Genetic diversity of the Critically Endangered Philippine Eagle Pithecophaga jefferyi ( Aves : Accipitridae ) and notes on its conservation

Acknowledgements: We gratefully acknowledge the Energy Development Corporation and the Philippine Eagle Foundation (PEF) and Department of Environment and Natural Resources Bidiversity Management Bureau for providing the samples used in this study. Abstract: The Philippine Eagle Pithecophaga jefferyi is a diurnal raptor endemic to the Philippines. Its distribution is restricted to remaining forests on the islands of Luzon, Samar, Leyte and Mindanao. The Philippine Eagle is classified as a Critically Endangered species under the IUCN Red List, with a high end estimated population of only 500 breeding pairs in the wild. Population decline has been attributed to continuing deforestation, particularly since the mid-1900s, and hunting. This study aimed to identify the effects of population decline on the genetic structure of the present population of the Philippine Eagle by sequencing 1132bp of the mitochondrial control region from 22 individuals. Control region haplotype diversity (h = 0.8960±0.05590) and nucleotide diversity (π = 0.006194±0.003372) are comparable with other accipitrid species. Maximum likelihood trees and network analysis show that the Luzon and Samar individuals come from different lineages, but both shared a common ancestral population with the Mindanao population. The genetic diversity, multimodal mismatch distribution for the control region and high frequency of lower class modes all indicate a recent bottleneck for the Philippine Eagle population. Possible strategies for conservation are discussed.


INTRODUCTION
Family Accipitridae includes eagles, kites, goshawks, hawk-eagles and Old World vultures.The Philippine Eagle Pithecophaga jefferyi is a diurnal raptor that was initially thought to be closely related to large carnivorous birds of the subfamily Harpiinae, such as the Harpy Eagle Harpia harpyja and the Crested Eagle Morphnus guianensis, based on size and morphology (Peters 1931).However, genetic analysis revealed that the Philippine Eagle is distinct from these species based on two mitochondrial regions and a nuclear intron, and more closely related to smaller species of snakeeagles of the subfamily Circaetinae (Lerner & Mindell 2005).The accipitrids are some of the most endangered species of raptors, with dwindling numbers attributed to habitat loss and other anthropogenic activities (Lerner & Mindell 2005).The Philippine Eagle is restricted to only four islands in the Philippines: Luzon, Samar, Leyte and Mindanao (Rabor 1971;Kennedy 1977).These eagles nest in tall trees usually in steep slopes near rivers and tributaries within mature dipterocarp forests (Kennedy 1977), and the observation that they prey on monkeys led to the common name Monkey-eating Eagle and the genus name Pithecophaga.Subsequent investigations revealed that Philippine Eagles also prey on other vertebrates, including mammals, reptiles and birds (Gamauf et al. 1998;Salvador & Ibanez 2006).These eagles have a breeding cycle that lasts around two years, with one egg being laid at a time (Kennedy 1981).t Philippine Eagle populations are continually being threatened by hunting and loss of habitat (Kennedy 1977;BirdLife International 2013), and continuing decrease in forest areas over the years has left this species under threat of extinction.A recent study estimated about 82-233 breeding pairs left in the wild in Mindanao and around 500 pairs throughout the Philippines (Bueser et al. 2003;Salvador & Ibanez 2006).IUCN classified these raptors as Critically Endangered because of their unique breeding characteristics and the unrelenting threats they face (Birdlife International 2013).
In most cases a decrease in the population size will lead to a decrease in genetic diversity.Low genetic diversity would normally increase the inbreeding rate, which in turn would lead to the lowering of the population's overall fitness and increase the extinction risk (Brook et al. 2002).The assessment of genetic diversity undertaken in this study will inform conservation actions for this species and address problems associated with captive breeding programs already in place.

