2009 Technical Report #15 Wildlife and Vegetation Surveys of Asuncion Island May 2008

Transcription

1 2009 Technical Report #15 Wildlife and Vegetation Surveys of Asuncion Island May 2008 Conducted by CNMI Division of Fish and Wildlife Department of Lands and Natural Resources Contributing Authors: Laura L. Williams, Paul Radley, Tony Castro, and Scott Vogt Survey Team Laura L. Williams, Paul Radley, Tony Casto, Paul Lisua and Cooper Schraudenbach

2 Table of Contents Introduction... 3 Asuncion Forest Surveys by Laura L. Williams... 6 Survey of Birds on Asuncion by Paul Radley Lizard and Small Mammal Surveys, Asuncion May 2008 by Tony Castro and Laura L. Williams Coconut Crab, Birgus latro Surveys on Asuncion, May 2008 by Scott R. Vogt Front page photo: Asuncion form the south upon approach (Photo: P. Lisua) 2

3 Introduction Asuncion is the third most northern of the fourteen islands of the Mariana archipelago within the political jurisdiction of the Commonwealth of the Northern Mariana Islands (CNMI). The island is located 19º 40 N longitude, 145º 24 East latitude, is 7.3 km 2 in area (Fig. 1), and at 891 m has the second highest elevation in the Marianas (Falanruw et al, 1989). Asuncion is generally vegetated by scrub, forest, fern and low grasses, with lava rubble extending from shore to peak on the south and west sides (Falanruw 1989; Obha 1994). The east side is much steeper and has been described as being mainly lava rubble, ferns and grasses (Falanruw 1989; Ohba 1994). The islands of Uracus, Maug, Asuncion and Guguan have been designated sanctuary islands by the CNMI Constitution (Article IX[2]) and no hunting, habitation, nor introduction of any non-native species is allowed (2 NMIAC ). Landing on a sanctuary islands is only allowed by special permit from the Division of Fish and Asuncion and several other islands north of Saipan are still volcanically active, Anatahan having erupted in 2003 and Pagan in Based on reports from early Spanish explorers Asuncion was thought to have erupted between 1660 and 1786 (Falanruw 1989) and is known to have erupted again in The neighboring island of Guguan also erupted in 1906 (Cruz et al 2000b). Unlike some of the other Northern Islands Asuncion has only been sporadically colonized since the arrival of the Spanish in the late 1600 s and does not have a feral animal population. The Spanish priest Sanvitores reported residents in 1669 (Falanruw 1989). The Japanese were involved to some extent on the island in the late 1930 s and 40 s and small groups of people have inhabited the island part time from the 1960 s on (Sylvario Mateo, Northern Islands Mayors Office pers. comm.). In contrast, Agrihan, Alamagan, and Pagan have maintained villages up through the early 1980 s and still have a few residents year round (Cruz et al 2000c; Cruz et al 2000g). Sarigan was a penal colony during the German administration (Spennemann 1999) and is now considered a sanctuary through the CNMI Division of Fish and Wildlife Regulations (2 NMIAC ). The islands of Agrihan, Pagan, Alamagan and Anatahan all support feral animal populations. There has been relatively little research on the flora and fauna of the northern islands. The floral species of Asuncion was investigated by Fosberg et al in 1975 (1979) and the vegetation was mapped by the U.S Forest Service in the 1980 s (Falanruw 1989). Bird species were recorded by Division of Fish and Wildlife Biologists in 1987 (1-2 June; Reichel et al.) and by a Biological Expedition to the Northern Mariana Islands conducted by the Natural History Museum and Institute from Chiba Japan in 1992 (Asakura and Furuki 1994). In 2000, the CNMI Division of Fish and Wildlife undertook expeditions to several of the northern islands to conduct Variable Circular Plot bird surveys, associated forest surveys, and lizard, rat, and coconut crab 3

4 surveys (Cruz et al. 2000b, Cruz et al. 2000c, Cruz et al 2000d, Cruz et al 2000e, Cruz et al 2000f). However, the planned expedition to Asuncion was cancelled in 2000 due to several tropical storms and typhoons in August of that year. The Division was finally able to conduct bird, lizard, coconut crab and forest surveys on Asuncion from May Figure 1. The Commonwealth of the Northern Mariana Islands 4

5 Literature Cited Asakura, A., and T. Furuki, eds Biological expedition to the Northern Mariana Islands, Micronesia. Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000a. Wildlife and vegetation survey, Aguiguan Technical Report #2. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000b. Wildlife and vegetation survey, Guguan Technical Report #3. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000c. Wildlife and vegetation surveys, Alamagan Technical Report #4. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000d. Wildlife and vegetation surveys, Anatahan Technical Report #5. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000e. Wildlife and vegetation surveys, Sarigan Technical Report #6. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000f. Wildlife and vegetation surveys, Pagan Technical Report #7. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000g. Wildlife and vegetation surveys, Agrihan Technical Report #8. Saipan, CNMI: CNMI Division of Fish and Fosberg, F.R., M.H. Sachet, R. Oliver A geographical checklist of the MicronesianDicotyledonae. Micronesica. 15(1-2). Obha, T Flora and vegetation of the northern Mariana Island. Pages in Biological expedition to the Northern Mariana Islands, Micronesia (A. Asakura and T. Furuki, eds.). Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Spennemann, D.H.R Aurora Australis: The German period in the Mariana Islands Occasional Papers Series Number 5. Division of Historic Preservation, Saipan, CNMI, U.S.A. 5

6 Asuncion Forest Surveys By Laura L. Williams, CNMI-DFW Summary The surveys conducted by the CNMI DFW from May 2008 are the first recorded quantitative forest surveys to be performed on Asuncion. All previous surveys utilized photographs and qualitative methods to describe the plant communities (Falanruw 1989; Ohba 1994). There were fifteen species of trees identified on Asuncion in the areas surveyed. Unfortunately, the endemic Terminalia rostrata was not observed. The native Neisosperma oppositifolia (Fagot) and Pandanus tectorius (Kafu) were the most frequent trees surveyed and had the greatest density. Ficus prolix (Nunu) and T. catappa (Talisai) had the greatest mean basal area (a measure of biomass). This is not surprising as they are much larger trees and therefore might be less dense. The most dominant species in the forest (a measure that combines biomass and density) were T. catappa, C. nucifera and N. oppositifolia. The four species with the greatest overall ecological influence as measured by the importance value ranking for the forest communities surveyed were from greatest to least importance Terminalia catappa, N. oppositifolia, P. tectorius and C. nucifera. The forests of Asuncion are much different than those of the other northern islands previously surveyed (Cruz et al 2000b-g). Asuncion s tree population is highly skewed towards the four dominant species. Other islands have much greater native species richness and all species are fairly common, not just a few species with the greatest frequency, density, and importance (Cruz et al 2000 b-g; Cruz et al 2003). However, on Asuncion the dominant species are several times more abundant than all of the other remaining eleven tree species. Introduction There has been minimal but continuous research on the flora of the remote Northern Mariana Islands. The plant species have been identified and documented for the past 75 years, including Kanehira (Ohba 1994) in 1934, Fosberg, Falanruw and Sachet in 1975 (Fosberg et al, 1977) and Fosberg et al ( 1979, 1982, and 1987). Extensive checklists have been developed for all plant species identified on each of the islands (Fosberg et al 1979, 1982 and 1987). Although Falanruw (1989) mapped the vegetation of plant communities of Asuncion, quantitative assessments were not performed. Ohba (1994) briefly characterizes the forests of each island visited (including Asuncion) during the multi-agency biological surveys sponsored by the Natural History and Museum Institute of Chiba Japan from May through June Quantitative surveys of forest and bird species were undertaken for the first time by the Commonwealth of the Northern Mariana Islands (CNMI) Division of Fish and Wildlife (DFW) in 2000 of all remote or uninhabited islands visited, including Aguiguan, Anatahan, Sarigan, 6

