Geology topics

Trevor Bak

Publications and source records attributed to Trevor Bak.

7 recordsLinked to original sources

USGS2026

Hakalau’s moving castle: How climate change and restoration are shifting an island fortress for forest birds

Hakalau Forest Unit of the Big Island National Wildlife Refuge Complex (hereafter, Hakalau) protects the largest area with the highest endemic forest bird diversity in Hawaiʻi, including four federally listed species. Hakalau’s higher elevation montane forest provides refuge from avian malaria ( Plasmodium relictum ), a primary driver of Hawaiian honeycreeper extinctions. However, recent declines in Hakalau’s birds at lower elevations could indicate that conditions have become suitable for disease vector Culex quinquefasciatus . We evaluated the statuses of Hakalau’s bird populations in the context of recent climatic changes using new survey data from point-transect distance sampling, producing abundance estimates from 1999 to 2024. We stratified our analysis across four elevation ranges (<1500 m, 1500–1700 m, 1700–1900 m, and >1900 m) and assessed trends for each species using state-space models (SSMs). We constrained population trajectories to be biologically realistic by incorporating population dynamic models within the SSMs. We observed highly species-specific abundance trends below 1500 m, predominantly stable to upward trends within 1500–1700 m, stable trends within 1700–1900 m, and upward trends above 1900 m. Declines in Hawaiʻi ʻamakihi ( Chlorodrepanis v. virens ) and endangered ʻakiapōlāʻau ( Hemignathus wilsoni ) abundance coincided with lengthening warm seasonal temperatures indicative of shrinking disease-free habitat below 1700 m. Above 1900 m, however, increases in nearly all species indicate that reforestation has likely restored disease-free habitat since 1999. While most species were stable to increasing overall, surveillance for mosquitoes and disease at lower elevations, documenting changes in habitat, and continuing bird population monitoring can help to gauge their long-term persistence at Hakalau.

Hawaii↗

USGS2025

Quantifying landscape-level biodiversity change in an island ecosystem: A 50-year assessment of shifts in the Hawaiian avian community

Hawaii has experienced profound declines in native avifauna alongside the introduction of numerous bird species. While site-specific population studies are common, landscape-level analyses of avian population dynamics are rare, particularly in island ecosystems. To address this gap, we used a density surface model to create a spatio-temporal projection of population densities and distributions across the Island of Hawai‘i, spanning nearly five decades (1976–2023). We incorporated environmental covariates of habitat, precipitation, and elevation, to further refine our projections. Our analysis encompassed nine native and six non-native bird species, inhabiting a range of ecological niches. We found five out of nine native species have declined in density and range size while four were stable. For non-native species, two were stable, one was decreasing, and three were increasing in density and range size. Our landscape projections can inform management by suggesting areas critical for habitat preservation and land acquisition for conservation, identifying where range fragmentation is occurring, and pinpointing locations of multi-species declines that are likely driven by a common cause. Our study demonstrates how long-term, landscape-level monitoring and analyses can advance understanding and addressing biodiversity loss, particularly in vulnerable tropical island ecosystems.

Hawaii↗

USGS2025

Automated methods for processing camera trap video data for distance sampling

Context Population monitoring is an essential need for tracking biodiversity and judging efficacy of conservation management actions, both globally and in the Pacific. However, population monitoring efforts are often temporally inconsistent and limited to small scales. Motion-activated cameras (‘camera traps’) offer a way to cost-effectively monitor populations, but they also generate large amounts of data that are time intensive to process. Aims To develop an automated pipeline for processing videos of ungulates (Philippine deer, Rusa marianna ; and pigs, Sus scrofa ) on Andersen Air Force Base in Guam. Methods We processed camera videos with a machine learning model for object detection and classification. To estimate density using distance sampling methods, we used a separate machine learning model to estimate the distance of target animals from the camera. We compared density estimates generated using manual versus automated methods and assessed accuracy and processing time saved. Key results The object detection and classification model achieved an overall accuracy >80% and F1 score ≥0.9 and saved 36.9 h of processing time. The automated distance estimation was fairly accurate, with a 1.1 m (±1.4 m) difference from manual distance estimates, and saved 16.8 h of processing time. Density estimates did not differ substantially between manual and automated distance estimation. Conclusions Machine learning models accurately processed camera videos, allowing efficient estimates of density from camera data. Implications Further adoption of motion-activated cameras coupled with automated processing could lead to continuous, large-scale monitoring of populations, helping to understand and address changes in biodiversity.