MATERIALS AND METHODS
Twenty-two individuals were studied.Feather samples from two Luzon individuals and a muscle sample from a Samar individual were provided by the Philippine Department of Environment and Natural Resources-Biodiversity Management Bureau (DENR-BMB).Blood samples from the 19 remaining individuals were provided by the Philippine Eagle Foundation (PEF) from their captive breeding facility in Davao City, Mindanao Island.Blood extraction was performed via venipuncture by a veterinarian during the annual medical check-up of the raptors; no animals were harmed.Samples from PEF consisted of 16 individuals originating from Mindanao forests and three individuals bred in captivity, whose parents originated from Mindanao forests as well (Table 1).Localities for each Philippine Eagle included in the study are shown in Fig. 1.
DNA from muscle samples was extracted using Wizard Genomic® DNA extraction kit (Promega, USA) while feather and blood samples were extracted using DNeasy Blood and Tissue Kit (Qiagen, USA) following manufacturers' instructions.
PCR conditions for the control region consisted of the following: an initial denaturation of 94 o C for three minutes; 35 cycles of 94 o C for one minute, 55 o C for 1.5 minutes, and 72 o C for 1.5 minutes; and a final extension of 72 o C for five minutes.The PCR products were run on a 1% agarose gel with ethidium bromide and the bands were viewed under UV light.
Bands from the gel that matched correctly with the expected size of the fragment were excised and purified using Qiaquick® Gel Extraction Kit (Qiagen, USA), following the manufacturer's instructions.Purified PCR products were sent to 1stBASE in Selangor, Malaysia for Mismatch analysis was done for the population and is shown in Fig. 4. Results for this analysis reveal a multimodal distribution.Test of goodness of fit was assessed using the raggedness index where a nonsignificant value indicates a non-significant difference between the observed and the simulated demographic expansion model.Based on the raggedness index for

6340
the test of goodness of fit of the observed data with demographic expansion model, the Philippine Eagle population has undergone a demographic expansion (r=0.02230,P-value=1.0000).
Tajima's D test, Fu's F S test and Fu and Li's F or D F , however, contradict the mismatch analysis and do not support demographic expansion.Results in these statistics contained negative values; however, none of these values were significantly different from zero (Table 3), thus population expansion is not supported.Tajima's D and Fu's F S are known to be more sensitive than mismatch distributions in detecting populations undergoing expansion (Fu 1997).Negative values of Tajima's D and Fu's F indicate an excess of single substitutions brought about by population expansion.In addition, Fu and Li's F and D F use an outgroup sequence to test for evidence of population expansion and are shown to be less sensitive to biases of small sample sizes (Fu & Li 1993).F ST among populations were generated (Table 4).All computed F ST were positive.Based on Sewall Wright's hierarchical F statistics, these F ST values indicated moderate to high genetic difference between the subpopulations.However, only the F ST for the overall population was significant (P-value = 0.0401).Results from AMOVA showed that most of the variations of the control region were found within populations (Table 5).