7 Guguan, Alamagan, Pagan, and Agrihan (Cruz et al 2000a-g). However, quantitative forest surveys (of both flora and fauna) were not conducted on Asuncion, Maug or Uracus due to seasonal and funding constraints. It has been the continuing intent of the CNMI-DFW to complete quantitative surveys of each of the northern islands. Quantitative flora and fauna surveys are essential to characterize the structure of the forest, the avian community, and for long term ecological studies. Asuncion offers a unique scenario for comparison to other island as it has had limited ecological disturbance in comparison to the other islands (Falanruw 1989). There has not been extensive and long term human settlements on the island and no feral animals are present (Falanruw 1989; Obha 1994). There has been reports of volcanic activity from 1660, 1786 and 1906 (Falanruw et al; Cruz et al 2000b), which has certainly modified the vegetation. However, several other islands have also had volcanic eruptions, Guguan in 1906, Pagan in the 1980 s, and Anatahan in Other islands are known to have steam vents coming out of the ground (Falanruw 1989; Cruz et al 200b-g; Cruz et al 2003). Forest Survey and Analysis Methods Trees were surveyed using the point-centered quarter method (Mueller-Dombois and Ellenberg 1974) at 21 stations (corresponding to the avian VCP stations) along four transects spaced 150 m apart (Fig. 2, 3a-c). Surveys occurred from May The lateral distance from point to nearest tree in each cardinal direction was measured, species identified, circumference measured and height estimated for all species 1.75m and greater. Canopy cover was determined using a densiometer (Model-C: Forest Densiometer sold through Forestry Suppliers, Inc.) according to the manufacturer s instructions and ground cover was estimated using the line-intercept method (Mueller-Dombois and Ellenberg 1974). Forest data from all transects were combined and analyzed for tree density, absolute density, absolute frequency, basal area, dominance, importance value, average height and percent canopy and ground cover (Mueller-Dombois and Ellenberg 1974). Absolute frequency was calculated as the number of stations in which a species occurs divided by the total number of stations multiplied by 100. The absolute density (AD) was calculated for the total number of trees supported in a 100m 2 or 1 hectare (ha) area and for each individual species within the area. Absolute density of all tree species in one hectare was calculated with the formula AD = area/d 2, where D = the mean distance of all distances to all trees at all stations. In order to determine the density of a specific tree species, the ratio of the number of quarters in which the species occurred to the total number of quarters was calculated. That ratio was then multiplied by the overall absolute density of trees to determine the density of each species per 100 m 2. Basal area (BA) was calculated with the formula BA = (DBH 2 ), where DBH is the diameter at breast height (1.4m) of each tree measured. Dominance of each species incorporates the average basal area of each species for all individuals surveyed times the species density in one hectare (as calculated above). Consequently, the dominance measure reflects the interaction of the biomass and density of a species, which yields the relative measure of the species ecological influence. However, The Importance Value (IV) is designed to rank the overall influence of species in a given community or ecosystem by incorporating dominance, density, and frequency (measure of percent distribution). The IV was determined by taking the sum of relative dominance, relative density 7

8 and relative frequency. Relative values were derived from the ratio of individual species to the sum of all species (Mueller-Dombois and Ellenberg 1974). Figure 2. Map of Asuncion with location of bird and forest transects and stations surveyed on Asuncion from May

9 Figure 3. Detailed map of bird and forest transects and stations surveyed on Asuncion from May Transect 2 Figure 3a. Location of camp for DFW expedition May The approximate location of the end of Transect 2 is delineated with an arrow and the lava flows near the peak are all brown. 9

10 Forest Survey Results Ten tree species were detected at the stations surveyed and an additional five species were observed opportunistically while on the island (Table 1). The species of tropical almond reported to be endemic to Asuncion Terminalia rostrata (Falanruw 1989) was not observed. Pandanus tectorius and Neisosperma oppositifolia had equal percent detections (31%) at stations along transects 1 and 2 (n=48) (Fig. 3). In contract Pandanus tectorius occurred in only 14% of the detections along the coastal transects (3&4). Along the coastal transects, 3 & 4 N.oppositifolia and Cocos nucifera had the greatest percent detections (30% and 28%), respectively (Fig. 4). The greatest species richness was found in the upland transects (9 diverse species) versus the coastal forest transects (6 species). Figure 3b. Landing site on Asuncion south of the camp. The forest paralleling the beach is the location of survey transect 3. 10

11 Figure 3. The percent detection of tree species at stations along upland transects 1 and 2 (n=48) on Asuncion Island, May Figure 4. The percent detection of tree species at stations along coastal transects 3 and 4 (n=36) on Asuncion Island, May

12 Two native species, N. oppositifolia and P. tectorius, had the greatest absolute frequency (i.e., detected at the most number of stations) being present at 57.14% of all stations. Terminalia catappa had the third greatest absolute frequency of 38.10% and Morinda citrifolia and Cocos nucifera absolute frequencies were 23.81% (Fig. 5). The remaining six species had very low absolute frequencies and were minimally distributed throughout the survey area. Figure 5. The absolute frequency % of ten tree species surveyed on four transects (n=82 stations) in Asuncion island May Figure 3c. Conducting a point center-quarter tree survey. 12

13 Table 1. The fifteen tree species observed on Asuncion in May 2008; ten were recorded at stations and five were observed opportunistically along transects. Species Status 1 Observed 2 Comment Carica papaya I O Cocos nucifera N S Agricultural Introduction; native to tropical America Erythrina variegata var. orientalis E S Variety is endemic Ficus prolixa N S Hibiscus tiliaceus N S Melanolepis multiglandulosa var. multiglandulosa E S Variety is endemic Morinda citrifolia N S Neisosperma oppositifolia E S Pandanus tectorius N S Pipturus argenteus N O Pisonia grandis N O Pithecellobium dulce I O Premna obtusifolia N S Agricultural introduction; native to S. American Terminalia catappa N S Trema orientalis I O Ornamental; Native to tropical Asia 1. E-Endemic; N-Native; I-Introduced 2. S-At a survey station ; O-Opportunistically Neisosperma oppositifolia and P. tectorius also had the greatest density in the areas surveyed at 7.21/ha and 5.54/ha, respectively (Table 2). Cocos nucifera and T. catappa had densities of 3.6 /ha and 3.33 /ha, respectively. All other tree species had less than a third of these densities. Ficus prolix and T. catappa had the greatest mean basal area with an average of 5, cm 2 /ha and 5, cm 2 /ha, respectively (Table 2). Despite the low density of Terminala catappa (3.33/ha) the high basal area resulted in a dominance value of T. catappa (17, cm 2 /ha) throughout the survey area, a result which is three times that of both C. nucifera (5, cm 2 /ha) and N. oppositifolia (5, cm 2 /ha) (Table 2). The four species with the greatest importance value for the forest communities surveyed were from greatest importance Terminalia catappa, N. oppositifolia, P. tectorius and C. nucifera (Table 3). Terminalia catappa appeared to be a favorite tree for fruitbat foraging in the evenings (based upon opportunistic observations) and lone individuals were found hanging from several. T. catappa during the daytime along transects (Fig. 6). 13

14 Table 2. The average basal area, density and dominance of tree species surveyed on four transects on Asuncion in May Species Avg. BA (cm2) SD Avg. BA Density #/ha Dominance cm 2 /ha C. nucifera E. variegata var orientalis F. prolixa H. tiliaceus M. multiglandulosa M. citrifolia N. oppositifolia P. tectorius P. obtusifolia T. catappa Figure 6. Mariana fruit bat (Pteropus mariannus) in a T. catappa tree on Asuncion Island, May

15 The highest importance value (index which incorporates dominance, density, and frequency) from greatest importance to least was Terminalia catappa (75.27%), N. oppositifolia (68.77%), P. tectorius (61.54%) and C. nucifera (39.94%), from the most important to the least, respectively (Table 3). Table 3. The Importance Value (IV) and the relative dominance (Rel. Dom), density (Rel. Den) and frequency (Rel. freq.) that the IV is derived from for tree species surveyed on four transects on Asuncion in May Species Rel. Dom (%) Rel. Den. (%) Rel. Freq (%) IV* Terminalia catappa Neisosperm oppositifolia Pandanus tectorius Cocos nucifera Morinda citrifolia Erythrina variegata var. orientalis Ficus prolixa Premna obtusifolia Melanolepis multiglandulosa Hibiscus tiliaceus *Importance Value the forest species ranking that is the sum of the characteristic of frequency, density and dominance (dominance is a value derived from density and basal area). Canopy and Ground Cover The overall mean canopy height for all transects was 4.92 m ±2.22m (Fig. 7). The average canopy height of the coastal forest is slightly greater 5.18m ±2.25m (transects 3 and 4) than the average inland forest which were 4.73m ±2.20m (Transects 1 and 2) (Fig. 7). The mean canopy cover for all areas surveyed was % ±13.09%. The average coastal forest canopy cover was higher at 91.22% ±6.65 and the average inland forest canopy was 87.93% ±19.87%. The average overall forest ground cover was 18.08% ±0.10%. The coastal forest average ground cover was less than the average inland forest ground cover, which was 13.33% ±32.79% and 21.82% ±28.57%, respectively (Fig. 8). 15

