*Current address: PRBO Conservation Science, 3820 Cypress Drive #11, Petaluma, CA 94954, USA.
Translocation and early post-release demography of endangered Laysan teal
Version of Record online: 27 MAR 2008
© 2008 The Zoological Society of London No claim to original US government works
Volume 11, Issue 2, pages 160–168, April 2008
How to Cite
Reynolds, M. H., Seavy, N. E., Vekasy, M.S., Klavitter, J. L. and Laniawe, L. P. (2008), Translocation and early post-release demography of endangered Laysan teal. Animal Conservation, 11: 160–168. doi: 10.1111/j.1469-1795.2008.00166.x
Re-use of this article is permitted in accordance with the Creative Commons Deed, attribution 2.5, which does not permit commercial exploitation.
- Issue online: 27 MAR 2008
- Version of Record online: 27 MAR 2008
- Received 15 October 2007; accepted 1 February 2008
- Laysan duck;
- Anas laysanensis;
- matrix model;
- asymptotic growth rate;
- wild reintroduction;
- effective breeding population
In an attempt to reduce the high extinction risk inherent to small island populations, we translocated wild Laysan teal Anas laysanensis to a portion of its presumed prehistoric range. Most avian translocations lack the strategic post-release monitoring needed to assess early population establishment or failure. Therefore, we monitored the survival and reproduction of all founders, and their first-generation offspring using radio telemetry for 2 years after the first release. Forty-two Laysan teal were sourced directly from the only extant population on Laysan Island and transported 2 days by ship to Midway Atoll. All birds survived the translocation with nutritional and veterinary support, and spent between 4 and 14 days in captivity. Post-release survival of 42 founders was 0.857 (95% CI 0.86–0.99) during 2004–2006 or annualized 0.92 (95% CI 0.83–0.98). Seventeen of 18 founding hens attempted nesting in the first two breeding seasons. Fledgling success was 0.57 (95% CI 0.55–0.60) in 2005 and 0.63 (95% CI 0.62–0.64) in 2006. The effective founding female population (Ne) was 13. We applied these initial demographic rates to model population growth. The nascent population size increased to >100 after only 2 years post-release (λ=1.73). If this growth rate continues, the size of the Midway population could surpass the source population before 2010.
The avifauna of many island ecosystems has been dramatically altered by the introduction of non-native species (Cuddihy & Stone, 1990; Blackburn et al., 2004; Steadman, 2006). This is especially true of the Hawaiian Islands, where, in the last 200 years, more than 39 species of birds have become extinct, and many others exist on only a subset of the islands they once inhabited (Olson & James, 1982). Today, there is an increasing understanding of the threat posed by non-native species and how these threats can be reduced through management and eradication (Cote & Sutherland, 1997; Flint & Rehkemper, 2002; Donlan et al., 2007). However, even after threats have been eliminated, the ability of rare species to recolonize restored areas may be limited by constraints on dispersal. In such cases, translocations of endangered species to parts of their original range can be an effective restoration strategy (Griffith et al., 1989; Saunders, 1994; Hutton, Parkes & Sinclair, 2007).
The success of translocations depends on the care and health status of the animals, the suitability of release sites (Lindenmayer, 1994) and the ability of the translocated animals to survive and reproduce. The World Conservation Union (IUCN, 1998) guidelines stress the need for post-release monitoring of all, or a sample of, translocated individuals. Early post-release monitoring can be used to assess the different phases of the re-establishment process and identify the possible reasons for failure in a timely manner, thereby allowing for adjustments to translocation procedures, or habitat management that can increase the odds of successful reintroduction (Armstrong, Castro & Griffiths, 2007; Seddon, Armstrong & Maloney, 2007).
The Laysan teal Anas laysanensis, also known as the Laysan duck (AOU, 1998), provides an example of the conservation challenges facing many island species and the utility of translocation strategies. The Laysan teal is a small Hawaiian endemic dabbling duck that feeds primarily on invertebrates and seeds (Reynolds, Slotterback & Walters, 2006) and is characterized as a relatively long-lived species (maximum known wild lifespan of 12 years) with a low reproductive rate (Moulton & Weller, 1984). Sub-fossil evidence indicates that the Laysan teal was once widespread on the Hawaiian archipelago, but that its range contracted about 1000 BP during the period of colonization by humans and rats (Olson & James, 1991; Cooper et al., 1996; Burney et al., 2001). In the 20th century, Laysan teal were restricted to a single population on Laysan Island, a remote atoll in the Northwestern Hawaiian Islands (Moulton & Marshall, 1996). The Laysan teal is listed as endangered by the US Fish and Wildlife Service (USFWS) and critically endangered by the IUCN (Butchart & Hughes, 2003). Given that there are no records of Laysan teal dispersing from Laysan Island and currently none in captivity suitable for release (Reynolds & Kozar, 2000), re-establishing the species within its previous range required translocation of wild birds.