Guam↗

USGS2025

Status and trends of forest bird populations at Hakalau Forest National Wildlife Refuge, 1987–2024

Since 1985, the Hakalau Forest Unit of the Big Island National Wildlife Refuge Complex (hereafter, Hakalau) has protected the largest endemic forest bird diversity in the State of Hawaii. This includes three endangered and one threatened species and their habitats. Hakalau’s vast area (155 km2), mostly high elevation (>1500 m) montane forest, provides refuge from avian malaria (Plasmodium relictum) vectored by introduced southern house mosquitoes (Culex quinquefasciatus). However, increases in the seasonal temperatures optimal for disease carrying mosquitoes associated with climate change have coincided with recent downward trends of native species in the closed forest area of the refuge (1450–1750 m), where disease is most likely to occur. Therefore, to inform refuge management with the most updated information on these populations and their trends, we analyzed forest bird survey data collected using point transect distance sampling with new survey data from 2021–2024. We stratified our analysis across management units, including the open forest and pasture since 1987, and the closed forest since 1999, and across four elevation ranges (<1500, 1500–1700, 1700–1900, and >1900 m) since 1999. We used distance sampling to estimate species- and strata-specific abundances and applied log-linear regression to detect trends across the timeseries. We found a continuation of previous trends, wherein most native forest birds declined in the closed forest strata (mostly below 1700 m) and increased in the highest elevation and pasture strata. Patterns were highly species-specific for the three lower elevations and open forest strata. ‘Apapane (Himatione sanguinea) and warbling white-eye (Zosterops japonicus), the most abundant native and introduced bird species, respectively, increased in nearly all strata. ‘I‘iwi (Drepanis coccinea) and Hawai‘i ‘amakihi (Chlorodrepanis virens virens), also common, decreased in closed forest, were stable in open forest, and increased in pasture. Both species were generally stable or increasing across all elevation bands, except Hawaiʻi ʻamakihi decreased in the below 1500 m and the 1500–1700 m elevation bands. All three of the federally endangered forest bird species declined in closed forest. Hawai‘i ‘ākepa (Loxops coccineus) also declined in the open forest and 1700–1900 m band and was the only species to decline overall. ‘Akiapōlā‘au (Hemignathus wilsoni) declined in closed forest but increased overall. ‘Alawī (Loxops mana, also known as Hawai‘i creeper) also decreased in the closed forest, but was stable to increasing in most strata and remained stable overall. Hawaiʻi ʻelepaio (Chasiempis sandwichensis) declined in closed forest, was stable in open forest and most elevation strata, and increased in pasture. ʻŌmaʻo (Myadestes obscurus) was stable in open and closed forest and at middle elevations and increased in the pasture. Introduced red-billedleiothrix (Leiothrix lutea) declined in closed forest, was stable in the open forest and at most elevations, and increased in the pasture. Coinciding with these changes, seasonal conditions for vector occurrence have continued to lengthen in the lower elevation strata through 2024, while Hakalau’s outplanting efforts continued to increase forest cover in the pasture, suggesting that disease-free habitat may have decreased in the closed forest and increased in the pasture. The declines in the closed forest and mixture of trends in the open forest and middle elevations bands suggest emigration into the higher elevation strata from lower elevation strata, and that possible threats have suppressed forest birds even at elevations >1500 m.