16 Figure 7. The average canopy height of all four transects surveyed (Total), the coastal transects (transects 3&4) and the upland forest transects (1&2) on Asuncion in May Figure 8. The mean percent canopy and ground cover for all transects surveyed (Total), coastal forest transects (3&4), and upland forest transects (1&2) on Asuncion in May

17 Discussion The overall forest (all transects) in the survey area is a T. catappa/ P. tectorius/ N.oppositifolia complex with an average canopy height of 5 m, 90% canopy cover and 19% ground cover. The importance value ranking of the three most important species is very close and therefore it is hard to define the forest community by a single or paired species complex (Table 3). Terminala catappa has a greater influence only by a small margin, which is clearly due to is very large over all biomass (as measured by basal area) compared to the other species (Table 2). There are some subtle difference between the coast and upland forests (transects 3&4 and 1&2, respectively). This study did not analyze the coastal and upland forests for all measures separately (i.e. frequency, density, basal area, dominance, and importance value); however the data do allow for some distinct observations. The coastal forest surveyed (transects 3 & 4) can be described as C.nucifera/N.oppositifolia/T. catappa forest complex. The percent detections were 28%, 30% and 17%, respectively (Fig. 3). The percent detections of C.nucifera in the coastal forest areas (28%) was much higher than the upland forest (transects 1 & 2) where it comprised only 7% of trees surveyed (Fig. 3 & 4). The influence of P. tectorius was much less along the coast, with 14% detection as opposed to 31% detection in the upland forest. The canopy height was variable with a mean of 5 m (±2.25m) reaching upwards to 7.5m (Fig. 7). The canopy cover was tightly clustered at a mean of 90% with very variable ground cover ranging from 10-50% (Fig. 8) The upland forest (transects 1 & 2), in terms of percentage of detections, could be described as a P. tectorius/n. oppositifolia forest complex. The percent detections were 31% for both species (Fig. 3 & 4). Terminala catappa is present in the upland forest but with a detection rate almost a third less at 13% (Fig. 3) and in the coastal forest it is (transects 3 & 4) 17% (Fig. 4). However, the dominance and importance of T. catappa are clear by its overall biomass as measured by the basal area (Table 2), which is an important influence not observable through density and frequency calculations. The species richness is greater in the upland forests, with nine species detected compared to six along the coastal transects (Fig. 3&4). Falanruw (1989) characterizes the forest between transects 1 and 2 as Terminalia forest. However, our survey data suggest the forest is more complex. The upland forest area is clearly also dominated by P. tectorius and N. oppositifolia, especially in the understory. It may be that Falanruw (1989) detected the T. catappa more readily as she primarily used aerial photos to develop her vegetation maps and forest community descriptions. Terminala catappa is generally a much taller tree than either N. oppositifolia or P. tectorius (Stone 1970) and it also has a distinct flametree canopy shape. This shape and height could mask the actual abundance of the other species when only using aerial photos. The other influential but less distinct species can be missed when quantitative studies were not performed and used to develop the forest descriptions. Falanruw (1989) did at some point conduct ground truthing, as the report contains pictures of the elusive endemic T. rostrata and contains reports of forest along the hillside that does not contain dense P. tectorius understory. The only areas found during this expedition that did not contain dense P. tectorius understory were in some areas along the coast. Ohba (1994) explored an area of Asuncion that began south of the present study s research area and extended up to lava fields (Fig. 9). Ohba (1994) describes the area he explored as consisting of a P.tectorius/ T. catappa association along the slopes and C. nucifera along the coast. 17

18 There were 10 tree species detected at survey stations on Asuncion which is less species richness than on other islands in the archipelago (Cruz et al 2000 b-g; Cruz et al 2003). On Agrihan, the island immediately south of Asuncion, 18 tree species were detected. Agrihan has a small feral animal population but they also have year round residents (Cruz et al 2000g). Guguan, also a sanctuary island without feral animals had 15 different tree species detected along similar transects (Cruz et al 2000b). On Alamagan, 23 tree species were detected. Feral animals are present however, the forest destruction does not seem to be as great as on the other islands with feral animals (Cruz et al 2000c). Despite destruction of the forest understory on (Cruz et al 2000f) 20 tree species were detected. There were 15 tree species found on Sarigan which was been devastated by feral animal destruction until their eradication in 1997 (Martin et al 2008). There were 16 species found on the southern side of Anatahan (Cruz et al 2003), which was also subjected to intense feral animal destruction. There were notable native species absent on Asuncion such as Aglaia marriannensis, Psychotria mariana, Ficus tinctoria and Pouteria obovata which are common on the other islands (Cruz et al 2000b-g). Despite a long absence of human settlements compared to other islands (Pagan, Alamagan, Sarigan and Anatahan) (Cruz et al 2000b-g) there is obvious signs humans have been present as noted by Falanruw and Fosberg (1989; 1979). The pattern of coconut forest is likely from cultivation and the presence of the non-native Pithecellobium dulce located at the landing location (Fig. 3b) indicates human influence. Falanruw (1989) describes what looked like abandoned farm plots in aerial photos in the 1970 s. All authors do agree however that the influence was minor (Falanruw 1989, Ohba 1994). The lack of species richness and the skewed abundance and dominance of the tree species present is unique to Asuncion. It has not been influenced by feral animal damage, nor is it likely due to the minimal human residence as other islands have been impacted to a much greater extent than Asuncion and have greater species richness. Nor is the lack of diversity likely size related. Asuncion is approximately 7.9 km 2, almost twice the size of Guguan and larger than Sarigan. An obvious distinction is Asuncion is either too far away for species to have arrived and colonized or further investigation in other areas is required. However, the sample size and methods were fairly consistent between islands surveyed. Figure 9. Approximate location of Ohba (1994) survey transect conducted on June 1, 1992 as part of the collaborative survey effort between the Natural History Museum and institute (Chiba, Japan) the CNMI DLNR and the University of Guam Marine Lab. 18

19 Acknowledgments We owe much to Captain Kelly Tenorio of the Super Emerald for delivering us safely to and from Asuncion. It was a comfort to see the Super Emerald in the harbor while we worked, especially during the storm. Our expedition physician, Cooper Schraudenbach not only was well supplied but he turned out to be great with the machete and coconut crab surveys! Patrick Santos of the Northern Islands Mayors Office is a great all around outdoorsman and it is always good to work with him! To the DFW Staff Paul Radley, Paul Lisa and Tony Castro you did a great job and everyone worked very hard. Gayle Martin was a great help in all areas of organization before, during and after the surveys and assisted with bird surveys when it needed. This expedition was funded by the State Wildlife Grant and the Wildlife Restoration Grant from the USFWS. Literature Cited Asakura, A., and T. Furuki, eds Biological expedition to the Northern Mariana Islands, Micronesia. Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000a. Wildlife and vegetation survey, Aguiguan Technical Report #2. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000b. Wildlife and vegetation survey, Guguan Technical Report #3. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000c. Wildlife and vegetation surveys, Alamagan Technical Report #4. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000d. Wildlife and vegetation surveys, Anatahan Technical Report #5. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000e. Wildlife and vegetation surveys, Sarigan Technical Report #6. Saipan, CNMI: CNMI Division of Fish and 19