Because of its extremely restricted geographic range (4.1 km2) and small population size (fluctuating between 300 and 600 individuals this decade), the Laysan teal has long been recognized as being vulnerable to extinction from chance events. In 1967, Laysan teal were translocated to Pearl and Hermes Reef (440 km northwest of Laysan Island) but population establishment failed within 1 year (Berger, 1981). In the early 1990s, the only Laysan teal population underwent a dramatic decline, further emphasizing the vulnerability of the species (Work, Meteyer & Cole, 2004).
In 2004, we initiated a second reintroduction by translocating Laysan teal to a mammalian predator-free atoll, Midway Atoll (USGS, 2005). Given the absolute lack of data on the ecology of the Laysan teal on islands other than Laysan, detailed post-release demographic, ecological and behavioral studies are valuable to assess whether the translocation site can support the species' life-history requirements. We conducted intensive monitoring for 2 years post-release to: (1) determine the fates of all released birds; (2) apply basic ecological information gained from radio tracking Laysan teal to adaptive management; (3) determine early demographic parameters to project population establishment or failure. We report estimates of age-specific survival rates, stage-specific breeding productivity and recruitment for modelling the likelihood of population persistence at Midway Atoll and describe detailed methods to add to the knowledge base of what is effective in successful avian translocations.
Laysan Island (hereafter Laysan) is a low (12 m maximum a.s.l.) 415 ha island 1463 km northwest of Honolulu (25°46′N, 171°44′W) accessible only by boat, typically a 5-day voyage (Fig. 1). Laysan is unique among the Northwestern Hawaiian Islands in having a large shallow hyper-saline interior lake. Laysan is dominated by native vegetation and, despite degradation due to human activities in the early 1900s (Ely & Clapp, 1973), now has one of the most intact ecosystems of the archipelago. Midway Atoll (hereafter Midway) comprises Sand (467 ha), Eastern (156 ha) and Spit islands located 1930 km from Honolulu (28°12′N and 177°22′W; Fig. 1). Midway has had relatively continuous human presence since 1871 and saw significant military activity from 1939 to 1996. As a result, Midway is a highly altered ecosystem dominated by non-native vegetation such as Casuarina equisetifolia forest, mixed exotic grass, forb, and shrub associations, Verbesina encelioides and a few native plants (Apfelbaum, Ludwig & Ludwig, 1983; Freifeld, 1993). Rats Rattus rattus arrived at Midway with cargo in 1943, but were exterminated by 1998 (Rauzon, 2001). Today, Laysan and Midway are part of the Papahānaumokuākea Marine National Monument, National Wildlife Refuges, home to millions of Pacific seabirds, and managed by the USFWS.
Midway is within the presumed pre-historic range of the Laysan teal. Although sub-fossil evidence or direct observations of Laysan teal on Midway are lacking, the species could well have gone undetected. Midway's paleoecological remains were destroyed when much of the island was scraped to within a few centimeters of the water table and refilled with dredge material and off-island soil (Jimenez & Rosendahl, 1994). Futhermore, a Western naturalist did not describe the atoll until 1901, many years after habitat alteration and introduction of livestock had already occurred, and at least 40 years after the ducks were already extirpated from Lisianski Island (26°04′N 173°58′W; Fig. 1; Olson & Ziegler, 1995), the last refuge of the ducks other than Laysan. To enhance habitat for re-introduction of Laysan teal at release sites, we created or enhanced fresh water seeps with native sedges Cyperus laevigatus and Cyperus polytachos, and planted native bunch grass Eragrostis variabilis, an important nesting habitat on Laysan (Moulton & Marshall, 1996; Reynolds, Crampton & Vekasy, 2007;Fig. 1).
We use bird banding terminology to describe age classes: hatch-year (HY) is from fledging and independence (≥65–70 days) until January 1, after which it becomes a second-year (SY). In the third calendar year, the bird becomes an after-second-year (ASY). Adult birds with uncertain ages are also called after-hatch-year (AHY). For the population matrix model, we use ‘post-fledging juvenile’ to refer to HY and SY sub-adult birds and ‘adult’ to refer to potentially breeding >1 year old. We refer to birds moved from Laysan to Midway as ‘founders’ and identify the year of translocation (2004 or 2005) as the ‘cohort.’ We refer to the first-generation young that hatched and fledged on Midway (offspring of the 2004 founders) as the 2005 F1 generation. Young that fledged in 2006 (offspring of the 2004 and 2005 founder cohorts) are the F1 2006 generation. The offspring of the F1 2005 generation are the F2 2006 generation.