DISCUSSION AND CONCLUSION
The sample size for this study was 22 individuals.Breeding pairs in the wild could be at most 500 (1000 individuals); however, this number may be an overestimate and it may go down to around 350 breeding pairs.The actual numbers may even decrease further if we assume that only around 40% of the habitat is used (Bueser et al. 2003).Nevertheless, the number of individuals used in this study is a good sample for a Critically Endangered species such as the Philippine Eagle (Images 1,2).
Genetic diversity of the Accipitridae was summarized in Table 3 in the study of Lerner et al. (2009).CR haplotype diversity and nucleotide diversity observed from the Philippine Eagle was comparable with values obtained from non-endangered raptors such as the Bearded Vulture Gypaetus barbatus (h=0.932,π=0.0292,Godoy et al. 2004) and White-tailed Eagle Haliaeetus albicilla (h=0.746,π=0.00680,Hailer et al. 2007).Genetic diversity was also comparable with other threatened raptor species such as the White-rumped Vulture Gyps bengalensis (h=0.76,Johnson et al. 2008), Harpy Eagle (h=0.906,π=0.0076,Lerner et al. 2009) and in the Red Kite Milvus milvus (h=0.610,π=0.0032,Roques & Negro 2005).These values were higher than some raptors with stable populations such as in the White-bellied Sea Eagle Haliaeetus leucogaster (h=0.350,π=0.000806,Shephard et al. 2005) and in Bonelli's Eagle Aquila fasciatus (h=0.542,π=0.00240,Cadahia et al. 2007).It should be noted that the Philippine Eagle, in contrast to these taxa mentioned, is an island species whereas the other raptors species inhabit huge distribution ranges.
High haplotype diversities and low nucleotide diversities may indicate a population bottleneck followed by demographic expansion (Grant & Bowen 1998).This, however, should not be used to disprove a recent population bottleneck, since long-lived species have been known to retain high genetic diversities even after experiencing a recent bottleneck (Hailer et al. 2006).Fu and Li's F and D F , Tajima's D, and Fu's F S do not support demographic expansion.In addition, the presence of high frequencies of lower class modes (i.e., 0 and 3 nucleotide differences in the mismatch distribution) is a pattern observed in simulations of populations that experienced a recent bottleneck, although other factors can also contribute to this scenario (Rogers & Harpending 1992;Lerner et al. 2009).Similar patterns of mismatch distribution have been observed in some subpopulations of birds that experienced a recent bottleneck (Harpy Eagles Harpia harpyja, Lerner et al. 2009; Prairie Chickens Tympanuchus cupido, Johnson et al. 2007).Similarly, the Philippine Eagle populations may have experienced a recent bottleneck.
One possible bottleneck event for the Philippine Eagle population would be the decrease in population attributed to a decrease in habitat over the years.In 1521, when the first Spanish colonizers set foot in the Philippine archipelago, it was estimated that 90% of Images 1 & 2. Philippine Eagle Pithecophaga jefferyi the country's land area was covered by forests, which decreased to around 71% in 1900 because of the increase in human population and agricultural demands (Environmental Science for Social Change 1999).In 1969 only an estimated 34.9% of forest covered the total land area of the country.This was also the time when Rabor (1968) warned the public about the declining population of the Philippine Eagle as a result of persistent forest exploitation and hunting (Salvador & Ibanez 2006).In 2003, only 23.9% (72,000km 2 ) of forest cover remained (Forest Management Bureau 2011), of which less than 10,000km 2 are considered primary forests (Lasco et al. 2001).
Although this bottleneck event is a possibility, there is a need to include additional data to prove this.It is necessary that data from other loci are evaluated to determine recent demographic changes that could have occurred.Though it is ideal to increase the sample size, this may prove difficult for the Philippine Eagle because of constraints imposed by difficulty in sampling as well as existing Philippine laws that prohibit such activities.
F ST indicates that the subpopulations are isolated from each other.On the other hand, AMOVA does not support this and a clear division between Luzon, Samar and Mindanao haplotypes were not observed in the MJ network and the ML tree.The MJ network seemed to indicate that the unique haplotype of the Samar sample came from a different phylogeographic origin from that of the Luzon samples, but both could have originated from Mindanao haplotypes.However, since the Philippine Eagle population of Luzon and Samar were underrepresented in this study, this conclusion may be premature, as additional samples are required to properly evaluate the scenario.Additional samples will also be needed in order to properly evaluate the divergence of the subpopulations.
Historical pattern of gene flow for the populations of the Philippine Eagle is also lacking.For now, we can say that island populations may be isolated because, despite current research efforts to monitor its population, documentation of inter-island migrations for the species was not observed.In this respect, populations could have solely dispersed during glaciation events as recent as the late Pleistocene when the Greater Mindanao was formed (Brown & Diesmos 2001), or further back as 11 million years ago when Mindanao was connected to Luzon (Hall 2002).Additional data is needed to confirm this.Use of nuclear markers such as microsatellites and single nucleotide polymorphisms (SNPs) on nuclear loci may also provide a more accurate interpretation about historical gene flow between island populations as well as give definitive answers about the genetic bottleneck events.
As mentioned, the high mitochondrial genetic diversity of raptors, including the Philippine Eagle, could be related to its high longevity.The species has been documented to live over 40 years in captivity, although persecution and habitat loss may have shortened their lifespan in the wild.Long-lived eagles such as the Whitetailed eagle Haliaeetus albicilla (Hailer et al. 2006), White-rumped vulture G. bengalensis (Johnson et al. 2008) and Harpy eagle H. harpyja (Lerner et al. 2009), similarly, have high mitochondrial genetic diversity.Species with a long lifespan can buffer against genetic diversity loss because of a large pool of juveniles that are more resilient to disturbances in the environment than adults (Kuo & Janzen 2004).However, as Johnson et al. (2008) pointed out, long lived species can only buffer against genetic diversity loss if pressures against their population number will last only for short periods of time.Prolonged low population numbers will eventually decrease the species' genetic diversity.