20 Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000f. Wildlife and vegetation surveys, Pagan Technical Report #7. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000g. Wildlife and vegetation surveys, Agrihan Technical Report #8. Saipan, CNMI: CNMI Division of Fish and Cruz, J. L. Williams, N. Hawley, J. Esselstyn, C. Kessler and O. Bourquin Wildlife and vegetation surveys and feral animal control on Anatahan Technical Report #10. Saipan, CNMI: CNMI Division of Fish and Fosberg, F.R., M.H. Sachet, R. Oliver A geographical checklist of the MicronesianDicotyledonae. Micronesica. 15(1-2). Martin, G., L. Williams, J.B.Cruz, N.Hawley, S.Vogt, B. Smith, O. Bourquin, S. Kremer, C. Kessler Wildlife and Vegetation Surveys of Sarigan Island, April Technical Report #14. Saipan, CNMI: CNMI Division of Fish and Obha, T Flora and vegetation of the northern Mariana Island. Pages inbiological expedition to the Northern Mariana Islands, Micronesia (A. Asakura and T.Furuki, eds.). Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Spennemann, D. H. R Aurora Australis: The German period in the Mariana Islands Occasional Papers Series Number 5. Division of Historic Preservation, Saipan, CNMI, U.S.A. Stone, B.C The flora of Guam. Micronesica6:

21 Survey of Birds on Asuncion By Paul Radley, CNMI DFW Summary Between 20 and 23 May 2008, CNMI DFW staff conducted Variable Circular Plot (VCP) surveys for birds along four transects established in 316 ha of forest cover on the Northern Mariana Island of Asuncion; 359 birds of eight species were detected. Detection data for four species of forest landbird encountered on Asuncion were analyzed in the program DISTANCE 5.0 to calculate density and abundance estimates. Densities of each, from greatest to least, are: Micronesian Starling (Aplonis opaca; 11.8 birds/ha, 95% CI ); Micronesian Honeyeater (Myzomela rubratra; 7.1 birds/ha, 95% CI ); Micronesian Megapode (Megapodius laperouse; 5.6 birds/ha, 95% CI ); Collared Kingfisher (Todiramphus chloris; 1.7 birds/ha; ). Although two past surveys listed the White-throated Ground Dove (Gallicolumba xanthonura) as uncommon or rare on Asuncion, none were detected on island in Megapodes, however, were quite common and detected in several cover types at various elevations and juveniles of the species were observed on two occasions. Introduction Asuncion is both the second highest island of the Northern Marianas and the second highest peak in Micronesia after Agrihan. This relatively young, symmetrically coned stratovolcano rises steeply to 891 meters and comprises a land area of 730 ha (Falanruw 1989, Berger et al. 2005). Falanruw (1989) mapped vegetation types of the island and reported 12% as barren areas (landslides, bluffs, lava flow and the caldera resulting from most recent eruption in 1906), 45% as sparse or low growth, and 43% as thickets, forests, and coconut. Total forest cover for Asuncion is reported by Falanruw (1989) as 316 ha, a figure that was used to calculate the forest landbird density and abundance reported here. In May 2008, CNMI DFW performed the first quantitative, systematic survey of avian populations on Asuncion. One previous research expedition was organized and carried out by DFW in 1987, which focused primarily on seabirds and Micronesian Megapode nesting in the northern island; the researchers were on Asuncion for only two days (1-2 June; Reichel et al. 1987). Another expedition was conducted for two days (1 and 7 June) in 1992 by the Natural History Museum and Institute, Chiba, Japan (Rice and Stinson 1992, Asakura and Furuki 1994). However, in both cases, as workers were on Asuncion for only short durations and no explicit details are presented concerning avian surveys carried out the on island, resultant data can only be considered as a record of species presence (Reichel et al. 1987, Rice and Stinson 1992, Asakura et al. 1994, Stinson 1994). 21

22 Avian Survey and Analysis Methods From May 2008, all species of birds detected while in route to and from Asuncion and while on island were recorded. Between 20 and 23 May, the CNMI DFW conducted Variable Circular Plot (VCP) or point-transect surveys in primarily forested habitat on Asuncion. Count stations (n = 20) were spaced at 100 m intervals along four transects; transects #3 and #4 each consisted of four stations and transects #1 and #2 six stations each (Fig.1). Surveys, which followed standard methods for point-transect distance sampling (Engbring et al. 1986), were conducted between 6:00 a.m. and 10:30 a.m. with count durations of 5-minutes per station. Weather conditions (rain, wind, and cloud cover) were documented and the estimated horizontal distance to all birds detected (visual and/or auditory) was recorded. Taped playbacks were not employed and the majority of birds were counted by song or call note. Following Falanruw (1989), total forested area (including native, coconut, and coastal thicket) was assumed to be 316 ha. As surveys were conducted primarily in forested habitat, this area figure was used to estimate avian population densities and abundance in program DISTANCE 5.0 (Thomas et al. 2006). Transects #1 and #2 were primarily comprised of native forest, transect #3 coconut forest, and transect #4 coconut forest with coastal thicket. Although station #6 on transect #2 was not entirely in forest cover, it was included in analysis as it was within primarily forested edge habitat. Density and abundance were estimated for breeding pairs of four species of forest bird native to Asuncion; Micronesian Megapode (Megapodius laperouse), Collard Kingfisher (Todiramphus chloris), Micronesian Honeyeater (Myzomela rubratra), and Micronesian Starling (Aplonis opaca). Numbers detected on Asuncion for each species were too low (< 80 detections) to directly calculate their estimates in DISTANCE. Thus, detection distances for these species were pooled with those from past VCP surveys on other islands in the Northern Mariana Islands, then post-stratified by island and year to estimate species density (breeding pairs per ha) and abundance (pairs per 316 ha of forest cover on Asuncion). To maximize model fit as necessary, data were right truncated (Buckland et al. 2001; Table 1) for all four species considered and additionally left truncated for two (Collard Kingfisher and Micronesian Starling). In the end, the model yielding the lowest AIC was used to select the most appropriate detection function for a given data set. All models incorporated half-normal detection functions (Table 1) and variances and confidence intervals were derived for all but one species (Collared Kingfisher) by bootstrap methods (within multiple covariates distance sampling) in DISTANCE from 999 iterations; the data for Collared Kingfisher enabled variance and intervals for the species to be determined only by conventional distance sampling (Thomas et al. 2006). 22

23 Figure 1. Survey transects 1-4 on the west side of Asuncion in which bird and forest surveys were conducted May Avian surveys (n = 12) that provided supplementary detection data for pooling with those from Asuncion included past results from Agrihan (Cruz et al. 2000g), Aguiguan (Cruz et al. 2000a, Esselstyn et al. 2003), Alamagan (Cruz et al. 2000c), Anatahan (Cruz et al 2000d, Cruz et al 2003a), the Bird Island Conservation Area, Saipan (Cruz et al. 2003b), Guguan (Cruz et al. 2000b), Pagan (Cruz et al. 1999, Cruz et al. 2000f), and Sarigan (Cruz et al. 2000e, Morton et al. 2000). For the purpose of analyses the assumption was made that all variables (e.g., forest type, climate, day length, etc) possibly affecting avian populations were the same (or as closely as possible) across the islands and areas taken into consideration. As the majority of stations on Asuncion were counted by P. Radley (transect #1 was counted by G. Martin, CNMI DFW, 23

24 because of rain and time constraints) and observer data were not available for any past surveys, differences in observed detection distances were not taken into consideration during analysis. Table 1. Model parameters and detection probabilities used to estimate population densities of forest landbirds detected on Asuncion, May Species Truncation Model + series expansion Micronesian Megapode 90.0 (R) Half-normal + cosine Collared Kingfisher 99.0 (R), 9.0 (L) Half-normal + cosine Micronesian Honeyeater 92.0 (R) Half-normal + cosine Micronesian Starling 90.0 (R), 5.9 (L) Half-normal + cosine Survey Results A total of 16 species of birds were detected while in route or on Asuncion from May 2008 (Table. 2). The vast majority of birds detected while on-island (18 24 May) were seabirds, large masses of which (primarily boobies, shearwaters, and Brown Noddies) were observed out to sea in the evening flying north, likely en-route to Maug or Uracas. On the evening of 22 May, several Wedge-tailed Shearwaters were heard calling not far from camp, providing evidence of breeding by the species on the island (Stinson [1994] listed this species as common on Asuncion). Additional species observed by boat en-route to Asuncion from Saipan were Matsudaira s Storm-Petrel (Oceanodroma matsudairae) and Spectacled Tern (Sterna lunata). By far the most numerous species detected along transects during VCP surveys was the White Tern (n = 123) followed by the Black Noddy (n = 76; Table 3); interestingly, Stinson (1994) listed the latter species as uncommon on Asuncion. Although seabirds that spend the majority of their time foraging in near-shore waters, both species nest in trees in forested areas. However, as these species (along with the Sooty Tern and Brown Noddy, both of which were also detected during surveys [Table 3]) are not technically forest landbirds they were not included in DISTANCE analyses. Micronesian Starlings (n = 57) were the most numerous forest landbird detected along transects, followed by the Micronesian Megapode (n = 35), both of which were detected more frequently on transects traversing native forest (starlings were highest along transects #1 and #2, Megapodes along transect #2; Table 3). The least numerous species detected was the Collared Kingfisher (n = 17) and the White-throated Ground Dove (Gallicolumba xanthonura), which Reichel et al (1987) and Stinson (1994) listed as rare or uncommon on the island, was not detected at all. 24

