Back to Laysan Duck Revised Recovery Plan Main Page

 

II. RECOVERY STRATEGY

 

A. PAST AND CURRENT CONSERVATION MEASURES

B. TRANSLOCATION: A PRIMARY RECOVERY TOOL

The greatest current threat to the Laysan duck is its small total population size and distribution limited to two locations on low-lying islands that are vulnerable to catastrophic events.  Ensuring the long-term viability of the Laysan duck depends upon (1) maintaining the source population and its habitat on Laysan Island, (2) maintaining the new population at Midway and improving habitat there, and (3) establishing the species on additional islands.  The immediate goal is to reduce the threats to the Laysan duck to the point that we can consider downlisting the species from endangered to threatened status.  The long-term goal is to recover the species; that is, to ensure that the threats to its persistence have been reduced so that it no longer requires protection under the Endangered Species Act and may be delisted.  This plan outlines the recovery actions that will reduce the risk of extinction for the Laysan duck by addressing the threats to the Laysan population, protecting and enhancing habitat quality, and reestablishing additional wild populations on other islands that are managed to ensure the long-term viability of those populations.

A. PAST AND CURRENT CONSERVATION MEASURES

A comprehensive restoration plan has been developed for Laysan Island (the Laysan Island Ecosystem Restoration Plan) that details the measures necessary to restore the ecosystem:  weed control; alien invertebrate identification and control; vegetation, invertebrate, and vertebrate monitoring; propagation and outplanting of native plants; plant and invertebrate restoration; pollen core studies; vertebrate restoration (including the Laysan duck); and snake-eyed skink eradication (Morin and Conant 1998).  Funding, time, and logistical constraints have prevented initiation of most of these projects, although some are underway.  This section presents those restoration projects and monitoring efforts designed specifically for the Laysan duck.  Recommendations for further recovery actions specifically geared to benefit the duck are presented in later sections of this recovery plan.

1. Laysan Duck Population Monitoring

The Laysan duck is a difficult species to monitor (Sincock and Kridler 1977).  The duck’s nocturnal and cryptic habits and seasonal differences in their use of the lake contribute to the difficulty of estimating the population size.  Line transect methods are unsatisfactory because of the negative impact on the ducks (e.g., flushing incubating females from nests, leaving eggs vulnerable to predators; Marshall 1992b) and the tendency to underestimate the population size (Sincock and Kridler 1977; Moulton and Weller 1984; Marshall 1992b).  Other negative effects of line transects include the inadvertent destruction of seabird burrows and disturbance of other ground-nesting birds.   

Lake counts on Laysan were used to generate an index of the population size in the last century and as recently as 1998.  Although not an effective method to estimate population size because use of the lake by Laysan ducks is seasonally and environmentally variable, these lake counts do provide an index of fluctuation in the population (Seavy et al.2009) and therefore are an efficient and useful tool for monitoring gross changes in numbers. .  Marshall (1992b) and others determined that the most accurate way to estimate the population size is by calculating ratios of marked to unmarked ducks at the lake at dusk.  This method requires that a portion of the population be marked.  Fall and winter yield the highest numbers of lakeside ducks for population estimates, although year-round monitoring is useful (Reynolds 2002).  Intensive banding was conducted most recently in 2004 and 2005, in order to follow broods and individual juveniles in preparation for translocation.  A percentage of the Laysan duck population on Laysan currently is marked with color bands.  Individuals have unique band combinations. 

The geographic isolation of the Laysan ducks on small islands (Laysan and Midway) makes the species well suited to a mark-resight method of population estimation (Lincoln-Peterson Index; see Table 2) because the population meets the “closed population” assumption of such a model.  There is no possibility of emigration or immigration, and during intensive monitoring and with high adult survivorship in this species, the mark-resight methods also meet the assumption of no births or deaths during the sampling period (Bibby et al. 1992).

Two monitoring methods are now used twice each month on Laysan Island to generate data that can be used in calculating Lincoln-Peterson estimates and measuring other population parameters: census walks and resighting surveys.  Field staff determine the ratio of marked to unmarked Laysan ducks during a 1-hour census walk around the lake before sunset (Marshall 1992b).  Birds are recorded as banded, unbanded, or unknown.  The numbers of broods and ducklings and the age class of ducklings are recorded.  Band reading is conducted for one to two hours before sunset.  Observers note the sex and band combination of each bird.  All ducklings and hens are identified, and the ducklings are assigned an age class. 

Along with the census walks, individual survival histories are used to determine the number of marked individuals in the population for Lincoln-Peterson estimates.  Resighting surveys provide data that can be used to determine population parameters such as survivorship, sex ratio, individual histories, brood production and breeder identification.

A protocol to monitor the status of the duck population at Midway is under development.  This project includes mark-resight methods to determine detection probabilities for comparison with all-wetland counts, and to calculate an initial population estimate. 

