The Scientific Bases for Preservation of the Mariana Crow (1997)

This chapter describes various recovery options for the aga, and compares the advantages and disadvantages of these options.

If current management efforts were to cease on Guam, the aga would most likely be extirpated on the island when the current adults die. There has been no nonassisted production of young in the wild since the 1980s (Wiles and others 1995), and only 5 young have fledged since 1990 (Aguon 1996). In the absence of an effective snake-control program, it is unlikely that the native crow population in Guam can be sustained, much less increased. Although the Rota population is not being actively managed, the population there is decreasing. Furthermore, unless effective control of the brown tree snake and quarantine measures are developed and implemented, there is a high probability that snakes would eventually colonize Rota and drive the aga to extinction.

Because of the unique circumstances posed by the brown tree snake on Guam, many management techniques developed for other endangered birds are inappropriate and cannot be applied to the aga. Although management efforts are evolving quickly, the current management procedures on Guam remain experimental.

Present management efforts regarding the Guam population have not resulted in either population stability or growth.

Reestablishment of a viable crow population on Guam is highly improbable unless more-effective snake-control procedures are developed and implemented on a larger scale in crow habitat. It must be noted, however, that the current experimental work on snake control being carried out by the Guam Department of Agriculture Division of Aquatic and Wildlife Resources (DAWR) and US Department of the Interior National Biological Service (NBS) personnel might become essential for developing better crow-management strategies on Rota or in places into which the crow might be introduced (see ''Translocation Options" later in this chapter). Although there is inherent benefit in continuing this work, it is not likely to result in the recovery of the Guam population. On Rota, the persistence time of the crow population could probably be increased through a combination of protection, public education, and habitat preservation, unless the brown tree snake colonizes the island. Without effective snake inspection, cargo quarantine, and control methods, such colonization is likely to occur and compromise the Rota population.

Intensive management involves the direct manipulation of nesting pairs in an attempt to increase the reproductive output of the population. The typical strategy is to remove eggs from wild pairs, artificially incubate them, hand-rear the resulting young, and either foster the young back into wild nests or release them with a soft-release technique (NRC 1992). Those methods have worked well on a variety of species, including European corvids (Delvaux 1991), North American corvids (Marzluff 1993; Marzluff and others 1994), and the 'alala (Hawaiian crow) (Derrickson 1994; Kuehler and others 1995).

The advantages and disadvantages of intensive management are discussed in a National Research Council report (see table 6.1 in NRC 1992), and they remain relevant for the aga with the following modifications based on recent experience with common ravens (Corvus corax), American crows (Corvus brachyrhynchos), black-billed magpies (Pica pica), and Hawaiian crows (Derrickson 1994; Kuehler and others 1995; Marzluff 1993; Marzluff and others 1994).

Intensive efforts to recover the 'alala began in 1992 after the Research Council committee's report (NRC 1992). Much has now been learned that is directly applicable to recovery of the aga. In particular, the technical aspects of observing, monitoring, and managing the wild population of 'alala are directly transferable to needed field studies and nest manipulations on Guam and Rota. Captive-propagation techniques developed for the 'alala are also directly applicable to the aga with slight modification, as discussed later in this chapter.

Advantages of egg removal are more certain now than they were before the 'alala-restoration project. Hatchability and survivability of eggs and chicks are increased in captivity, and the annual cohort of young is much larger with intensive management than without it (Derrickson 1994; Marzluff 1993; Marzluff and others 1994). Most of the problems previously encountered with this technique have been overcome. Optimal techniques for artificial incubation have been developed and already applied to aga eggs (Brock 1996). A key is to allow parents to incubate eggs for 5–7 days before egg removal. The concern that all nesting pairs will not lay another clutch is also diminished. Corvids almost always renest after clutch removal (Marzluff 1993; Marzluff and others 1994); given the propensity of aga to renest after natural failure (Morton 1996a), we expect them to renest readily after clutches are removed. Hatching and rearing chicks in captivity have the disadvantage of requiring expanded staff and facilities, as in the case of the 'alala.

Soft methods of release range from controlled release over a period of days or weeks (hacking) to more elaborate systems involving holding birds for extended periods in outdoor flight cages in the habitat where they are likely to be released. Disadvantages associated with this technique (NRC 1992) are greatly reduced by the knowledge gained during 'alala restoration. The technique remains labor-intensive and expensive, but there is now less concern the captive-reared birds will not learn appropriate survival or social skills. Captive-reared corvids readily integrate with wild birds after release and have no apparent problems in socializing, foraging, or avoiding predators (Banko 1996; Marzluff 1993; Marzluff and others 1994). If birds are to be held for long periods (over 6 months), wild birds can be housed with captive-reared birds as tutors to facilitate social integration (Marzluff and others 1994).