25 Table 2. All avian species (n = 16 from 10 families) detected at Asuncion, May 2008 (taxonomy, sequence, and nomenclature follow Wiles 2005). Common Name Micronesian Megapode Wedge-tailed Shearwater Red-tailed Tropicbird Masked Booby Brown Boody Red-footed Booby Great Frigatebird Intermediate Egret Pacific Reef-Heron Sooty Tern Brown Noddy Black Noddy White Tern Collared Kingfisher Micronesian Honeyeater Micronesian Starling Scientific Name Megapodius laperouse Puffinus pacificus Phaethon rubricauda Sula dactylatra Sula leucogaster Sula sula Fregata minor Egretta intermedia Egretta Sacra Sterna fuscata Anous stolidus Anous minutes Gygis alba Todiramphus chloris Myzomela rubratra Aplonis opaca Table 3. Totals for all birds detected over four transects on Asuncion, May Transects #1 and #2 were native forest, transect #3 coconut forest, and transect #4 coconut forest with coastal thicket. (Percent occurrence is the number of stations occupied divided by the number of stations sampled [n = 20], multiplied by 100.) Species Total Det. by Transect #1 #2 #3 #4 Total Det. No./ Trans. No./ Stat. % Occur. Micronesian Megapode Sooty Tern Brown Noddy Black Noddy White Tern Collared Kingfisher Micronesian Honeyeater Micronesian Starling Total

26 Population Density and Abundance As a result of pooling detection data from Asuncion with those of past surveys on other Northern Islands, density estimates were produced for all four species encountered on the island (Table 4); detections for all four species alone were too low to estimate densities in DISTANCE. The Micronesian Starling had highest density for forested habitat on Asuncion, followed by the Micronesian Honeyeater and the Micronesian Megapode; the lowest density by far was exhibited by the Collared Kingfisher (Table 4). Table 4. Density estimates (birds/ha) and 95% confidence intervals (Lower and Upper 95% CI) for each species detected on Asuncion, May Included are comparisons of these species across four other Northern Mariana Islands (listed from north to south). (Per hectare density estimates for Saipan were determined from Camp et al. [in press].) Species Micronesian Megapode Collard Kingfisher Micronesian Honeyeater Micronesian Starling Island Density Estimate SE Lower 95% CI Upper 95% CI Asuncion Agrihan Sarigan Sarigan Saipan Aguiguan Asuncion Agrihan Sarigan Sarigan Saipan Aguiguan Asuncion Agrihan Sarigan Sarigan Saipan Aguiguan Asuncion Agrihan Sarigan Sarigan Saipan Aguiguan

27 Species abundance is simply a function of density (Table 4) multiplied by the total area of suitable habitat or cover (i.e., 316 ha of forest cover on Asuncion). Assuming that both sexes do not sing, call, or otherwise vocalize, abundance estimates for Collared Kingfishers ( individuals [mean = 525]), Micronesian Honeyeaters ( [mean = 2253]), and Micronesian Starlings ( [mean = 3732]) could potentially be interpreted as a rough estimate of the number of breeding pairs of each on Asuncion (assuming only singing males of each species were counted and that each male is paired). However, there are problems with this approach; (1) birds recorded by sight only during surveys were not (and in some cases could not) be differentiated by sex, (2) the detection type (i.e., heard or seen) by species for each detection recorded is unknown, and (3) no discrimination was made while recording sound detections between song or call (in all three species, both sexes do at least call). Although Micronesian Megapodes do produce sexually dimorphic calls (often given in a duet; U.S. Fish and Wildlife 1998) no discernment was made between the genders of vocalizations while recording detections for the species. Thus, the individuals (mean = 1757) estimated for the island should not be considered an indication of the number of breeding pairs but rather the mean of the total adult population on island. Discussion As explained previously, to compensate for the small sample-size of detections acquired per species on Asuncion, it was necessary to augment these data by pooling them with distance data from other islands for the purpose of analysis. When analyzing post-stratified data in DISTANCE, the program produced separate density estimates for all the islands taken into account. Included in Table 4 are density estimates for four islands in the Mariana Archipelago of the four species of forest landbird detected on Asuncion. Although detection data from a maximum of 12 surveys on eight islands were pooled with those from Asuncion for the purpose of analysis, only four surveys on three islands (those on Agrihan, Sarigan, and Aguiguan) were used consistently in DISTANCE in an un-modified format (Table 4). (The per hectare density estimates for Saipan included in Table 4 were calculated from Camp et al. [in press]). Densities of Micronesian Megapodes (5.56/ha), Collard Kingfishers (1.66/ha), and Micronesian Starlings (11.81/ha) are higher on Asuncion than the four other islands compared in Table 4 (detections of megapodes on Saipan were too low for DISTANCE generated density estimates for that island). However, the density of Micronesian Honeyeaters (7.13/ha) on Asuncion is higher only than that for Saipan; the greatest density of this species occurs on Agrihan (22.39/ha), the next island south of Asuncion (Table 4). In the case of megapodes, kingfishers, and starlings, higher densities may be associated with the relatively low level of human presence and human introduced variables on Asuncion, such as feral animals and nonnative predators, and domestic livestock. The positive effects of lower human presence may be especially relevant in regards to the megapode. Apart from its distant proximity to the nearest center of human population (i.e., Saipan), Asuncion, like Sarigan, is designated as a CNMI Wildlife Sanctuary Island and contemporary human presence should by definition be relatively low. Both of these islands sustain the highest densities of megapodes amongst islands compared in Table 4, a trend that is not evident for kingfishers and starlings. 27

28 Lower densities of Micronesian Honeyeaters on Asuncion compared with other northern islands in Table 4 (i.e., Agrihan and Sarigan) may primarily be related to the timing of their colonization of the island. Table 4 indicates that honeyeater densities increase in a northward direction along the archipelago. However, the species may have more recently arrived on Asuncion than the other three species of forest landbird detected during surveys there. In further support of this hypothesis, honeyeaters are listed by Stinson (1994) as rare on Maug and as not occurring on Uracas, the next islands north of Asuncion, respectively. Additionally, it is possible that the larger megapodes, kingfishers, and starlings are better at dispersing relatively long distances over water than the much smaller honeyeaters. Although there are no data to support this assertion for the other species, there is evidence that megapodes are relatively strong dispersers (Pratt et al. 1980, U.S. Fish and Wildlife Service 1998). Micronesian Megapodes Both total detections (Table 3) and DISTANCE generated density estimates (Table 4) for the Micronesian Megapode suggest that the species is relatively common on Asuncion. In fact, when compared to other islands in the archipelago, the estimated densities of the species on Asuncion are relatively high (Table 4). Megapodes were detected in all cover types surveyed and at various elevations on the island. These results are not consistent with those of past surveys of the island; however, these were few and of short duration, and focused primarily on seabirds (Reichel et al. 1987, Rice and Stinson 1992, Stinson 1994). In 1987, during a CNMI DFW survey of on 1 and 2 June, Reichel et al. (1987) saw one megapode and heard another low on Asuncion s southwest slope. During the 1992 research expedition to the Northern Islands by the Natural History Museum and Institute, Chiba, Japan, only one megapode was heard at lower elevation on the south side of the island (Rice and Stinson 1992, Stinson 1994). These results lead past researchers to conclude that megapodes were rare or uncommon on the island (Reichel et al. 1987, Stinson 1994), with a total population estimate of less than 25 individuals (U.S. Fish and Wildlife Service 1998). Serving as indisputable evidence of breeding by megapodes on the island, juveniles (Fig. 2) were seen on Asuncion on two separate occasions in On 19 May, after completing the cutting of transect #2, P. Radley and C. Schraudenbach climbed to the bottom (at ~430 m elevation) of an exposed pumice field (19 41 N, E; Fig. 3a and 3b) on the west side of the island while attempting to summit the volcano. During the fairly steep descent from the bottom of this field, three juvenile megapodes were flushed from an area of low fern and scrub vegetation (Fig. 4) growing from course pumice substrate between 430 m and 275 m elevation. Several adult megapodes were heard calling from clumps of low trees and shrubby vegetation widely dispersed throughout this area. This broad area of fern layered pumice on the west slope of the island above forested habitat may serve as nesting grounds for megapodes on Asuncion. 28