2.  Ecosystem Conservation and Monitoring

(a)  Weed control and vegetation monitoring

In 1991 we initiated a program to eradicate the nonnative grass Cenchrus echinatus on Laysan Island.  Full-time crews of one or more technicians have maintained these eradication efforts year-round.  C. echinatus is highly invasive, forming dense mats that crowd out the native bunchgrass Eragrostis variabilis, which is the preferred nesting habitat for the duck on Laysan.  Eradication efforts have been highly effective.  No C. echinatus has been found on Laysan since April 2002 (C. Rehkemper, USFWS, pers. comm. 2008).

Beginning in 1999, seeds of the endangered plant Mariscus pennatiformis ssp. bryanni were collected and propagated on Laysan.  Seeds and cuttings of another endangered plant, Chenopodium oahuense, also were gathered.  Seeds of the native palm Pritchardia remota were obtained from Nihoa Island and taken to Laysan for propagation, and work has begun on the propagation of the bunchgrass Lepturus repens (Depkin and Lund 2001).  Current native plant propagation efforts on Laysan include the following species: Capparis sandwichiana, Chenopodium oahuense, Lepidium bidentatum var. o-waihiense, Lepturus repens, Mariscus pennatiformis ssp. bryanni, Pritchardia remota, Santalum ellipticum, and Solanum nelsonii.

At Midway Atoll, extensive habitat restoration was undertaken in preparation for Laysan duck reintroduction, including propagation and outplanting of native species such as Eragrostis variabilis, Scaevola sericea, and various sedge species. 

(b)  Invertebrate monitoring

Arthropod sampling and identification were conducted opportunistically in 1999 and 2000 by Nishida (1999, 2000).  Continued incursion of alien arthropods was documented.

In 2007, extensive invertebrate sampling was conducted on Laysan as part of efforts to characterize habitat for future translocation of the endangered Nihoa millerbird (Acrocephalus familiaris) to the island (MacDonald 2008).  These collections, when analyzed, may provide additional information about the available prey base for Laysan ducks.

(c)  Ant control experiment

A pilot project to remove introduced ants from Spit Island, Midway Atoll, was conducted in 2001 and 2002.  Fire ants (Solenopsis geminata) were thought to be eliminated but began to reappear 1 year after the pesticide was applied (C. Swenson, U.S. Fish and Wildlife Service, pers. comm. 2002).  As Midway is a reintroduction site for the Laysan duck, the successful eradication of fire ants would be beneficial to the success of that program.  Methods to eradicate ants from other islands would improve opportunities for ecosystem restoration, which would also benefit Laysan ducks.

(d)  Mosquito control at Midway

Mosquito control on Sand Island at Midway has been underway since the fall of 2003 (J. Klavitter, Midway Atoll National Wildlife Refuge, pers. comm. 2009).  The primary breeding sites for the insects are the sewer and septic tanks of the waste water system.  Secondary sites include discarded metal and wood holding water in the junkyard near the Seaplane Hangar, abandoned buildings with leaky roofs, and the drinking water storage tanks.  In the spring and summer of 2003 the mosquito population was observed to be relatively high, as evidenced by several thousand Laysan albatross chicks with severe infections of avian poxvirus, which is spread by mosquitoes.  In the fall of 2003, shade cloth was used to create physical barriers to prevent mosquitoes from accessing the waste water and drinking water systems in an effort to prevent breeding.  In addition, the majority of items holding standing water in the junkyard and in abandoned buildings were removed or permanently drained.  After management actions were performed, mosquito numbers have decreased dramatically as has the incidence of avian pox in Laysan albatross.  From 2004 to the present, fewer than 100 albatrosses each year have been affected by pox. 

As a precaution to prevent mosquitoes from breeding in the wetlands created in preparation for Laysan ducks, mosquito fish (Gambusia sp.) were introduced to approximately one-third of the wetlands beginning in 2003, and "mosquito dunks" (Bacillus thuringienis israelensis) were used in the remaining wetlands.  This management was effective and was used until 2006.  After 2006, mosquito dunks were not needed; the population of Laysan Ducks rapidly increased and now appear to be effectively controlling mosquitoes in the wetlands.

(e)  Lake and brine fly sampling.

Every other week the salinity, water temperature, and water depth are measured in the lake at the permanent depth gauge along the east edge, as well as in two adjacent freshwater seeps(USFWS 2001).  Brine flies are monitored as an index of food abundance for the duck.  Fly abundance at the lake serves as a predictor of duck breeding. 

3.  Captive Populations

In the late 1950s, 33 ducks were removed from Laysan and transferred to captive breeding facilities around the world.  Offspring from those birds were used to found a colony at the former Pohakuloa Endangered Species Facility in Hawai`i, and seven wild Laysan ducks were later added to that flock in an effort to improve breeding.  This program was discontinued in 1989 because of costs and because at that time, prior to the discovery of subfossils throughout the Hawaiian Islands, little justification existed for releasing Laysan ducks on other islands.  Some of the birds were shipped to mainland facilities, and individuals older than 8 years were euthanized (Reynolds and Kozar 2000b). 