Conservation of threatened but genetically diverse species should include ways to maintain genetic diversity in the populations in the wild.One approach is to focus on protecting and ensuring the survival of wild populations in local nest sites.For example, conservation programs in Central Europe increased the survival of H. albicilla populations by locating and protecting nests.During winter, they also leave uncontaminated carcasses for the eagles to feed on.It is believed that this practice led to the maintenance of the species' genetic diversity (Hailer et al. 2006).
This conservation practice on ensuring the survival of the juveniles can be applied to the Mindanao Philippine Eagle population in order to maintain its degree of genetic diversity.In the conservation of Philippine Eagles in Mindanao, local government and communities play a substantive role in undertaking this conservation practice if nests have been identified within their area.Current conservation efforts for the Philippine Eagle are primarily led by the Philippine Eagle Foundation since its establishment in 1987 through captive breeding and reintroduction to the wild of juveniles and rehabilitated individuals from where they were collected (Salvador & Ibanez 2006).One program of PEF that conforms to the aforementioned conservation practice is the "adopt-a-nest" program where participants are enlisted to monitor and report back to PEF personnel about the breeding status within the nest.As such, this program should be given emphasis by conservationists and the Philippine government.In addition, several laws, such as the Republic Act 9147: Wildlife Resources Conservation and Protection Act, were enacted to protect the remaining populations of the Philippine Eagle as well as to promote its conservation.
It should be clear that although this study observed a high genetic diversity for the remaining population of the Philippine Eagle (with focus on the Mindanao population), this does not exempt this species from the danger of extinction as long as the main threats to this species, such as habitat destruction and hunting, remain.Caution must also be exercised when observing patterns for high genetic diversity in populations that have undergone a recent bottleneck since this could mask high rates of genetic drift (Kuo & Janzen 2004).The main objective of captive breeding programs is for increasing the population of an endangered species for later introduction or reintroduction.It is undisputedly an expensive way to save a species such as the Philippine Eagle, not to mention other criticisms against captive breeding such as increase in inbreeding rate and genetic adaptation to captivity as discussed by Frankham (2008).However, with proper management of breeding individuals (Frankham 2008) and addition of individuals from the wild (Johnson et al. 2008), these problems can be prevented.Nevertheless, success of captive breeding programs can only be realized if captive bred individuals can survive and reproduce in the wild upon release.That is why it is still essential to ensure, for the Philippine Eagle, that a good range of high quality habitat is maintained and prohibiting hunting is enforced.Successes in restoring population numbers have been observed in Bald Eagles Haliaeetus leucocephalus and Peregrine Falcon Falco peregrinus (www.fws.gov).
If Philippine Eagle protection and habitat quality maintenance are properly executed, it will also result in the protection of other species that fall within the protected area (Salvador & Ibanez 2006).Recently, the Philippine government, through the DENR-BMB, has provided additional support to stem the tide of extinction of the Philippine Eagle.More support from other institutions will definitely increase the effectiveness of these conservation efforts and eventually reverse the trend of extinction the Philippine Eagle is facing.
This study demonstrates the current genetic characteristics of the Philippine Eagle population.It is not possible to fully evaluate the genetic differentiation between island populations at this time because of low number of individuals, included for the study, from the islands of Luzon and Samar.Mismatch analysis indicate a genetic bottleneck that could have coincided with the rapid loss of forests in the mid-1900s.The Philippine Eagle showed comparable genetic diversity with other members of the family Accipitridae despite experiencing population decline.However, continued low population numbers may inevitably lead to a decrease in genetic diversity.Within the PEF, additional individuals may be needed in order to maintain the genetic diversity of the Philippine Eagle population within the captive breeding area.As a means to increase or maintain the genetic diversity of population in the wild, conservation efforts should focus on increasing juvenile survival.
For future studies, it is recommended that the sample size of the Luzon and Samar population be increased so that more haplotypes may be identified and can be used to determine the direction of migration of the Philippine Eagle and to be able to estimate genetic differentiation between populations.Information from this will help determine if introduction of individuals from Mindanao to the other islands is advisable.Museum specimens may also be used to increase the number of individuals.Other sensitive markers such as microsatellites and single nucleotide polymorphisms (SNPs) may also be used in order to be able to assess the genetic diversity of the nuclear loci.
Fig. 3.A cluster of Mindanao haplotypes (H1, H2, H4, H5, H6, H9 and H10) form the torso of the network.The network shows that each of the Luzon (H12 and H13) and Samar (H14) haplotypes stem from the torso on different branches, although the Luzon haplotypes are closer to the cluster of Mindanao haplotypes compared to the Samar haplotype.This may indicate that the Samar individual came from a different phylogeographic origin from that of the Luzon individuals.Mismatch analysis was done for the population and is shown in Fig.4.Results for this analysis reveal a multimodal distribution.Test of goodness of fit was assessed using the raggedness index where a nonsignificant value indicates a non-significant difference between the observed and the simulated demographic expansion model.Based on the raggedness index for

Figure 3 .
Figure 3. Median-joining network for the mitochondrial control region haplotypes of the Philippine Eagle.Circles indicate haplotypes and sizes indicate the number of individuals representing that haplotype.Black circle=Mindanao haplotypes, white circle=Luzon haplotypes, light grey circle=Samar haplotype, small dark grey circles=hypothetical median vectors.Numbers between lines denote the number of mutations between adjacent haplotypes.

Figure 4 .
Figure 4. Mismatch distribution analysis for the mitochondrial control region of the Philippine Eagle.

Table 3 . Results for Tajima's D, Fu's F S , and Fu and Li's F and D F of the control region.
Tajima's D Fu's F S test Fu and Li's F Fu and Li's D F