29 Figure 2. Juvenile Micronesian Megapode on Asuncion, 20 May 2008 (photo: P. Lisua). Figure 3a. Lower edge of pumice field (~430 m elevation) facing down-slope on Asuncion, 19 May 2008 (photo: P. Radley). Juvenile Megapodes were flushed from area of low vegetation and ferns below pumice field. 29

30 Figure 3b. From lower edge of pumice field (~430 m elevation) facing up-slope towards the summit of Asuncion, 19 May 2008 (photo: P. Radley). Figure 4. Low ferns and vegetation above forest edge similar to, and in the vicinity of, areas of cover from which juvenile and adult Megapodes detected, Asuncion, 19 May 2008 (photo: C. Schraudenback). 30

31 Another juvenile was observed and photographed (Fig. 2) in primarily native vegetation on the afternoon of 20 May as G. Martin and P. Lisua set traps for coconut crab surveys along transect #1. The bird evidently flushed out of low vegetation and flew across the transect in front of the crew, landing a short distance away where it was photographed (Fig. 2). Acknowledgments Expedition physician, Cooper Schraudenbach, assisted with cutting transects and setting stations for VCP surveys. Gayle Martin (Alaska Department of Game and Fish, Anchorage) assisted in the field with bird surveys when needed. Tina de Cruz provided data from past DFW and USFWS surveys of the Northern Islands for pooling with Asuncion data, and, along with Fred Amidon (U.S. Fish and Wildlife Service, Honolulu, Hawaii) provided comment on the draft manuscript. Rick Camp (University of Hawaii, Honolulu) provided invaluable methodological insight and assisted greatly with all DISTANCE analyses. Literature Cited Asakura, A., and T. Furuki, eds Biological expedition to the Northern Mariana Islands, Micronesia. Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Asakura, A., T. Ohba, S. Miyano, T. Furuki, T. Kurozumi, and H. Harada Outline of the biological expedition to the Northern Mariana Islands, Micronesia. Pages 1-11 in Biological expedition to the Northern Mariana Islands, Micronesia (A. Asakura and T. Furuki, eds.). Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Berger, G.M., J. Gourley, and G. Schroer Comprehensive Wildlife Conservation Strategy for the Commonwealth of the Northern Mariana Islands. Unpublished report to the National Advisory Acceptance Team, U.S. Fish and Wildlife Service, Portland, OR. 358 pp. Buckland, S.T., D.R. Anderson, K.P. Burnham, J.L. Laake, D.L. Borchers, and L. Thomas Introduction to Distance Sampling: Estimating abundance of biological populations. Oxford: Oxford University Press. Camp, R.J., T.K. Pratt, A.P. Marshall, F.A. Amidon, and L.L. Williams. (in press) Recent status and trends of the land bird avifauna on Saipan, Mariana Islands, with emphasis on the endangered Nightingale Reed-Warbler Acrocephalus luscinia. Bird Conservation International 31

32 Cruz, J., L. Arriola, N, Johnson, S. Vogt, K. Miller, B. Camacho, S. Rasa, and S. Camacho Annual Report Fiscal Year 1999; Federal Aid in Wildlife Restoration Program. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000a. Wildlife and vegetation survey, Aguiguan Technical Report #2. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000b. Wildlife and vegetation survey, Guguan Technical Report #3. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000c. Wildlife and vegetation surveys, Alamagan Technical Report #4. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000d. Wildlife and vegetation surveys, Anatahan Technical Report #5. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000e. Wildlife and vegetation surveys, Sarigan Technical Report #6. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000f. Wildlife and vegetation surveys, Pagan Technical Report #7. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000g. Wildlife and vegetation surveys, Agrihan Technical Report #8. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Williams, N. Hawley, J. Esselstyn, C. Kessler, and O. Bourquin. 2003a. CNMI wildlife and vegetation surveys and feral animal control, Anatahan Technical Report #10. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Williams, N, Johnson, V. Camacho, and J. Salas. 2003b. Baseline avian and vegetation surveys, Bird Island Wildlife Preservation Area Technical Report #11. Saipan, CNMI: Division of Fish and Engbring, J., F.L. Ramsey, and V.J. Wildman Micronesian forest bird surveys, 1982: Saipan, Tinian, Aguijan, and Rota. U.S. Fish and Wildlife Service Report. 32

33 Esselstyn, J., J. Cruz, L. Williams, and N. Hawley Wildlife and vegetation surveys, Aguiguan Technical Report #9. Saipan, CNMI: CNMI Division of Fish and Falanruw, M.V.C Vegetation of Asuncion: A Volcanic Northern Mariana Island. Resour. Bull. PSW-28. Berkeley, CA: Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Dept. of Agriculture. Morton, J.M., M. Lusk, G. Hughes, G.J. Wiles Occurrence and density of avifauna and Mariana fruit bats on Sarigan, Commonwealth of the Northern Mariana Islands, in July USFWS, Honolulu, Hawaii. Pratt, H.D., J. Engbring, P.L. Bruner, and D.G. Berrett Notes on the taxonomy, natural history, and status of the resident birds of Palau. Condor 82: Reichel, J.D., S. Taisacan, S.C. Villagomez, P.O. Glass, and D.T. Aldan Field Trip Report: Northern Islands, 27 May 5 June. CNMI Division of Fish and Wildlife, Saipan. Rice, C.G., D.W. Stinson Field Trip Report: Chiba Institute Northern Islands Trip, May and 24 May 9 June. CNMI Division of Fish and Wildlife, Saipan. Stinson, D.W Birds and mammals recorded from the Mariana Islands. Pages in Biological expedition to the Northern Mariana Islands, Micronesia (A. Asakura and T. Furuki, eds.). Natural History Research, Special Issue 1. Chiba, Japan: Natural History Museum and Institute. Thomas, L., J.L. Laake, S. Strindberg, F.F.C. Marques, S.T. Buckland, D.L. Borchers, D.R. Anderson, K.P. Burnham, S.L. Hedley, J.H. Pollard, J.R.B. Bishop, and T.A. Marques Distance 5.0. Release 2. Research Unit for Wildlife Population Assessment, University of St. Andrews, UK. U.S. Fish and Wildlife Service Recovery Plan for the Micronesian Megapode (Megapodius laperouse laperouse). U.S. Fish and Wildlife Service, Portland, OR. Wiles, G.J A checklist of the birds and mammals of Micronesia. Micronesica 38:

34 Lizard and Small Mammal Surveys, Asuncion May 2008 Antonio S. Castro and Laura L. Williams, CNMI-DFW Introduction In 2000 the Division of Fish and Wildlife conducted baseline biological surveys, including lizard and small mammal surveys, from Anatahan to Agrihan (Cruz et al 2000b-g; Cruz et al 2003). The surveys on Asuncion are part of the ongoing effort to conduct baseline quantitative surveys on all the remote Northern Mariana Islands. Methods Lizard Surveys Diurnal ground lizards were sampled along bird and forest transect 2, between May 2008 (Fig. 1; Fig. 2). Two Victor glue traps were placed flush to the ground at 10 stations 15 meters apart along the transect (Bauer and Sadlier, 1992; Rodda et al., 1993) which was primarily native forest habitat. For three consecutive days, traps were set in the morning ( ), remained for 6-7 hours per day, and were monitored four times per day. Nocturnal lizards were sampled along transect 2 at the same 10 stations every 15 meters (Fig. 1). At each station, one Adhesive Victor sticky glue trap was stapled to a tree at one meter from the ground at 1700 for three consecutive nights. Traps were removed each morning and replaced the next evening with new traps. All captured lizards were removed from the Victor glue traps using vegetable oil, and were weighed, measured and preserved. Lizards were preserved with an injection of 10% formalin in the abdomen then soaked in 10% formalin. Varanus indicus (Monitor lizard) were also sampled along transect 2 (Fig.1). Three flat loop snare traps were established every 75 meters (Berger and Mueller, 1959) and baited with rotten chicken. The traps remained along the transect for four days (96 hrs), the bait was replaced every two days, and the traps were monitored every hours. 34