Source population monitoring and translocation
On Laysan, we monitored the population size and trend (Reynolds & Citta, 2007; Camp et al., in press) and the breeding success of the source population from March to September to target non-sibling juveniles as candidates for translocation in 2004–2005 (see also Reynolds et al., 2007). Thirty sub-adult (HY and SY) ducks were fitted with radio transmitters on Laysan to facilitate relocation and capture during 2004 and 2005. These young birds were independent, could fly and had not yet formed obvious pair-bonds or attempted to breed on Laysan at the time of translocation.
Two teams captured and processed candidate Laysan teal for translocation. We captured birds between 21:00 and 06:00 h local time on 1–2 October 2004, and 2–3 October 2005. The capture team included one to two radio trackers, two bird catchers using headlamps and hand nets, a small-boat paddler and two to three ‘runners’ to transport the birds within cotton bird bags, from the capture site at the lake, 2–3 km to the processing team at the camp. The processing team included data recorders, a veterinarian and biologists to measure and handle birds. The processing team conducted a physical exam of the birds, and identified the sex, age and brood identity for each bird. In 2004, we used body condition indices (Harder & Kirkpatrick, 1996) and morphometric sexing to select appropriate translocation birds. After morphometric sexing of juveniles proved to be unreliable in 2004, we used vent morphology to sex all HY candidates in 2005.
All birds underwent a complete external physical examine. We collected baseline pathology information, and birds were administered subcutaneous fluids (50 mL kg−1 lactated ringers), nutritional support (∼10 mg kg−1 of oral administered, Ensure, Abbott (TM) Abbott Laboratories, Abbott Park, IL, USA) and parasite treatments (200 μg kg−1 of ivermectin) for the nematode Echinuria uncinata, a significant cause of mortality in Laysan teal (Work et al., 2004). Birds were placed in individual transport boxes with food and water. After sunrise, transport boxes were loaded on small boats to board the ship. Once onboard, biologists replenished food and water for birds every 6 h and cleaned the cages. Birds were weighed with 500 g Pesola™ spring scales and given subcutaneous fluids and oral (Pesola (TM), Pesola Prazisionswaagen Ag, Rebmattli, Switzerland) every 12 h.
At Midway, we transported ducks to aviaries on Sand or Eastern Island. Laysan teal were held for 3–13 days in groups of two per aviary, and were provided cover and bathing pools. In the aviaries, ducks were offered locally collected live invertebrates and seeds and commercial mash (Nutrena Premium Poultry™, Cargill, Inc., West Wayzota, MN, USA) with meal worms Tenebrio molitor. We trimmed primary feathers to prevent long-range dispersal from Midway after release, and freed birds in groups of four at four pre-determined release sites (Fig. 1). Because trimmed flight feathers reduced the foraging range of the ducks, we offered supplemental food at the release sites as a precautionary measure. Initially, approximately eight cups of poultry mash were offered five times per week, and then, we tapered this regime gradually to no supplemental food after 3 months post-release.
All birds were given unique color band combinations and backpack harness (Sirtrack®, Havelock North, New Zealand)-mounted radio transmitters. Birds <400 g carried a 6–9 g transmitter with a shorter battery life. Larger birds carried an 11–12 g transmitter with a longer battery life (ATS and AVM custom models) and a duty cycle to turn the signal off 12–14 h day−1. All transmitters were <3% of the bird's body mass. Bird locations were determined one to six times per week by homing with hand-held three-element Yagi antennas and Telonics® (Telonics, Inc., Mesa, AZ, USA) receivers. All birds required multiple recaptures to replace radio transmitters by radio warranty dates, or after expiration. Color band resights were made at the release sites and select wetlands 1–2 h before sunset with a × 40–60 spotting scope.
In both years of the post-release monitoring, because all females were fitted with radios transmitters, we were able to find, monitor and record the outcome for nearly all nesting attempts. Nest attendance was determined by checking telemetry signals two to six times per week. To prevent disturbance or abandonment of nesting hens, nest contents were checked only if telemetry data suggested an incubation break or early failure, or after the normal 29-day incubation period. All ducklings were color banded at 20–60 days. A subset of juveniles were fitted with radio tags when they were 60–90 days of age. All carcasses suitable for necropsy were sent to the USGS National Wildlife Health Lab (Honolulu Field Station) for complete examination to determine the cause of death. In other cases, suspected causes of death were obtained from field signs.