The birds produced by mainland zoos were deemed unsuitable as candidates for reintroduction to the wild because of the potential loss of adaptations for life in the wild (McPhee 2003) and loss of genetic diversity (Frankham 1995) as well as poor breeding records, the possibility of hybridization in captivity (Reynolds and Kozar 2000b), and the risk of introducing new pathogens and parasites to wild Laysan ducks.  Surveys of zoos and private collections in 1999 indicated that 211 Laysan ducks were held in 32 collections worldwide, all descended from fewer than 19 founding pairs (Reynolds and Kozar 2000b).  Initially, birds bred well in captivity, but over time breeding success has decreased, possibly as a result of inbreeding depression.  Average clutch size for captive broods declined from 7.3 eggs in 1984 (Marshall 1992; Moulton and Marshall 1996) to 4.9 in 1999 (Reynolds and Kozar 2000a).  Some captive populations may also suffer from genetic “pollution”; birds have been kept in mixed flocks, and Laysan ducks in three facilities are known to have hybridized with a koloa, a northern shoveler (Anas clypeata), a cinnamon teal (Anas cyanoptera), and a wood duck (Aix sponsa).  Only 15 percent of facilities surveyed kept pedigrees for their Laysan ducks (Reynolds and Kozar 2000a).  For additional discussion, see Hybridization and Introgression section below.

4.  Pearl and Hermes Reef Translocation

Aware of the threats facing the Laysan duck, 40 years ago biologists attempted to establish a new population on Pearl and Hermes Reef, approximately 440 kilometers (273 miles) northwest of Laysan (see Figure 2).  In March of 1967, five males and seven females were captured on Laysan Island and transported to Pearl and Hermes Reef for release.  The first two birds released flew directly out to sea and disappeared.  The remaining 10 ducks had their wings clipped to prevent flight until after the annual molt (Berger 1981).  An expedition in May discovered two dead Laysan ducks, cause of death unknown.  In July a female was found incubating a nest of six eggs, but the nest later failed.  Only two ducks were seen during a visit to the island in September of that year, and none were seen on successive trips (Sincock and Kridler 1977).  Inadequate monitoring of the released birds prevented identification of causes of mortality.  However, a combination of factors probably doomed the effort: the marginal habitat and lack of permanent sources of fresh water, small number of founding birds, and random factors.  No further translocations were attempted until the 2004 translocation to Midway.

5.  Midway Atoll Translocations

In the 2004 and 2005 breeding seasons, duck broods on Laysan were closely monitored and juvenile ducks selected as candidates for translocation to Midway.  These ducks were fitted with radio transmitters so that their condition could be tracked through the summer and fall, and to facilitate their capture in early October.  Twenty and 22 ducks were successfully moved from Laysan to Midway in 2004 and 2005, respectively.  Following the 2-day trip by sea, the ducks were placed in field aviaries to ensure their recovery to pre-capture body condition, acclimate them to their new home, and familiarize them with local food sources.  After several days, ducks were released two or three at a time into wetlands created for them.  In 2005, seven months after the first translocation, the first Laysan duck nests were found at Midway.  Successful breeding seasons in 2005, 2006, and 2007 led to rapid growth of the population.  As of 2007 Midway harbored an estimated 200 Laysan ducks (Reynolds et al. 2007a), and the 2008 breeding season produced a large number of fledged juveniles as well.  Unfortunately, a botulism outbreak in August of 2008 resulted in the loss of more than 150 ducks at Midway, and although at this writing it is generally agreed that the refuge harbors at least 200 ducks, we do not have a new estimate of population size in the atoll.  This event has necessitated a re-evaluation of wetland management at Midway to facilitate rapid response to future botulism outbreaks and minimize the mortality of Laysan ducks.

The translocation protocols were designed to address the issues discussed in the previous section.  Some analyses of reproductive, demographic, and other data collected during radio-tracking and other monitoring efforts at Midway have been provided in reports by the USGS Pacific Island Ecosystems Research Center (e.g., Reynolds et al. 2006a) and peer-reviewed publications (e.g., Reynolds et al. 2007a, 2008); more are forthcoming.

B. TRANSLOCATION: A PRIMARY RECOVERY TOOL

Translocation is the deliberate release of animals to the wild to establish, reestablish, or augment a population (Griffith et al. 1989).  It is used as a conservation tool to mitigate threats to a species by placing individuals at locations that are free of those threats, as a short-term or long-term means of increasing a species’ chance of survival, or as part of a program to restore a particular biotic community.  There is an urgent need to translocate Laysan ducks to additional islands and establish new populations, especially for the first two of these reasons.  The restoration of the Laysan duck as a component of the native ecosystems on these islands is also desirable.

1.  Justification for Translocation within Hawai`i

The discovery of Laysan duck bones on the Main Hawaiian Islands and our knowledge that the species previously inhabited Lisianski Island provide a sound biogeographic foundation for reintroducing the Laysan duck to additional islands throughout the archipelago (Olson and Ziegler 1995; Cooper et al. 1996).  Ecosystem restoration and the reestablishment of wild Laysan ducks on other islands are needed to reduce the risk of extinction.  Reintroduction of the Laysan duck also would represent the restoration of a missing component of the Hawaiian avifauna on these islands.  The restoration of Laysan ducks to additional islands will reduce the risk of extinction from events and processes that may affect the species in the two locations in the Northwestern Hawaiian Islands where it presently occurs and restore the species to ecosystems where it previously existed.