35 Figure 1. Map of Asuncion with location of bird and forest transects. The lizard and small mammal surveys were performed on transect 2 from surveyed on Asuncion from May Figure 2. Lizard preservation with formalin on Asuncion May

36 Small Mammal Surveys Rats were surveyed using Victor snap-traps along transect 2 (Fig. 1). Ten traps were tied 1 meter above the ground 15 meters apart. Traps were baited with copra and peanut butter. Traps were set at 1700 and checked the next morning at for three consecutive nights. The traps were not placed on the ground because of considerable Micronesian Megapode (Megapodius laperouse laperouse) activity and there was concern that the endangered megapodes might be injured. Results Lizard Surveys Two skink species were caught during the diurnal surveys, Emoia caeruleocauda (Blue-tailed skink)(fig. 3) and Emoia slevini (Mariana skink)(fig.4). The abundance of E. caeruleocauda was many times greater than that of E. slevini (Table 1). The only lizard (gecko) caught during the nocturnal surveys was Gehyra oceanic (Oceanic gecko) (Table 1)(Fig.5). There were no monitor lizards caught during the 216 trap hours of the monitor surveys, and no monitor lizards were observed opportunistically on Asuncion. Small Mammal Survey There were only two rats caught within the 48 trap hours. However, many rats were observed opportunistically during the evening coconut crab surveys. The species of the captured rats have not been identified yet. Table 1. The Presence and abundance of lizards trapped on Asuncion, May, 2008 Diurnal Nocturnal Total # Trap Stations # Trap Hours Emoia caeruleocauda (Blue-tailed skink) 50 Emoia sleveni (Mariana skink) 3 Gehrya oceanic (Oceanic gecko) 9 Lizards/100 Trap hrs

37 Figure 3. Emoia caeruleocauda (Blue-tailed skink) (picture by Gordon Rodda) Figure 4. Emoia sleveni (Mariana skink) (picture by Gordon Rodda). 37

38 Anatahan 1 Sarigan Guguan Alamagan Agrihan Asuncion Figure 5. Gehrya oceanic (Oceanic gecko) (picture by Gordon Rodda). Table 2. The distribution of lizard species detected in the Northern Mariana Islands during baseline biological surveys conducted in 2000, 2006 and 2008 (Cruz et al. 2000b-g; Cruz et al. 2003; Martin et al. 2006). Species Cryptoblepharis poecilopleurus (Oceanic snake-eyed skink) Emoia caeruleocauda (Blue-tailed skink) Emoia sleveni (Mariana skink) Emoia atrocostata (Tide pool skink) Lepidodactylus lugubrus (Mourning gecko) Gehrya mutilata (mmutilating gecko) Gehrya oceanic (Oceanic gecko) Nactus pelagicus (Slender-toed gecko) Varanus indica (Monitor lizard) 38

39 Discussion Lizard species richness varies throughout the northern islands. On Asuncion three lizard species were detected, only one of which was a gecko. On Anatahan six lizard species were detected, four were geckos (Cruz et al. 2003). Alamagan was found to have three skinks, and no geckos (Cruz et al. 2000d), and Guguan (Cruz et al. 2000b) had 2 skink and 4 gecko species. Agrihan (Cruz et al. 2000g) was only found to have one skink species. The distribution of skink species between the islands is variable. The endemic E. sleveni (Mariana skink) has only been detected on Sarigan (Cruz et al. 2000e; Berger et al. 2007), Asuncion, Alamagan (Cruz et al. 2000c) and Pagan (Scott Vogt in 1999 personal communication). The indigenous Cryptoblepharis poecilopleurus (Oceanic snake-eyed skink) has been found on Sarigan Anatahan, Guguan (Cruz et al. 2000b, 2000d, and 2000e) and Asuncion. Emoia caeruleocauda (Blue-tailed skink) has been found on Asuncion, Anatahan, Alamagan, Agrihan, Sarigan (Cruz et al 2000b-g) and Pagan (Scott Vogt in 1999 personal communication). The rare indigenous Emoia atrocostata (Tide pool skink) is recorded only from Alamagan and Guguan (Cruz et al. 2000c and 2000b). The gecko species distribution between islands is also variable. On Anatahan and Guguan three species were detected (Cruz et al. 2000d and 2000b) whereas on Asuncion and Sarigan (Cruz et al. 2000e) only one species has been detected. There were no geckos found in 2000 on Alamagan and Agrihan (Cruz et al. 2000c and 2000g). The introduced Gehrya oceanic (Oceanic gecko) was found only Asuncion, Guguan and Anatahan (Cruz et al. 2000b-g). Gehrya mutilata (Mutilating gecko) has been found on Anatahan and Guguan (Cruz et al. 2000b; Cruz et al. 2003) but was not detected on Asuncion. The indigenous Nactus pelagicus (Slender-toed gecko) was found on Anatahan and Sarigan (Cruz et al. 2000d and 2000e) but not on Asuncion. The indigenous Lepidodactylus lugubrus (Mourning gecko) has been detected only on Guguan (Cruz et al. 2000b). Monitor lizard (Varanus indicus) have been documented on Sarigan (where they appear to flourish) (Martin et al. 2006), Anatahan, Pagan, Agrihan, and Alamagan (Cruz et al. 2000b-g) whereas they have not been detected on Guguan (Cruz et al. 2000b) and Asuncion. The lack of monitor lizards on Asuncion is probably important to the Megapode population. Literature Cited Bauer, Am and R.A. Sadlier The use of mouse glue traps to capture lizards. Herpetological Review, 23: Berger, D.D. and H.C. Mueller The Bal-chtri: A trap for birds of prey. Bird Band. Assu. 30: Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000a. Wildlife and vegetation survey, Aguiguan Technical Report #2. Saipan, CNMI: CNMI Division of Fish and 39

40 Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000b. Wildlife and vegetation survey, Guguan Technical Report #3. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000c. Wildlife and vegetation surveys, Alamagan Technical Report #4. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000d. Wildlife and vegetation surveys, Anatahan Technical Report #5. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000e. Wildlife and vegetation surveys, Sarigan Technical Report #6. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000f. Wildlife and vegetation surveys, Pagan Technical Report #7. Saipan, CNMI: CNMI Division of Fish and Cruz, J., L. Arriola, N. Johnson, and G. Beauprez. 2000g. Wildlife and vegetation surveys, Agrihan Technical Report #8. Saipan, CNMI: CNMI Division of Fish and Cruz, J. L. Williams, N. Hawley, J. Esselstyn, C. Kessler and O. Bourquin Wildlife and vegetation surveys and feral animal control on Anatahan Technical Report #10. Saipan, CNMI: CNMI Division of Fish and Martin, G., L. Williams, J.B.Cruz, N.Hawley, S.Vogt, B. Smith, O. Bourquin, S. Kremer, C. Kessler Wildlife and Vegetation Surveys of Sarigan Island, April Technical Report #14. Saipan, CNMI: CNMI Division of Fish and Rodda, G.H., M.J. McCoid and T.H. Fritts Adhesive trapping II. Herpetological Review, 24:

41 Coconut Crab, Birgus latro Surveys on Asuncion, May 2008 By Scott R. Vogt Introduction The coconut crab (Birgus latro) is the largest terrestrial invertebrate in the world with a large native range from the Pacific to the Indian Ocean. It is an important game species throughout most of its range (Brown and Fiedler 1991, Fletcher and Amos 1994, Kessler 2006). In the Mariana Islands coconut crabs are popular food on holidays and at fiestas. In the Commonwealth of the Northern Mariana Islands (CNMI) there is a season for crab hunting with size, sex and bag limits (2 NMIAC 85). Even with these regulations the population is overharvested on the CNMI s inhabited islands, as it is in most of its range (Kessler 2006, USFWS 2001, U.S. Navy 2008). In the CNMI there is the commonly held belief that the uninhabited islands north of Saipan (aka the Northern islands) hold vast numbers of crabs. However, there has been little to no systematic crab surveys performed on these islands and crab abundance and demographics are unknown. This study is the first time that a systematic sampling method was used to measure crab abundance and demographics on Asuncion. The only other similar survey within the northern Mariana Islands has been on Sarigan (Martin et al. 2008). Methods This work was performed on Asuncion from May 19-22, 2008 using the modified methods of Fletcher and Amos (1984) as described by Cruz (2005). Crabs were sampled along two coconut crab transects (Crab 1 and Crab 2) (Fig. 1) at 100 baited stations established every 20 meters. The coconut crab transects were located partially along bird and forest transects 1, 3 and 4 (Fig. 1). The habitat was a mix of coconut forest closer to the coast and native forest on the inland sections of both Crab transect 1 and 2. Bait consisted of an aged coconut with milk still inside and cat food. A triangular shaped cut approximately 4 x 3 cm was made into the husk of the coconut exposing the interior meat (Fig. 3). Canned cat food was placed on the coconut near the interior opening. The coconut was then tied with wire to a tree at the station. Crab transect 1 contained 50 traps which were baited and surveyed on May 20 and 21. Crab transect 2 was bated and surveyed on May 21 and 22, All captured crabs were weighed, measured for thoracic length and width (Fig. 2), sexed, marked and released at the capture site (Fig. 3-4) (Fletcher and Amos 1994; Cruz 1995). Crabs were marked by painting a number on the carapace with brightly colored fingernail polish. This ensured that crabs were not double counted. Pesola balances and digital calipers were used for 41

42 the weights and measures. Abundance is expressed as the number of crabs captured divided by the number of trap nights (Fletcher and Amos 1994; Cruz 1995) and is not absolute population estimate but rather an index of the relative abundance. A trap night is defined as one station set for one night. Figure 1. The location of Coconut crab surveys performed on Asuncion from May Crab transects were established partially along bird and forest transects 1, 3 and 4 and crossed native and coconut crab forest. 42

43 Figure 2. The Thoracic width (TW) and length (TL) measurements taken from all crabs captured at stations or along transects between May 2008 on Asuncion. Figure 3. Coconut crab bait station along crab transect 1 on Asuncion May

44 Results There were 35 crabs captured over a total of 200 trap nights for a catch rate of 0.18 crabs/ trap night (Table 1). Only four females were captured. The mean weight of the males (n=31) was over twice that of females (n=4), ±748.03gm and ±302.15gm respectively (Table 2). There was less variation between the sexes for mean thoracic length and width. Males mean thoracic length and width was 61.10±11.61 mm and ±33.56 mm respectively, whereas females thoracic length and width was 45.00±10.83 mm and 90.00±21.34mm respectively (Table 2). Figure 4. Measuring the thoracic width of a coconut crab caught along transects on Asuncion May 2008 by Patrick Santos and Paul Lisua. 44

45 Table 1. The weight, thoracic length and width of coconut crabs caught on 2 transects during two nights from May 2008 on Asuncion. Sex Thoracic Length Total Weight (gm) Thoracic Width (mm) (mm) M F M M M M M M M M M M M F M F M M M M M M M F M M M M M M M M M M M

46 Table 2. The mean thoracic length, width and total weight of all coconut crabs, males and females caught at bait stations along two transects on Asuncion on May Thoracic Length (mm) Total Weight (gm) Thoracic Width (mm) Total Mean (n=36) ±sd ± ± ± % Confidence Confidence Interval Males Mean (n=31) ±sd ± ± ± % Confidence Confidence Interval Females Mean (n=4) ±sd ± ± ± % Confidence Confidence Interval Standard deviation Discussion Overall the catch rate recorded on Asuncion 0.18 crabs/trap night or bait is similar to that of Sarigan in 2006 (coconut and native forest habitat) (0.09 crabs/trap night) (Martin et al., 2008), Tinian (Leuaeana leucocephala and native forest) (0.11 crabs/trap night) (U.S. Navy 2008) and Saipan (L. leucocephala forest) (0.24 crabs/bait) (Kessler 2006). It is surprising however that the rate is so low for a population that likely has little or no harvest pressure. Unlike Tinian (U.S. Navy 2008) and Saipan (Kessler 2006), but similar to Sarigan, the population demographic was skewed to larger crabs on Asuncion (e.g. > 40 mm in thoracic length) (Fig. 5). The Island of Diego Garcia is a small atoll in the Indian Ocean that serves as a base for the U.S. Navy. There is restricted access to the island and half of the atoll is a fenced conservation area. This is one of the few areas in the world with a non-harvested crab population and therefore can serve as a reference population for comparisons with the Mariana Island populations (Scott Vogt, U.S. Navy, unpublished). Coconut crab bait stations along transects were conducted (40 trap nights) within the conservation area, which is a coconut forest. The population demographics on Diego Garcia are more evenly distributed amongst the size classes than in any of the populations in the Mariana Islands (Fig. 5). Additionally, the catch rate on Diego Garcia was 1.78 crabs/trap night, over 10 times higher than Asuncion (Scott Vogt, United States Navy, unpublished data). 46

47 Thoracic length (mm) Figure 5. Coconut Crab Size Demographics for Select Marianas Islands and Diego Garcia If Asuncion coconut crabs were being harvested the population size demographics would be similar to those on Saipan and Tinian with the population skewed towards smaller crabs as larger crabs are more likely to be harvested (Fig. 5). However, similar to Sarigan, the exact opposite was seen with over 90% of the sample being over the legal size limit (3 inches wide) (2 NMIAC 85) (Fig. 6). In light of the demographic data we do not believe that the Asuncion crab population is being subjected to significant harvest pressure. The lack of smaller crabs and low densities indicates very low juvenile recruitment, and not harvest. It is possible that the observed low juvenile recruitment is the result of an unknown sampling bias or small sample size; however the same sampling technique did detect the small crab cohort (< 3 inches across the back) on other islands. Therefore, the negligible small crab cohort on both Asuncion and Sarigan (Martin et al., 2008) is likely not a sampling artifact but due to other factors. Coconut crabs fertilize their eggs on land but release the eggs to hatch in the ocean where there is a marine larval stage lasting 2-4 weeks (Greenway 2003). It is possible that the lack of a fringing reef on Asuncion allows currents to disburse many of the larvae, causing juvenile recruitment into the adult cohorts to be low and/or infrequent (Greenway 2003). Juvenile recruitment of crabs on any given island might actually be from larvae that originated on another island and were carried by currents (Greenway 2003). Brown and Fiedler (1991) documented very low recruitment on Vanuatu. They estimated that substantial juvenile recruitment occurred only once every 5-10 years (Brown and Fiedler 1991). After further research they concluded that 47

48 coconut crab larvae are transported by surface currents which are highly influenced by the prevailing wind conditions at time of release (Brown and Fiedler 1991). Figure 6. The percent of coconut crabs over/under the legal hunting size of 3 inches wide (2 NMIAC 85) on Saipan (Kessler 2006), Tinian (US Navy 2008), Diego Garcia (Scott Vogt unpublished data), Sarigan (Martin et al. 2008) and Asuncion (from surveys May 2008). The longevity of coconut crabs allows populations to sustain themselves even with low recruitment; however, excessive human harvest can put this in jeopardy. Populations with low juvenile recruitment can be easily over-harvested as the adults are replaced at a lower rate than what is harvested. Compounding this is the slow growth rate for coconut crabs which take 8-10 years to reach the CNMI legal size limit (Brown and Fiedler 1991). If other uninhabited islands north of Saipan show a similar demographic tendency as Asuncion and Sarigan then great care should be taken in harvesting these populations. Fletcher and Amos (1994) estimated that the average annual natural mortality of a harvestable (e.g. legal size) crab population was 5% and stated that harvest rates should not exceed this. More research is warranted on this issue in the CNMI since this is an important species to the local culture and because of the belief (quite possibly erroneous) that large numbers of crabs exist in the Northern islands. 48