All adult females and most males (n=42) were radio tagged during most of the 2-year study period, all fledglings in 2005 were also radio tagged and all 2006 fledglings were given unique color band combinations. Reproductive success was recorded for all females in 2005 and 2006. Because we monitored all nesting attempts, we report the average number of young fledged (f) for each cohort. We also report information on clutch size, the proportion of females attempting to nest, the proportion of females that nested successfully (one or more eggs hatch) and the proportion of ducklings that survived to independence/post-fledgling (c. >65 days)
Survival (S) of the 2004 and 2005 founder cohorts and the 2005 F1 generation (radio tagged) was estimated using the Kaplan–Meier method (Kaplan & Meier, 1958; Cox & Oakes, 1984). For the 2006 F1 and F2 post-fledglings, <25% of HY were given temporary radio-tags; thus, we estimated first-year survival using the ratio of known live birds (based on recaptures, resights, radio tracking and carcass recoveries) to the total number banded during the previous year (Johnson, 1996), with 95% binomial confidence limits (Zar, 1999). For 2006 post-fledglings without radio tags, not seen, recaptured or collected as carcasses, we assumed mortality if there were no sightings of an individual 1 year after it was marked. We report apparent nest and fledgling success (the number of ducklings surviving until independence and flight).
We used a subset of demographic data collected on Midway from 2004 to 2006 to construct a two-stage matrix model for the female population (McDonald & Caswell, 1993; Caswell, 2001). The model represented population growth as
where ni is a vector with the number of individuals in each stage at time i and A is the population projection matrix. We used a projection interval from i to i+1 of 1 year and assumed a post-breeding census. The projection matrix A had two stages: SY [birds in their first (potential breeding) year] and ASY [birds in at least their second (potential breeding) year]. The projection matrix was
where the matrix elements are vital rates (fecundities and survival) calculated using the survival (S) and female fecundity (f) for adult (a) and juvenile (j) birds. Adult survival was based on fates of the founding individuals that were radio-tracked over the 2-year post-release period. For the juvenile stages, we used the survival estimate for the 2006 F1 and F2 generation. For adult fecundity, we used the average number of young fledged in 2006 by the six 2004 ASY founder hens. This subset of founders had been on the island for more than a year and were all capable of flight. We expected that their reproductive success would be less impacted by the translocation stress, feather trimming, and supplemental feeding that occurred at release. For SY fecundity, we used the average number of young fledged in 2006 by the first five F1 females hatched on Midway in 2005. Like the subset of ASY founders, we chose this subset because we felt that it would not be impacted by translocation handling, and provided the best approximation for the future reproductive success of SY females on Midway. Because we used a post-breeding census, we multiplied fecundity by stage-specific survival. In the absence of 2006 juvenile sex ratios, we assumed a 50:50 ratio and multiplied the total fecundity values by 0.5 to limit it to female production.
To investigate the relative importance of each stage to the population growth rate, we calculated the elasticities of the matrix elements (de Kroon et al., 1986; Caswell, 2001). Elasticity is the proportional change in the population growth rate (λ) with respect to proportional changes in the matrix elements. We also used the population matrix to project the population size over the next 4 years. For the initial population size, we used the adult female population size at the end of the 2006 and then, assuming a 50:50 sex ratio of juveniles, multiplied by two to calculate the total population size. We compared this hypothetical population projection with two benchmarks: (1) the historic size of the Laysan Island population and (2) the projected population if Midway supported the same density as Laysan (assuming total island areas were suitable habitat). All matrix calculations were conducted using the function from the ‘popbio’ package (Stubben & Milligan, 2007) in R (R Development Core Team, 2005).
Means are presented±standard deviation, except where noted otherwise.
Before initiating the translocations, we documented a successful breeding season on Laysan (Reynolds et al., 2007). Twenty and 22 translocation candidates were captured between 21:00 and 06:00 h on Laysan on 1–2 October 2004 and 2–3 October 2005, respectively, and transported by ship for 2 days. We primarily targeted HY (independent juveniles) for translocation from 57 ducklings of 20 broods fledged on Laysan in 2004, and 56 ducklings of 20 broods fledged in 2005. Not all radiotagged candidates were captured, and, thus, some individuals of unknown identification were also captured and translocated (n=7). The transit time to Midway was c. 40 h in 2004 and 50 h in 2005. The translocated birds spent a mean of 9.4±4.0 days in captivity, including 42.5–60.4 h in a transport box and 2.0–13.7 days in holding aviaries on Midway's Sand and Eastern Islands.
Despite oral and subcutaneous nutritional support, during transit, both adult and juvenile birds lost significant body mass (paired t2004=7.03; d.f.=19, P<0.005; paired t2005=9.2; d.f.=21, P<0.005). In 2004, adults lost significantly more body mass than juveniles during transit (tstat=−2.95; d.f.=19, P=0.004), but in 2005 we tube fed adults more during transport, and there were no differences by age. The mean change in body mass during transit ranged from −11 to +1% (n=42). The total mean change in body mass from capture on Laysan until release into the wild on Midway was an 11±13.8% gain (n=20) in 2004, but a 6±12.1% loss (n=22) in 2005. During 2004, the mean body mass increased significantly in the aviaries (paired t2004=−5.10, d.f.=19, P=0.003; ); however, in 2005, mass differences between release into the aviary and release into the wild were not significant (paired t2005=0.1, d.f.=21, P=0.42). Compared with juveniles (n=31), adults (n=11) lost significant body mass in the aviaries in both years (tstat=−4.7; d.f.=23; P<0.0005).