A key to successful release of corvids is to raise them in groups and allow them to acclimate to the release area in a flight cage for several weeks or months before release. If corvids are raised with other conspecifics from hatch through release, puppet-rearing and tutoring are not necessary to enhance survival or breeding. Captive-reared black-billed magpies and American crows raised without puppets or tutors are breeding successfully in the wild.

The goals of a release project influence the type of rearing used. If the goal is to produce corvids that will readily assimilate with wild birds, tutoring and releasing birds when they are several months old can be beneficial (Marzluff and others 1994). However, if the goal is to colonize a new area and managers want released birds to remain near the release site, techniques for improving socialization

and rapid acquisition of independence, such as tutoring, puppet-rearing, and maximizing time before release, might actually be detrimental (Marzluff and others 1994).

The advantages of soft releases listed in the Research Council report (NRC 1992) are all relevant to aga. The ability to colonize vacant portions of the range is especially beneficial. In fact, releases in vacant areas are the most likely to establish new breeders in the population quickly, apparently because territorial defense in occupied habitats reduces the ability of released birds to establish their own territories. The only hand-reared corvids known to breed in the wild were released in previously vacant areas.

Fostering nestlings to wild pairs is the least expensive and most natural way to release birds born in captivity. It will probably be an important option for aga when reproductive success in the wild returns to normal, and it remains a good option when only 1 chick is raised in captivity. Results of experiments with American crows, black-billed magpies, and common ravens suggest that chicks will be readily adopted even when several weeks old (Marzluff and others 1994). Fostering older chicks (10–20 days old) is preferred because they are less vulnerable to parental neglect or aggression. Older chicks can be substituted for newly laid eggs without apparent problems (Marzluff and others 1994).

Released birds should be closely monitored so that knowledge about limiting factors, sociability, and productivity can be gained. Radiotelemetry techniques have been developed for 'alala and tested successfully on aga, and 1 radio-tagged aga has nested in the wild (Morton 1996a).

Conservation programs that involve translocation (moving animals between populations, moving animals into unoccupied former habitat, or releasing animals into new habitat) are biologically risky, expensive, time-consuming, and usually unsuccessful (Ebenhard 1988; Griffith and others 1989; Snyder and others 1996). Translocation should never be attempted except as a last resort, and even then only after comprehensive planning to ensure the long-term survival of both the donor and recipient populations and their communities. Various potential translocation options for the aga are discussed below with recognition of the concerns raised above regarding the success of past programs for other species.

There are many more crows on Rota than on Guam, so an argument could be made to translocate some birds to Guam as a population enhancement strategy. Such a translocation would presumably serve 2 immediate objectives: the Guam population would be increased, and reproductive problems related to senescence would be rectified. However, with the current tree-barrier method, even a relatively small increase in the Guam population will require substantial increases in time, personnel, and resources. It is very unlikely that a large increase in the Guam population will be possible until snakes are controlled in or excluded from extensive areas of suitable habitat. Although removal of eggs or nestlings from the Rota population would probably entail little social or demographic disruption, given the renesting potential of this species, such removal cannot be justified biologically until the size and demography of the Rota population are accurately determined. Removal of birds from Rota might also jeopardize the present, politically sensitive habitat-preservation efforts on that island and hence the long-term security of its population.

Because snake problems on Guam are so severe, it might be advisable to move some or all of the Guam crows to Rota. Moving some of the crows to Rota might increase the genetic diversity of the Rota population, preserve the genetic diversity of the Guam population by incorporating it into the Rota population, increase the demographic stability of the Rota population, and provide a greater chance of continued survival and reproduction of individuals from Guam. Problems to be dealt with include being comfortable with genetic mixing of the populations. Tarr and Fleischer (1995) have noted that the genetic differences between populations are not great enough to raise concern and have suggested that the demographic contribution might be more critical than genetic considerations at this point.

However, given the advanced age of the majority of the Guam birds, their demographic contribution to the Rota population would be minimal. Furthermore, taking birds from Guam would weaken the Guam population. Therefore, this option should be considered only after population-enhancement efforts on Guam are deemed unsuccessful. In addition, habitats would need to be carefully assessed because the aga uses somewhat different habitats on the two islands. Finally, both habitat availability and the social structure of the population on Rota would need to be characterized to ensure that suitable vacant habitat is available and that translocating birds into the population would cause minimal social disruption.

Removing all crows from Guam and putting them all on Rota would be a major event in the species's history and should be undertaken only after careful consideration. The translocated birds might have a better chance of surviving on Rota, but the extirpation of the Guam population would have far-reaching effects. First, having all aga on 1 small island is inherently unsafe and would substantially increase the risk of extinction of the species as a whole because of typhoons, disease, predators, or progressive habitat fragmentation and modification. Second, loss of another native forest bird from Guam could make further habitat-restoration efforts and snake-eradication research more difficult to defend. Third, loss of a native species from Guam would have to be discussed among many local groups, especially Chamorro groups in whose culture the species has been inherent for centuries. These drawbacks could be mitigated if, after brown tree snake control is substantially improved, birds from Rota were reintroduced to Guam.