The Laysan duck has proven to be an excellent candidate for translocation.  The species is adapted to a harsh environment, flexible in its foraging and breeding behavior, large enough to carry radio transmitters with high battery capacity (to facilitate monitoring of released birds), and the flight feathers can be trimmed to prevent dispersal from the release site.  On a predator-free island, clipping flight feathers does not compromise the duck’s survival, foraging, or breeding, and the feathers are replaced with the next molt.  With adequate food, water, cover, and protection from mammalian predators, the Laysan duck breeds well in the wild.  The birds are unlikely to affect rare invertebrates at translocation sites because the ducks seem to select the most abundant prey available (Reynolds et al. 2006b). 

2.  Hybridization and Introgression

Hybridization is the interbreeding of individuals from genetically distinct populations, and introgression is gene flow between populations of individuals that hybridize (Rhymer and Simberloff 1996).  There is some concern that Laysan ducks might hybridize with koloa or mallards.  Hybridization and introgression with mallards has contributed to the decline of other duck species in New Zealand, Australia, and Hawai`i (notably the koloa; Rhymer and Simberloff 1996).  However, Laysan ducks are genetically distinct from mallards and koloa (Rhymer 2001), and they may have co-existed with koloa on the main islands in the past, factors that suggest Laysan ducks are less likely to hybridize in the wild (Reynolds and Kozar 2000a; Pratt and Pratt 2001), although they may hybridize in captivity.  As a precaution, however, mallards should be eliminated at translocation sites to prevent hybridization or competition of mallards with either of the native endangered duck species.  A multi-agency group is developing a comprehensive statewide approach to the feral mallard problem, and research is currently underway to develop reliable criteria for distinguishing between koloa, hybrids and female mallards.  A program to remove feral mallards and hybrids, including public education and outreach, will be proposed by the group once identification and removal methods have been refined and tested in the field.

3.  Source Population

The existing captive flocks of Laysan ducks are unsuitable for release into the wild for several reasons: (1) the pedigrees of these birds are unknown because studbooks have not been maintained; (2) careful breeding to maintain genetic diversity has not taken place; (3) these captive populations have become increasingly adapted to captivity over multiple generations (more than 40 years); (4) captive ducks on the mainland may be reservoirs for diseases to which Hawaiian birds have no immunity; and (5) captive Laysan ducks kept in mixed-species flocks have been documented to hybridize with other species (see Prospects for Reintroduction of Captive Birds, below).  Unless a new captive flock is created that is managed specifically for the purpose of establishing additional wild populations, only wild-source individuals should be used for translocation (Reynolds and Kozar 2000a).  Translocation success with wild-caught animals often is greatest when animals are removed from high density and increasing source populations (Griffith et al. 1989).  These conditions are rare for endangered species, but such conditions do occur periodically on Laysan Island.

Of primary concern to managers, then, is the population trend on Laysan and whether the population can withstand the removal of individuals to reestablish the species elsewhere in Hawai`i.  For the first translocations, to Midway, the best age class and the number of ducks to remove from the source population were explored with population simulations for several removal scenarios using the RAMAS AGE program (version 2.0; Reynolds and Kozar 2000a).  The program simulates age-structured population fluctuations and can be applied to predict population size and persistence.  Simulations incorporating translocation removals show that removal of up to 20 percent of juvenile birds for 5 years had the least significant impact on population projections.  Removal of breeding birds accelerated the time to extinction and caused a greater decline in the population than removal of juveniles.  Removal of adult females from Laysan, especially during periods of lower population density, could exacerbate decline in the source population by decreasing production.  Therefore, only juvenile ducks should be removed from Laysan Island, and they should be removed during periods of high density or population growth to avoid adverse effects to the source population. 

Duckling mortality on Laysan is often attributable to trauma, and is correlated with the density of adult females (Reynolds and Work 2005, Reynolds et al. 2007).  Thus, limited brood rearing habitat is suspected to increase mortality from overcrowding, potentially contributing to the density dependence that has been observed in population fluctuations (Seavy et al 2009).  Habitat enhancement and other management should be explored to increase duckling survival and to provide more juveniles for translocation.

4. Founding Population

Translocation of fledged juveniles from different broods is ideal to maximize the genetic representation of the species in the new population.  The genetic variability in Laysan ducks is presumed to be low, but currently we have no data to validate that assumption. 

The age and sex of the translocated birds are important variables in producing a self-sustaining population.  As mentioned above, fledged juveniles are the preferred candidates for translocation, based on population viability analysis and the behavior of birds during the 2004 and 2005 translocations to Midway.  Also, an equal or slightly male-biased ratio would be preferable, as a slight bias toward males promotes male-male competition and female choice, an important stimulant for breeding activity in many dabbling duck species (McKinney and Brewer 1989). 