After the 2004 translocation, we confirmed the sex ratio of juvenile founders based on plumage dimorphism. The sex ratio was highly skewed towards males (14:6). We improved this ratio with individuals translocated in 2005 (10:12). The final age distribution of the founders was 31 HY juveniles (2.5–5 months old), eight SY sub-adults (∼14–18 months), or AHY (≥14 months) and two ASY (≥24 months) birds (Table 1).
The first three nests on Sand Island were discovered on 27 April 2005, 7 months after the translocation. Five of six females attempted nesting, and four were successful (i.e. hatched ≥1 egg). Four females re-nested at least once and two double brooded successfully (i.e. produced and fledged two sets of ducklings). The ASY female was the only hen that did not attempt nesting in 2005. In 2006, nesting was confirmed on both Sand and Eastern Islands. Nine of 12 females from the 2005 founder cohort nested successfully, and five first-generation (F1) females also successfully nested and fledged ducklings (Tables 2 and 3).
|Year||Cohort||Island||No. of females||Successful nestsa||No. of females to fledge a duckling|
|Clutch size||2005–2006||All||7.0 (sd=1.1; n=45)|
|Probability of a ♀ attempting to nest||2005||Founders 2004||0.83 (95% CI=0.75–0.97)|
|2006||All||0.96 (95% CI=0.93–0.99)|
|Probability of ♀ nesting successfully (1≥eggs hatch including re-nesting)a||2005||Founders 2004||0.83 (95% CI=0.75–0.97)|
|2006||All founders||0.72 (95% CI=0.69–0.76)|
|2006||F1 2005||1.0 (95% CI=0.90–1.0)|
|Probability of ducklings surviving to independence/post-fledgling (c. 60 days)||2005||Founders 2004||0.57 (95% CI=0.55–0.60)|
|2006||All founders||0.63 (95% CI=0.62–0.64)|
|2006||F1 2005||0.19 (95% CI=0.16–0.20)|
|Annual adult survival||2004–2006||All founders||0.92 (95% CI=0.83–0.98)|
|Annual post-fledgling juvenile survival||2005||F1 2005||0.90 (95% CI=0.71–1.0)|
|2006||F1,2 2006||0.82 (95% CI=0.81–0.84)|
|Mean ducklings fledged per ♀||2006||All founders||3.08 (se 0.71)|
|F1 2005||0.60 (se 0.30)|
Seventeen of 18 founding hens attempted nesting, producing 46 nests in the first two breeding seasons (Table 2). The effective population (Ne or the number of hens passing on their genes to the next generation) was 13 of 18 founder females. Successful breeding males are more difficult to determine; however, 13 of 24 male founders appeared to form pair bonds with successfully breeding females. Apparently, in the first breeding season, one male produced three successful broods. Sixty-seven juveniles survived to fledging during the first two breeding seasons.
Translocation and aviary survival (Ŝ) was 1.0. Annual post-release Ŝ of founders was 0.92 (95% CI: 0.83–0.98). Founders' Ŝ over 24 months was 0.83 (se=0.07). We estimated that 2005 F1 generation Ŝ of 11 post-fledgling juveniles (all with radio transmitters) during the first year was 0.90 (95% CI: 0.71–1.0), and 0.82 (95% CI: 0.81–0.84) for 2006 F1 and F2 generations (Table 3). The sex ratio of the 2005 F1 juveniles surviving to SY was 50:50. Additional demographic estimates are given in Table 3.
Mortality factors identified from recovered carcasses of founders and F1 generations were varied and included trauma from an Albatross attack (Phoebastria sp.) (n=1), an overhead power line collision (n=1), raptor predation (probably Falco peregrinus) (n=3), transmitter entanglement (n=1) and yolk sac infection (n=1; USGS National Wildlife Health Lab).
Population size and projections
The population size on Midway increased from zero before the translocation in 2004 to 104 by December 2006. The female population size at the end of 2006 was 28 adults and 28 juveniles (assuming half of the F1 and F2 fledglings were female). For the population projection, adult Ŝ was 0.92 and post-fledging Ŝ was 0.82. In 2006, the six ASY founder hens fledged an average 3.2 young per hen, and F1 SY hens fledged an average of 0.6 females per hen (Table 2). The resulting matrix for the female population was
The asymptotic growth rate of the Laysan teal population was λ=1.73. The elasticities of λ to changes in vital rates were high for adult survival (0.30), adult fecundity (0.34) and juvenile survival (0.30), but much lower for juvenile fecundity (0.05).