In most cases, the viability of severely declining species is improved when populations are established in several locations. Maintenance of multiple populations decreases the chance that one catastrophic event would cause extinction. Furthermore, genetic diversity might be better maintained because each population

differentially experiences the effects of genetic drift, so fewer unique alleles are lost to the species overall (although each population will lose unique alleles). Finally, if each population expands, there will be greater demographic stability for the species.

Because the aga has occurred only on Guam and Rota during historical times, any effort to establish additional populations will entail releasing birds outside their historical range. Such introductions are generally not successful (Griffith and others 1989). Translocation of the aga to other islands would have to be evaluated carefully from the viewpoint of its contribution to persistence of the species and its effects on the flora and fauna of the other islands. Of course, the number, sex, social status, age, and so forth of the crows to be introduced would have to be considered.

The most likely places to search for new sites would be nearby islands where the habitat would be appropriate and the crow's presence would have minimal effects on other species. Some of the more-northerly, uninhabited islands or even the small island of Agiguan, between Rota and Tinian (Figure 1-1), might be appropriate; however, investigations to determine potential environmental impacts would have to be conducted before a selection is made. Other possible sites include islands in Palau, Pohnpei, or the Marshall Islands, where forest-bird assemblages are similar to those in the Mariana archipelago.

Movement of all of Guam's birds to a new site would have to be considered in light of potential genetic consequences. Because the Guam population appears to have greater genetic diversity than the Rota population, it might be advantageous to introduce some of Guam's birds to a new island and others to Rota to increase genetic diversity in an established population that is likely to persist at least in the short term. The same would hold true if only some of Guam's birds were to be translocated.

Translocation of some Rota birds to a new site should be undertaken with caution because it is the only successfully reproducing population of aga. However, if a complete and accurate census of the Rota crows indicates that there are unpaired birds, birds without nest sites, or birds whose territories are slated for destruction by development or that eggs can be harvested without adversely affecting productivity, there is probably good reason to invest in a third population.

Releasing the captive birds back into the wild warrants serious consideration. Although the captive birds potentially constitute a third population themselves, they have not yet reproduced well and therefore cannot be considered a viable population at this point. The captive program was initiated for research and training purposes and not specifically to establish a viable captive population. Because husbandry techniques have been established for corvids in general, there is less need for a captive population. If a captive population were required in the

future, one could be established on an emergency basis. It might be better to put the captive birds back into the wild if there were places where they could survive and contribute to population recovery or if monitoring their behavior would provide important information on their biology. This option keeps the focus of the program on research on and recovery of wild populations. Because most of the captive birds were wild-caught, they should demonstrate normal survival on release into the wild.

Given that all the captive birds in the Mariana Archipelago Rescue and Survey program were derived from the Rota population, their return to Rota would have minimal beneficial effects on the wild population's demographic or genetic structure. Furthermore, the risk of introducing diseases acquired in captivity to the population on Rota must be weighed carefully. Even if the risk is very low, the consequences would be so serious with respect to species survival that this option should be avoided. In contrast, given the precarious state of the Guam population and the low probability of short-term survival, the risks to the aga associated with introducing the captive birds into the Guam population are much more acceptable. Even if translocating the captive birds to the Guam population would not increase population viability substantially in the short term, it should allow for the refinement of practical translocation methods and the continuation of critical research on snake management and control, habitat use and preference, and the role of senescence in reduced productivity.

The Guam birds could be moved into captivity until effective snake-control methods are available. However, this plan presents numerous problems. First, it would constitute extirpation of the Guam population. Second, husbandry techniques have largely been worked out, so there is no purpose in keeping birds in captivity unnecessarily. Third, removing birds to captivity would preclude conducting critical field research. Fourth, removing birds from the wild could subvert current efforts to control the brown tree snake and to preserve wild populations and habitats. Fifth, captive breeding is a conservation technique with a number of important, but often-overlooked limitations, including difficulties in establishing self-sustaining captive populations, poor success in reintroducing captive-bred stock to the wild, progressive domestication, disease outbreaks, preemption of alternative conservation methods, high program costs, and difficulties in maintaining administrative continuity (Snyder and others 1996). Finally, establishing a viable captive population by using Guam birds alone would be impractical, given the age of most of the birds and the apparent age-related reproductive problems. Even if those problems could be overcome, offspring of the captive population would have to be released into new areas unless effective, large-scale methods of snake control or eradication were developed and implemented on Guam.