Hawaii↗

USGS2025

Forest bird population status on Saipan, a small oceanic island

Tropical oceanic islands are critical biodiversity hotspots where population monitoring can help to determine the status and trends of rare and endangered species. Saipan is the second largest island in the Mariana Islands and contains many endemic and range-restricted bird species. Surveys of forest birds were conducted on Saipan using point-transect distance sampling, starting in 1982. To determine population status for Saipan avifauna, we generated abundance estimates based on surveys conducted in 1982, 1997, 2007, and 2018. We also assessed community composition by exploring if there were changes in assemblage diversity, total forest bird abundance, and ratio of native to non-native populations over time. For the 2018 survey, there were 5791 detections of 18 species during 242 counts, with six of 18 species detected at > 50 % of the counts and with relatively high abundance, while 12 species were detected at < 50 % of the counts and were uncommon or rare. The most abundant species was Bridled White eye (nosa', Zosterops conspicillatus saypani ) at > 40 birds/ha and the least common species was the Saipan Reed Warbler (gå'ga' karisu, Acrocephalus hiwae) at < 1 bird/ha. We found that the community composition is becoming slightly more even over time and the overall community abundance is stable (around 375,000 birds), but non-native species have become more common while native species have remained stable. Results from this study can be used to inform conservation management of forest birds in Saipan and contribute to a broader understanding of forest bird status in the Marianas.

Saipan↗

USGS2024

Nihoa and Laysan Island passerines population abundances, trends, and habitat utilization

Nihoa and Laysan Island, part of the Northwestern Hawaiian Islands, are host to three endangered passerine species—Nihoa finch (Telespiza ultima), Nihoa millerbird (Acrocephalus familiaris kingi), and Laysan finch (Telespiza cantans). Using point-transect distance sampling survey records from 2010 to 2022 for Nihoa and 2013 to 2019 for Laysan Island, we estimated the density and abundance of all three species. We also compared densities between habitats for both islands, classifying Nihoa habitat as Eragrostis (grass) or mixed shrub and Laysan Island as open or dense habitat. The population of Nihoa finch remained stable with slight fluctuations among years, with a population size of 6,592 (4,954–8,655) birds for the most recent sampling year, 2022. Laysan finch also remained stable with a total population in both dense and open habitat of 17,657 (11,994–23,320) for the most recent sampling year, 2019. The Nihoa millerbird significantly increased on both Nihoa and Laysan Island with a global population of 1,907 (1,291–2,766) in 2019, the most recent year both islands were sampled. Overall, the populations of these three endangered species were stable or increasing. The increase of Nihoa millerbird on Laysan Island, after being translocated to the island in 2011 and 2012, represents the successful establishment of a second viable population, leading to a decrease in the species’ IUCN extinction risk status (from critically endangered to endangered). Continued population monitoring of these endangered species can help inform future management actions and ensure their preservation into the future.

Hawaii↗

USGS2022

2021 Tinian Island forest bird abundance estimates

The U.S. Navy, through Micronesian Environmental Services, surveyed landbirds in the Military Lease Area on Tinian Island in May and June 2021 using point-transect distance sampling methods. There were 2,074 individuals of 14 species detected during 123 point counts. Six species were detected during >50% of the counts and were observed at relatively high abundances, while eight species occurred at <50% of the counts and were uncommon to rare. Densities of native landbirds in the Military Lease Area ranged from the uncommon Mariana kingfisher ( Todiramphus albicilla ) at 0.46 birds/ha (95% confidence interval [CI] = 0.33–0.63) to the very abundant bridled white-eye ( Zosterops conspicillatus ) at 102.63 birds/ha (95%CI = 86.70–122.91). Most distances recorded during the 2021 Military Lease Area survey were rounded to distance intervals of 0 and 5. Measuring exact distances of detected animals is preferable to collecting distances grouped into bins or rounding. Direct comparison with previously published estimates was not possible because of changes in the sampling frame; however, densities of six species were greater, two were smaller, and one was similar to the 2008 survey estimates for the Hagoi, Diablo, and Masalog regions. Our findings indicate that the landbird community in the Military Lease Area appears to be dynamic and resilient.

Military Lease Area, Tinian Island↗