Birds selected for translocation should be treated for echinuriasis and other diseases before removal to the transfer sites.  Echinuria uncinata is unknown in waterbirds in the Main Hawaiian Islands, and the risk to those species would be substantial if juvenile birds from Laysan harboring the parasite were transferred to the main islands either for release or for propagation of a captive flock (T. Work, pers. comm. 2002).  The anti-parasite medication ivermectin is known to eliminate nematodes in other waterfowl, and has been used successfully in other endangered duck species during translocation in New Zealand (Gummer 1999). 

When logistics permit, the ideal release procedure includes a period of acclimation in an on-site enclosure.  This type of release may restore loss of body condition during interisland transport, may encourage recognition of novel prey, and improve site fidelity (Kleiman 1989).  An aviary on or near the release site is ideal for temporarily housing translocated birds.  Laysan ducks are known to be aggressive towards one another, and separate pens may be necessary for some individuals.

On Laysan, reproduction is highly variable, and few or no ducklings are produced in some years, so planning for multi-year translocations may be required.  The target number of founders for Midway was 50 individuals; this number was thought to be a reasonable minimum to reduce the risk of inbreeding depression and improve the chances of establishing a healthy, growing population (USFWS 2004).   Of the original 42 founders, only about 25 of these, and fewer females than males, are known to have bred.  Nonetheless, the rate of population growth at Midway (see Reynolds et al. 2007a) suggests that this number was sufficient to establish a healthy, self-sustaining population.  Post-hoc analyses of translocations of other island species provide corroboration for this conclusion.  Study of a similar situation (small number of founders, strongly male-biased sex ratio) in a translocated population of New Zealand robins (Petroica australis) determined that supplementation with additional founders was not necessary (Armstrong and Ewen 2001).  The study of genetic variability and success of translocations in the South Island saddleback described above (Taylor and Jamieson 2008; see the Genetic Considerations section) suggests that, although more founding breeders is preferable to fewer, target numbers of individuals for founding new populations of wide ranging and genetically diverse (e.g., continental) species may not apply to island species that already have undergone multiple, severe bottlenecks.

Supplemental translocation may occasionally be required in reintroduction programs to ensure population persistence by increasing population growth, responding to a catastrophic decline, and/or maintaining or improving genetic variability.  However, the status of the nascent population and necessity of such additional translocations must be assessed before undertaking them (Armstrong and Ewen 2001).  It is important to consider the potential for translocations of additional founders not only to be unnecessary, but to result in the use of scarce conservation resources that would be better applied to other projects (Armstrong and Seddon 2008).  In the case of the Laysan duck, translocation is a costly undertaking that necessitates trade-offs on other fronts.  We acknowledge that declaring the Midway translocation a complete and unqualified success is premature (e.g., Seddon 1999); we don’t know what threats to this new population may crop up in the future.  For example, although we were aware of isolated cases of botulism in migratory birds at Midway, we could not have predicted an outbreak in Laysan ducks of the magnitude that occurred in August of 2008.  However, based on initial assessments of the translocation (Reynolds et al. 2006a, 2007a), the Laysan ducks at Midway do not appear to exhibit demographic indications (i.e., slow or no population growth owing for example to reduced hatchability or reduced duckling survival) that additional translocations are necessary now to ensure the population’s persistence (B. Bowen, University of Hawaii, pers. comm. 2008).

5.  Selecting and Evaluating Release Sites

For a translocation to be successful, the primary threats that led to the species’ initial decline or extirpation must be controlled.  Poor habitat quality is the most common reason for the failure of translocations (Griffith et al. 1989; Veitch 1995).  In the case of the Laysan duck, mammalian predators on the main islands need to be controlled at proposed translocation sites.  Sufficient food, water sources, vegetative cover, and breeding sites also must be available at the release location.  Each site must be carefully evaluated for the presence and quality of these resources, and appropriate restoration or enhancement, as well as predator removal or control, is a prerequisite for translocation.  Translocation plans for each proposed restoration island or site should be developed to suit the logistical feasibility of the site and the status and availability of source birds.

The presence of mammalian or other introduced predators will seriously jeopardize the success of any translocation effort (Armstrong and McLean 1995; Veitch 1995; Towns et al. 1997).  Only habitats where mammalian predators are absent or sufficiently controlled should be considered for translocation sites.  Possible methods for control of predators at translocation sites in the Main Hawaiian Islands include fences, toxicants, trapping, and shooting, or some combination of these.  Predator exclusion fences are under development but not yet in regular use in Hawai`i; research and trials are taking place, however, in Hawai`i, New Zealand, and elsewhere in the Pacific.  In addition, even the predator-free Northwestern Hawaiian Islands will require varying degrees of restoration in the form of pest and weed control, or freshwater seep creation or restoration.  Depending on the condition of the release site and the status of resources necessary to support Laysan ducks (fresh water, prey base, vegetative cover), translocation of ducks to a temporary aviary setting prior to release can occur simultaneously with some habitat restoration efforts.  Intensive management of Laysan ducks at translocation sites, such as the provision of supplemental food and water, may be required until habitat restoration efforts are complete.