If the current rate of population growth continues, the population is projected to reach the Laysan Island population size in 2009, and reach a similar density if all emergent land area was appropriate habitat in 2010 (Fig. 2). In 2003, the estimated population size on Laysan (4.1 km2) was 523 ducks (Reynolds & Citta, 2007), which is a density of 126 ducks km−2. If this density is applied to the entire land area of Midway (6.2 km2), the atoll would support ∼785 ducks (Fig. 2).
To date, the translocation of Laysan teal to Midway Atoll has been a successful endeavor. Like Laysan's, Midway's population is vulnerable to catastrophes such as tsunamis, hurricanes, accidental introduction of rats or other predators and disease. However, with the exception of global sea-level rise, it is unlikely that disasters will strike both atolls simultaneously. Thus, two populations provide some degree of insurance against catastrophic events. Given the uncertainty associated with re-establishing endangered animal populations, and the economic expense of these activities, we believe that our efforts provide information that can inform future translocation efforts for Laysan teal and other island endemics birds.
Our methods used both intensive monitoring pre-translocation on Laysan and post-release at Midway. On Laysan, before the translocation, we used long-term monitoring to confirm an increasing source population (USFWS, 2004; Camp et al., in press). Additionally, nest monitoring helped to determine the Laysan teals' breeding phenology, verify a successful breeding season and allow for our logistic preparations (Reynolds et al., 2007). Then brood monitoring allowed for selection of suitable juveniles as candidate founders for translocation. Our aim was to remove birds from Laysan during a period of population growth, so as not to negatively impact the source population with translocation removals. We also sought to maximize the available genetic variation of Midway's founders with non-siblings. Only birds in good body condition were selected for translocation. We believe that the careful selection of translocation candidates, plus the non-stop transit to Midway, and intensive nutritional support during the translocation, contributed to total transfer survival.
Juvenile birds maintained and recovered body condition better than adults during transport and aviary holding. However, transport and post-release survival did not differ, between adults and juveniles, and rapid release of adults was successful. We attributed higher release weights in 2004 (11% gain) compared with 2005 (6% loss) to a shorter transit time (10 h faster in 2004) and consumption of commercial mash after release into the aviaries. In 2005, some of the commercial mash had been in storage for ≥1 month and was not appealing as a novel food. Future translocations should prioritize selecting HY and sub-adult founder candidates, minimizing transport and holding times and ensuring that aviary mash is very fresh.
Post-release monitoring allowed us to record demographic rates that describe a rapidly growing population on Midway. Contrary to expectations, we observed that females translocated as juvenile birds were more likely than those translocated as adults to fledge ducklings successfully during the first year post-release. Midway's adult and post-fledgling survival is similar to estimates from Laysan Island (Reynolds & Citta, 2007); however, re-nesting rates, clutch sizes and duckling survival were higher than at Laysan (Moulton & Marshall, 1996; Reynolds & Work, 2005).
Using vital rates observed on Midway in 2006, we projected that the population would grow to a larger number of birds than those currently existing on Laysan in <10 years (Fig. 2). This projection and the post-release demography show that the early stages of the Laysan teal population establishment have been successful. Our projection did not include habitat suitability or consider vegetation management activities that may influence vital rate (i.e. mowing). Because demographic variability, environmental stochasticity and density dependence were not included, it is likely that the observed population growth rate will be less than our projection. To verify long-term population persistence, population monitoring should be continued given this uncertainty.
Although clearly important, demographic monitoring can be expensive and logistically challenging on remote islands. Our elasticity analyses suggest potential approaches to focusing the monitoring process. Because the combined elasticity for adult survival and fecundity (0.64) was substantially greater than for juvenile survival and fecundity (0.35), we infer that small changes in adult vital rates will have a greater effect on the population trend than will similar changes to juvenile vital rates. Therefore, given limited resources, if measuring the fecundity and survival of adults would be valuable vital rates to monitor. Furthermore, these results suggest that management decisions to maximize the population growth rate should prioritize actions that will maintain or increase adult survival and fecundity. However, we caution that elasticity values may be substantially different when the population size is larger-either as a result of density-dependent modifications to vital rates (de Kroon, van Groenendael & Ehrlén, 2000), changes in vegetation or management that influence survival or reproductive rates or because this nascent population does not have a stable age distribution (Fefferman & Reed, 2006).
