The Palila (Loxioides bailleui), the last member of the once speciose finch-billed Hawaiian honeycreeper clade (Drepanidinae) in the main Hawaiian Islands, faces critical conservation challenges as an endangered species. Understanding the drivers of its decline is essential for effective management. We used additive decomposition models to examine temporal trends in climatic variables (temperature, precipitation, drought) and Normalized Difference Vegetation Index (NDVI), a vegetation health metric hypothesized to be associated with long-term trends in Palila abundance at landscape (250 m) scales on the Island of Hawai'i. A breakpoint analysis identified 2005-2009 as critical years of Palila decline. Vegetation health metrics at the 250 m scale lined up well both spatially and temporally with trends in Palila declines, with a significant browning from January 2004 to January 2014. Given the strong correlation between vegetation health and drought metrics at the landscape scale (r = 0.75, p < 0.001), NDVI changes appeared driven by drought. To enable the future projection of habitat quality in this area, we explored a stepwise linear regression to explain the variation in MODIS NDVI in recent years. We found that 87 % of the variability in NDVI can be explained by wet season precipitation and vapor pressure deficit from the previous dry season. The model is largely driven by a strong positive correlation between wet season precipitation and NDVI (r = 0.72, adjusted p < 0.001). Areas that maintained a low likelihood of NDVI decline throughout the time series and experienced increases in predicted Palila count represent potential drought microrefugia for the species. This higher elevation microrefugia is likely resilient against decreases in wet season precipitation through supplemental water retention from fog drip. While NDVI rebounded after 2014, Palila have not recovered. Our analysis highlights the importance of trend decomposition for monitoring endangered species with limited rebound potential due to small population dynamics and indicate continued warm, dry conditions may prevent Palila recovery without intervention.

[This corrects the article DOI: 10.1371/journal.pone.0288306.].

Palila (Loxioides bailleui) are critically endangered Hawaiian honeycreepers specializing on m & amacr;mane (Sophora chrysophylla) seeds and restricted to Mauna Kea volcano on the Island of Hawai'i. Recently, the population was estimated to decline by 89% between 1998 and 2021, despite decades of ungulate removal, fence construction, m & amacr;mane regeneration, fire suppression, and predator control. To inform managers with the most recent update on the status and trends of the Palila population, we analyzed annual bird survey data collected using point-transect distance sampling since 1998, including new annual survey data from 2022, 2023, and 2024. Prior to analysis, we predicted the population trajectory would change between 2021 and 2024 because of continued management actions promoting habitat recovery. We used distance sampling, log-linear regression, and state-space modeling to produce the new estimates and analyze trends across the time series. The 2022 population estimate was 367 to 742 birds (95% confidence interval; point estimate: 545), the lowest in recorded history. The 2023 and 2024 estimates of 374 to 842 birds (point estimate: 596) and 412 to 970 birds (point estimate: 666) were the second and third lowest in our time series, respectively. Our estimates for years before 2022 show population fluctuations between 4000 to 6800 birds from 1998 to 2005, then a steep decline through 2010. For the next decade, abundance fluctuated around 1000 birds, before declining again in 2021 to less than 700 birds. From 1998 to 2024, the population declined by more than 90%, or 205 birds per year, with 100% statistical support for an overall downward trend, despite significant management efforts and research. The greatest threats facing the Palila, if familiar, are not being eliminated swiftly enough to promote their recovery. The currently small and range-limited population is vulnerable to future climate-related events such as drought and fire. Continued monitoring can help to assess the response of Palila to adaptive management actions and changing environmental conditions.

Annual point counts are commonly used to monitor birds to track population densities across space and time. Palila (Loxioides bailleui) are surveyed annually in the first quarter, but we recently instituted quarterly sampling that offers a unique opportunity to improve estimator precision. We conducted point-transect distance sampling point counts during the first quarter of 2020 through 2024, and the second through fourth quarters in 2022 and 2023, and the second quarter in 2024. The reduced sampling intensity during the quarterly counts, however, requires model-based methods to estimate abundance to the entire sampling frame. We modeled spatial and temporal correlation using a soap film smoother within a generalized additive modeling framework, a density surface model, fitted to palila counts each quarter for the five-year timeseries to track changes in population abundances. Our results indicate that palila maintained a high-density hotspot throughout the five-year timeseries; however, the extent of the hotspot declined substantially over the timeseries while densities within the hotspot declined from about 3 birds/ha in 2020 to about 1 bird/ha in 2024, which resulted in a 66% decline in palila abundances over 5 years. Density surface model estimates give on average a confidence interval width that was 74.7% shorter than the associated distance sampling confidence interval widths. Our results indicate that palila may benefit most if management actions were applied within the remaining hotspot. Additionally, this temporally fine-grained sampling provides information on seasonal movement patterns and resource tracking, and population response to management and conservation actions. Our spatially explicit, model-based approach is applicable to a wide range of monitoring programs, particularly those with inconsistent, opportunistic spatial coverage.