Literature reviews and site visits to areas where Laysan ducks might be reintroduced were conducted in 1998 and 1999 (Reynolds and Kozar 2000a).  Biological characteristics and non-biological suitability features of these sites are summarized in Appendices 1 and 2.  Biological factors considered included habitat assessment, vegetation characteristics, invertebrate abundance, fresh water presence or absence, potential predators, and the need for restoration and/or predator control efforts.  Non-biological features included physical characteristics of the island, logistical feasibility (e.g., ease of post-release monitoring), and existing infrastructure or management.  Twelve Northwestern Hawaiian Islands and eight Main Hawaiian Islands were assessed in terms of their suitability for the reestablishment of the Laysan duck (Appendix 1).  Of the 20 islands considered, eight sites were judged to be promising potential translocation sites in the short- to medium-term: Midway Atoll National Wildlife Refuge, Lisianski Island and Nihoa Island (Hawaiian Islands National Wildlife Refuge), and Kure Atoll (State of Hawaii) in the Papahānaumokuākea Marine National Monument (Northwestern Hawaiian Islands), and the islands of Kaho`olawe and Kaua`i in the Main Hawaiian Islands (Appendix 2).  Descriptions of all eight sites and brief discussions of their biological and physical suitability and management needs are presented in the next section.

In 2003, 13 scientists and land managers participated in a structured ranking of these sites to determine an initial location where Laysan duck translocation would be most feasible, cost effective, and likely to succeed.  Logistical feasibility and cost are especially critical considerations in the remote Northwestern Hawaiian Islands, and these factors weighed as heavily as biological suitability in the site ranking. 

Through this process, Midway Atoll was identified as the best site for trial releases of Laysan ducks, and two translocations were carried out, as described in the previous section, “Past and Current Conservation Measures.”  Here we provide additional details about our evaluation of Midway as a translocation site, and of current concerns now that Laysan ducks have become established there.  This discussion provides an example of the process and types of considerations that would be undertaken for any release site.

Midway Atoll lies at 28° 12’ N, 177° 22’ W, approximately 1,840 kilometers (1,143 miles) northwest of Honolulu (Figure 1).  The atoll’s land area covers 625 hectares (1,544 acres) and is composed of two main islands, Sand Island (467 hectares [1,154 acres]) and Eastern Island (156 hectares [385 acres]), and a smaller islet, Spit Island (2 hectares [5 acres]).  Like Laysan, Midway Atoll is a National Wildlife Refuge managed by our agency and, with its surrounding waters, is included in the Papahānaumokuākea Marine National Monument, co-managed by our agency, the National Oceanic and Atmospheric Administration, and the State of Hawai`i.  Midway is staffed by permanent USFWS personnel and can support chartered air service from Honolulu.  Rehabilitation of habitat at Midway and close monitoring of translocated ducks was therefore more logistically feasible than it would be on an uninhabited island.

In 1998, 1999, and 2001, biologists traveled to Midway to evaluate the atoll as a potential release site for Laysan ducks, and specifically to evaluate the creation of wetland habitat to provide the fresh water essential to supporting a self-sustaining population of Laysan ducks (Reynolds and Kozar 2000a).  The water table is less than 2 meters (6.6 feet) below the land surface in some parts of the atoll, providing suitable conditions for the creation of additional wetlands.  Since 2002, several small wetlands have been created on Sand and Eastern Islands.  Two of these wetlands on Sand Island and two on Eastern served as release sites for translocated ducks.

Midway has experienced many introductions of highly invasive nonnative plant species over the years, including Verbesina encelioides and Cenchrus echinatus.  The invertebrate fauna on Midway Atoll is dominated by exotics.  Vegetation restoration is a high priority and is ongoing in selected parts of the atoll, but if broad-scale herbicides, pesticides, and heavy equipment are used, Laysan ducks could be negatively affected; close coordination between Refuges and Ecological Services programs of USFWS will be necessary to ensure that conflicts between habitat restoration and Laysan duck recovery at Midway are resolved efficiently.

Many introduced invertebrates likely are prey items for the Laysan duck (Reynolds and Kozar 2000a).  However, introduced predatory arthropods such as fire ants and big-headed ants may need to be controlled until techniques for ant eradication are developed (or the impacts of ants on the duck’s prey base and nests are judged to be insignificant).  Fire ants were discovered at Midway Atoll in 2000.  Results of a pilot project to eradicate ants from Spit Island using bait treated with the toxicant Maxforce (hydramethylon) indicate that fire ants can be controlled by periodic (possibly annual) applications of granular ant toxicants (C. Swenson, U.S. Fish and Wildlife Service, pers. comm. 2001).

Although rats have been eradicated from Midway, mice (Mus musculus) have not, and their abundance (they occur on Sand Island only) has increased markedly since rats were removed in the mid 1990s.  Mice currently don’t appear to have any impact on Laysan ducks or Midway’s seabirds, but they may become a problem for birds in the future.  On Gough Island in the South Atlantic Ocean, for example, mice have developed the capacity to prey on and kill albatross chicks, and have a significant effect on their reproductive success (Wanless et al. 2007).  Sand Island should be monitored for evidence of adverse impacts of mice on Laysan ducks: direct impacts such as predation and/or indirect effects such as limitation of food resources. 