Our post-translocation monitoring demonstrated that the Laysan teal has unexpected reproductive plasticity and is capable of using novel habitats. On Midway, Laysan teal reproduced more quickly than they do on Laysan. Their low reproductive rate on Laysan may reflect carrying capacity or food limitation, and the hyper-saline ecosystem is apparently not essential to their ecology. Furthermore, translocated ducks on Midway used a wide variety of vegetation types for nesting and foraging that are absent on Laysan (i.e. v. encelioides). The recovery plan calls for conservation of the source population on Laysan, re-establishing populations on four to five other islands to increase the species resistance to demographic and environmental uncertainties and periodic genetic supplementation from the Laysan population (USFWS, 2004). The outlook for this conservation strategy is optimistic, given the ability of Laysan teal on Midway to nest and forage in novel and non-native vegetation and their potential to reproduce more rapidly than they do on Laysan.
We suspect that the small number of founders, limited land area (0.1 km2) and low elevation (maximum 3 m) susceptible to storm surge, plus the marginal habitat contributed to the disappearance of Laysan teal translocated to Southeast Island of Pearl and Hermes Atoll in 1967–1968. Evidence from our translocation to Midway suggests that if translocation sites have fresh water, abundant vegetative cover and a rich prey base, additional populations on mammalian predator-free islands have a good chance of success, further reducing the species risk of extinction, and replacing a missing component of Hawaiian ecosystems.
We thank E. Baldwin, J. Breeden Jr., B. Casler, B. Christenson, the American Islander Crew, J. Dhundale, H. Freifeld, N. Jarret, M. Johnson, J. Kelly, K. Kozar, S. Kropidlowski, M. Laut, E. Lund, A. Marshall, Midway's fire-crew, D. Palawski, C. Rehkemper, J. Shore, G. Shuman, C. Vanderlip, J. Walters, R. Woodward, T. Bodeen, J. Leineke, E. Flint and USFWS volunteers for assistance with this project. T. Work provided veterinary care during capture, transport and aviary holding. N. Shema constructed the transport boxes; M. Williams and P. McClelland at New Zealand Department of Conservation offered excellent advice based on their experience with the Campbell Island Teal. This was an interagency project. Funding was raised through grants from the National Fish and Wildlife Foundation, USGS QRP grants, Friends of Midway Atoll and USFWS Pacific Islands Ecological Services. In kind, staff and logistical support was provided by Pacific Island Ecosystems Research Center, USGS National Wildlife Health Lab, Midway Atoll NWR, National Oceanic Atmospheric Agency, the Wildfowl & Wetland Trust, USFWS Pacific Islands Fish and Wildlife Office and the Hawaiian Islands National Wildlife Refuge. D. Armstrong, J. M. Reed, J. Walters and T. Work provided reviews of this paper. J. Breeden Jr., L. Crampton, C. Forbes Perry assisted with revisions. Use of product names is for descriptive purposes and does not imply endorsement by the US Government.
- AOU (1998). Checklist of North American birds. Washington, DC: American Ornithologist's Union.
- 1983). Ecological problems associated with disruption of dune vegetation dynamics by Casuarina equisetifolia at Sand Island, Midway Atoll. Atoll Res. Bull. 261, 1–19. , & (
- 2007). Using adaptive management to determine requirements of re-introduced populations: the case of the New Zealand Hihi. J. Appl. Ecol. 44, 953–962. , & (
- 1981). Hawaiian birdlife. Honolulu, HI: University of Hawaii Press. (
- 2004). Avian extinction and mammalian introductions on Oceanic Islands. Science 305, 1955. , , , & (
- 2001). Holocene lake sediments in the Maha'ulepu caves of Kaua'i: evidence for a diverse biotic assemblage from the Hawaiian lowlands and its transformation since human arrival. Ecol. Monogr. 71, 615–642. , , , , , & (
- 2003). 2004 Red list changes for waterfowl. Threatened Waterfowl Res. Group Newslett. 14, 82–84. & (
- A statistical test to show a negligible trend: comment. Ecol. Appl. in press. , , &
- 2001). Matrix population models, construction, analysis, and interpretation. Sunderland, MA: Sinauer Associates Inc. (
- 1996). Ancient DNA and island endemics. Nature 381, 484. , , , , , & (
- 1997). The effectiveness of removing predators to protect bird populations. Conserv. Biol. 11, 395–405. & (
- 1984). Analysis of survival data. New York, NY: Chapman and Hall. & (
- 1990). Alteration of native Hawaiian vegetation: effects of humans, their activities and introductions. Honolulu, HI: University of Hawai'i Press. & (
- 2007). Recovery of the Galápagos Rail (Laterallus spilonotus) following the removal of invasive mammals. Biol. Conserv. 138, 250–254. , , , , & (
- 1973). The natural history of Laysan Island, Northwestern Hawaiian Islands. Atoll Res. Bull. 171, 1–361. & (
- 2006). A vital rate sensitivity analysis for nonstable age distributions and short-term planning. J. Wildl. Mgmt. 70, 649–656. & (
- 2002). Control and eradication of the introduced grass, Cenchrus echinatus, at Laysan Island, Central Pacific Ocean. In Turning the tide: the eradication of invasive species, occasional paper of the IUCN species survival commission no. 27: 110–115. Veitch, C.R. & Clout, M.N. (Eds). Auckland: IUCN-The World Conservation Union. & (
- 1993). Patterns of vegetation and nesting seabirds at Midway Atoll, Northwestern Hawaiian Islands. Master's thesis, University of Oregon, Eugene. (
- 1989). Translocation as a species conservation tool: status and strategy. Science 245, 477–480. , , & (
- 1996). Physiological methods in wildlife research. In Research and management techniques for wildlife and habitats: 275–306. Bookhout, T.A. (Ed.). Bethesda, MD: The Wildlife Society. & (
- 2007). Reassembling Island ecosystems: the case of Lord Howe Island. Anim. Conserv. 10, 22–29. , & (
- IUCN (1998). Guidelines for re-introductions. Prepared by the IUCN/SSC Re-introduction Specialist Group. Cambridge, UK: IUCN.
