. Off-channel tracts of habitat along, but disconnected from, the Platte River require perpetual investments of capital and maintenance to provide adequate nesting areas for terns and plovers when they are no longer being used by industry; intensive management, including native predator exclusion and control (Keldsen & Baasch, 2016), are required to achieve and maintain reproduction by the two species in these areas.
On-channel habitats, such as those used by the birds on the historical CPR and contemporary LPR, existed or presently exist (LPR) only in resilient, dynamic river systems and are maintained by hydrological and geomorphic processes and benefit a diversity of species (Alexander et al., 2013;Currier et al., 1985). A decision to formally withdraw from river restoration and shift focus to maintaining relatively small and intensively managed tracts of off-channel habitat in the CPR disregards consequences beyond the scope of these two species and relegates the status of least terns and piping plovers in this region to species that are conservation reliant-imperiled species whose threats can only be managed rather than eliminated (Goble, Wiens, Scott, Male, & Hall, 2012;Scott, Goble, Haines, Wiens, & Neel, 2010). Decisions to render a species conservation reliant have been questioned (Goble et al., 2012;Scott et al., 2010) because, even though species recovery goals may be achieved, populations are only maintained through perpetual human intervention.
Dynamic, albeit altered, river systems such as the Platte River and others in the Great Plains, which presently maintain nesting habitat used by least terns and piping plovers, play an important role in the ongoing recovery of both species.

| CON CLUS IONS
We appreciate the authors' efforts toward modeling sandbar availability in relation to river hydrology; however, their analysis has shortcomings which limit the study's usefulness. These shortcomings, as well as incomplete characterizations of the species' ecology and the historical record, negate the author's assertions that least tern and piping plovers are not adapted to occupying and nesting on river sandbars on the Platte River system. Decisions relegating imperiled species to conservation reliant status need to be made only after considering the full range of tradeoffs and consequences.