6.  Other Prospective Translocation Sites

(a)  Northwestern Hawaiian Islands.

(i)  Lisianski Island.  Lisianski Island is Laysan’s nearest neighbor in the northwest Hawaiian chain, and is known to have previously supported Laysan ducks.  Loss of plant cover in the mid-1800s resulted in shifting sands that filled the island’s freshwater source.  Since Laysan ducks occurred previously on Lisianski, we know that with adequate management the island can support the species, thus Lisianski is a potential translocation site.  In the event that Lisianski is chosen as a translocation site for Laysan ducks, wetland habitat must be restored to provide a source of fresh water, and development of a Lisianski ecosystem restoration plan is recommended.  The restoration of the wetland on Lisianski would pose logistical challenges, as the remote location of the island would preclude use of the heavy construction equipment that would normally be used for such an operation. A brief pilot study took place in 2006 to evaluate the hydrology of Lisianski and the feasibility of hand excavation to restore small fresh water seeps (Meyer 2006).  Only one of 11 test pits (maximum depth 7 feet [2 meters]) excavated in this initial effort struck groundwater, and this was a very thin layer; the conclusion of the pilot study was that hand-excavation to create perennial seeps on Lisianski is not feasible (Meyer 2006).  Using heavy equipment to excavate a wetland on Lisianski may be possible, but presents significant logistical and regulatory hurdles to overcome.  Other options, such as catchment ponds or pumping groundwater from a lens that may exist at greater depth, have yet to be investigated.

(ii)  Nihoa Island.  Nihoa Island also is considered a potential translocation site after experimental translocations are made to other islands.  At 68 hectares (168 acres), the island is large enough to support small numbers of Laysan ducks.  Native plants and arthropods are abundant.  Freshwater seeps occur naturally on Nihoa, eliminating the need to develop water sources or conduct other restoration for Laysan ducks.  Nihoa is considered the most pristine of the Northwestern Hawaiian Islands, and an assessment of the potential impacts of Laysan ducks on the island’s terrestrial biota should be conducted prior to translocation (Reynolds and Kozar 2000a).  For example, the endemic cone-headed katydid Banza nihoa already may be negatively affected by the introduced grasshopper Schistocerca nitens and perhaps by several ant species (E. Flint, pers. comm. 2006).  Laysan ducks are likely to eat the most abundant palatable prey available at any translocation site and so may not affect Banza nihoa, but the risk to the katydid of predation by Laysan ducks should still be considered.  We suspect, however, that human impacts to Nihoa (associated with a translocation effort) are the primary risk.  Technology for remote or automated post-release monitoring to eliminate the need for human presence on Nihoa should be explored.

(iii)  Kure Atoll.  Kure Atoll consists of two separate islets comprising 100 hectares (247 acres) of land area.  Kure Atoll once supported a U.S. Coast Guard LORAN (long range navigation) station, but little of the infrastructure remains.  The atoll is managed by the State of Hawai`i, which eliminated rats on the islets in 1994.  The islands support a large number of arthropods and have a moderate amount of nesting cover for ducks (Reynolds and Kozar 2000a).  Further restoration work is necessary before the atoll would be a suitable translocation site for Laysan ducks.  In 2005, laboratory tests found groundwater from Kure to be free of contaminants, and in 2006 the State’s field crew created a small freshwater seep and outplanted native wetland plants in and around it (C. Vanderlip, Hawaii Division of Forestry and Wildlife, pers. comm. 2009).  Additional sources of fresh water, perhaps from rainwater catchments or additional excavated wetlands such as those at Midway, could be created to sustain Laysan ducks there.  The State currently plans to investigate the ecological impacts of big-headed ants (Pheidole megacephala) at Kure and is seeking support for a comprehensive weed control program to eliminate or control the spread of Verbesina encelioides in the atoll.  Because alien species control necessarily involves extensive physical disturbance and the use of pesticides, the majority of such work should take place before Laysan ducks are released in the atoll (C.Vanderlip, pers. comm. 2009). 

(b)  Main Hawaiian Islands.

(i)  Kaho`olawe.  Kaho`olawe has great potential as a translocation site for Laysan ducks in the Main Hawaiian Islands.  Translocation of Laysan ducks to Kaho`olawe already has been recommended by the Kaho`olawe Island Restoration Commission (Social Science Research Institute 1998).  A former U.S. Navy bombing range, the island was transferred to the State of Hawai`i in 1994, and for the next 10 years the Navy worked to remove as much remaining live ordnance as possible and meet the State’s objectives for preservation of archeological sites and environmental restoration on Kaho`olawe.  The Navy’s work on Kaho`olawe was completed in 2004.  Goats were removed, and the planned ordnance removal was completed.  Restoration of native vegetation is ongoing on the island, which now supports moderate nesting cover and a wide variety of arthropods.  Ephemeral wetlands exist on Kaho`olawe but need significant enhancement to support Laysan ducks.  Rats have not been seen on the island since 1971, but their bones have been found in owl pellets since then (Snetsinger 1994).  The presence of rats may be equivocal, but the presence of cats is certain.  If cats and other mammalian predators are removed, Kaho`olawe will have excellent potential as a release site for Laysan ducks (Reynolds and Kozar 2000a).