- 1994) Archaeological subsurface testing for supplemental cultural resources overview survey: Midway Atoll. Report 1453-031194, PHRI Inc., Honolulu, HI. & (
- 1996). Population analysis. In Research and management techniques for wildlife and habitats: 275–306. Bookhout, T.A. (Ed.). Bethesda, MD: The Wildlife Society. (
- 1958). Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53, 457–481. & (
- 2000). Elasticities: a review of methods and model limitations. Ecology 81, 607–618. , & (
- 1986). Elasticity: the relative contribution of demographic parameters to population growth rate. Ecology 67, 1427–1431. , , & (
- 1994). Some ecological considerations and computer-based approaches for the identification of potentially suitable release sites for reintroduction programmes. In Reintroduction biology of Australian and New Zealand Fauna: 1–5. Serena, M. (Ed.). Chipping Norton: Surrey Beatty and Sons. (
- 1993). Matrix methods for avian demography. In Current ornithology: 139–184. Power, D. (Ed.). New York, NY: Plenum Press. & (
- 1996). Laysan duck. The birds of North America 242, 1–20. & (
- 1984). Biology and conservation of the Laysan duck (Anas laysanensis). Condor 86, 105–117. & (
- 1982). Fossil birds from the Hawaiian Islands: evidence for wholesale extinction by man before western contact. Science 217, 633–635. & (
- 1991). Descriptions of thirty-two new species of birds from the Hawaiian Islands: part I, non Passeriformes. Ornithol. Monogr. 45, 1–88. & (
- 1995). Remains of land birds from Lisianski Island, with observations on the terrestrial avifauna of the northwestern Hawaiian Islands. Pacific Sci. 49, 111–125. & (
- R Development Core Team (2005). R: A language and environment for statistical computing. Version 2.1.1. Vienna, Austria ISBN 3-900051-07-0, http://www.R-project.org
- 2001). Isles of refuge. Honolulu, HI: University of Hawai'i Press. (
- 2007). Post-fledging survival of Laysan ducks. J. Wildl. Mgmt. 71, 383–388. & (
- 2007). Laysan Teal nesting phenology and site characteristics on Laysan Island. Wildfowl 57, 54–67. , & (
- 2000). History and current status of the Laysan Duck (Anas laysanensis) in captivity. ‘Elepaio, 60 ( 7 ): 59–65. & (
- 2006). Diet composition and terrestrial prey selection of the Laysan Teal on Laysan Island. Atoll Res. Bull. 543, 181–199. , & (
- 2005). Mortality in the endangered Laysan Teal, Anaslaysanensis: conservation implications. Wildfowl 55, 31–48. & (
- 1994). Translocations in New Zealand: an overview. In Reintroduction biology of Australian and New Zealand fauna: 43–46. Serena, M. (Ed.). Surrey Chipping Norton, UK: Beatty and Sons. (
- 2007). Developing the science of reintroduction biology. Conserv. Biol. 21, 303–312. , & (
- 2006). Extinction and biogeography of tropical pacific birds. Chicago: University of Chicago Press. (
- 2007). popbio: construction and analysis of matrix population models. Version 1.0.1. Vienna, Austria ISBN 3-900051-07-0, http://www.R-project.org & (
- USFWS (2004). Draft revised recovery plan for the Laysan duck (Anas laysanensis). Portland, OR: US Fish and Wildlife Service.
- USGS. (2005). Translocation of endangered Laysan ducks to Midway Atoll National Wildlife Refuge (2004-5). USGS Fact Sheet 2005-3128, November 2005.
- 2004). Mortality in Laysan ducks (Anas laysanensis) by emaciation complicated by Echinuria uncinata on Laysan Island, Hawaii, 1993. J. Wildl. Dis. 40, 110–114. , & (
- 1999). Biostatistical analysis. Upper Saddle River, NJ: Prentice-Hall. (