(ii)  Kaua`i.  Of the other Main Hawaiian Islands, Kaua!i may be the best choice for reintroduction of the Laysan duck because it is the only island that may still be free of the Indian mongoose, a predator that would pose a major threat to Laysan ducks.  Other significant predators, including rats, cats, and dogs, occur on the island and would have to be controlled prior to a release of ducks (and probably in perpetuity).  Multiple sites on Kaua`i could be suitable for Laysan duck release (see Appendix 1), including two existing National Wildlife Refuges.  These sites have extensive areas of suitable habitat and nesting cover and abundant sources of food and fresh water. 

(iii)  Other Main Hawaiian Islands.

Ni`ihau, O`ahu, Maui, Moloka`i, Lāna`i, and Hawai`i all have sites that potentially could support Laysan ducks.  Managed wetlands occur on O`ahu, Maui, and Hawai`i, and of the five islands listed above, these three may provide the best opportunities for establishing self-sustaining Laysan duck populations.  All of these islands, however, have significant problems with introduced mammalian predators which would have to be addressed through either control efforts or exclosures before they could be considered as suitable translocation sites for Laysan ducks.

7.  Prospects for Reintroduction of Captive Birds

(a)  General Issues. 

The original Laysan duck recovery plan recommended maintaining captive flocks bred to ensure pure strains for eventual reintroduction to the wild (USFWS 1982).  Unfortunately, this plan was never realized.  Hybridization, incomplete population statistics, and harmful genetic change in captivity make the existing captive ducks and their future offspring poor candidates for reintroduction (Reynolds and Kozar 2000b).  Genetic change in a captive environment can decrease reintroduction success in two ways:  1) genetic variation may be lost through limited breeding opportunities, and 2) animals may become adapted to the captive environment (Frankham 1995; McPhee 2003).  In zoos, natural selection pressure on many features required for survival in nature, such as hunting and foraging abilities, is relaxed.  Over long periods in captivity, natural selection acts to maximize fitness in a captive environment, thus the individuals surviving and breeding are those pre-adapted to captive conditions (Frankham 2008).  A review of translocation efforts for various animal species from 1973 to 1986 found a vastly different success rate between wild-caught (75 percent) and captive-reared (38 percent) individuals (Griffith et al. 1989).

Captive breeders can minimize genetic adaptations to captivity by specifically managing captive flocks for reintroduction to the wild.  Techniques to minimize genetic changes include reducing time spent in captivity, regularly introducing wild genes, using only the offspring of wild birds for release, and releasing birds into wild or semi-wild habitat temporarily, until suitable habitat within their previous range can be restored (Frankham 1994; Reynolds and Kozar 2000b).

Disease is an additional risk in translocating captive-reared birds, especially birds from mainland facilities.  Confinement and mixing with other birds often increases the likelihood of disease transmission in captive flocks (Friend and Thomas 1990).  The Avian Disease Working Group, an association of captive breeders and veterinarians, rejected the idea of reintroducing any captive mainland birds to Hawai`i based on logistical, fiscal, and quarantine restraints as well as the risk of disease introduction (USFWS 1994). 

(b)  Northwestern Hawaiian Islands. 

We believe it would be feasible to use captive-bred birds for introduction to the Northwestern Hawaiian Islands if these birds came from a new captive flock specifically managed for such releases, but considering the urgency of establishing another wild population, using wild, parent-raised fledglings from Laysan for reestablishment on other islands in the northwestern chain is more expedient, easier logistically, and perhaps more successful.  A captive breeding program would take years to produce suitable numbers of offspring for release.  Disease risks on the main islands are higher, and these risks may be minimized if translocations of birds to islands in the northwestern chain are of individuals from other northwestern islands.

(c)  Main Hawaiian Islands. 

The single remaining natural population of Laysan ducks, on Laysan Island, is likely to remain the best source for founders of new populations, because in the timeframe for recovery of the species, Laysan will always harbor the greatest reservoir of the species’ diversity; other populations will always be a subset of that reservoir.  Translocation of wild birds from Laysan for the establishment of wild populations in the Main Hawaiian Islands may be feasible, but this possibility is limited by both the logistics and the “critical mass” needed for the establishment of large self-sustaining populations without depleting the source.  If multiple populations are established in the Northwestern Hawaiian Islands and they reach carrying capacity, subsequent removal of hatch-year birds from more accessible Northwestern Hawaiian Islands (such as Midway) for Main Hawaiian Island populations may be a feasible option.

A captive breeding facility, managed for establishing additional wild flocks of Laysan ducks, is another possible strategy for establishing populations in the Main Hawaiian Islands.  Eggs taken from Laysan Island may be the best way to found the captive flock because eggs are easier to transport than live birds, and egg removal would have the least impact on the population dynamics of the Laysan birds.  First-generation (F1) offspring from those eggs would be released to found the new wild flocks on the Main Hawaiian Islands when other restoration requirements have been met at potential release sites.