Overexploitation, habitat conversion, and introduced species have caused unprecedented extinctions and heavily degraded native bird populations in island ecosystems. In the Hawaiian Islands, stemming these losses has proven difficult as the highly specialized avifauna are often impacted - among other things - by poorly understood trophic disruptions as well as persistent climatic shifts. Here we investigate these dynamics by examining the trophic positions of 7 different taxa of producers and consumers across the last century in the subalpine mamane (Sophora chrysophylla) forest ecosystem on the island of Hawai'i. From museum collections and contemporary sampling, we analyzed the stable nitrogen (delta 15N) values of producers and consumers to identify trophic and source amino acids and derive trophic enrichment constants specific to this food web. This enables us to reconstruct the diet of the palila (Loxioides bailleui), a critically endangered finch whose population recently declined 90 %. Our results show that from 1890 to 2006, the palila trophic position declined from 2.6 to 2.2, with cascading implications for its diet. Bayesian mixing model reconstructions indicate that palila trophic position changes likely arose from a 76 % decline (69.3 % to 16.6 % of diet) in the consumption of native moth caterpillars, and a 172 % increase (30.7 % to 83.4 %) in native plants. From the available ecosystem variables, exploratory Bayesian multiple regressions selected surface temperature changes, and the interactions of surface temperatures with drought and caterpillar parasitism as the primary drivers of these trophic changes. Despite the predicted increases of warming and drought, management interventions may build resiliency in this unique island ecosystem.

Long‐term ecological studies are invaluable for detecting changes over time. Forest restoration has been a conservation priority in Hawaiʻi, where invasive species have negatively impacted native bird habitat. During 1993–1994, a study was conducted of Hawaiʻi ʻelepaio ( Chasiempis sandwichensis ) nest site selection and forest composition in mesic montane forest along Mauna Loa Road in Hawaiʻi Volcanoes National Park. We returned to the site and repeated the methods used in the earlier study to record Hawaiʻi ʻelepaio nest site selection (tree species, tree and nest height, and tree girth [diameter at breast height, DBH]) from 2015 to 2017, and measured forest composition (tree density, relative abundance, height, and DBH) in 2019. Three times as many nests were found in ʻaʻaliʻi ( Dodonaea viscosa ) as in koa ( Acacia koa ) in both time periods. Heights of koa and ʻaʻaliʻi did not change, but their DBH increased over time. The relative height of nests in ʻaʻaliʻi trees did not change, but nests in koa were placed higher in the crown during the later study. Overall tree density increased from 1,619 (95% confidence interval [CI] = 1,499–1,740) trees/ha to 2,583 (95% CI = 2,176–3,096) trees/ha. Koa relative abundance was 53% of total trees earlier and 45% later, ʻaʻaliʻi remained at 47%, and māmane ( Sophora chrysophylla ) increased from 0 to 8% over time. Our results indicate that changes in forest composition affect nesting behavior, but in ways that are not necessarily simple or consistent.

AbstractWe investigated the reproductive ecology and effects of egg parasitism on the Samoan swallowtail butterfly (Papilio godeffroyi), which survives only on Tutuila Island, American Samoa, after having disappeared from the much larger islands of Upolu and Savai‘i in independent Samoa. During monthly surveys of its only known host plant, Micromelum minutum, across eight sites in 2013 and 2014, we collected eggs, eggshells, larvae, pupae, and pupal exuviae. Live specimens were reared under laboratory conditions to determine reproductive outcomes, developmental rates, and sex ratios, as well as parasitoid attack frequencies, brood sizes, and sex ratios. Sixty‐six of 448 (14.7%) eggs produced larvae, 47 of which became adults. The sex ratio was approximately even overall and within each developmental stage. Eggs were slightly larger on individual host trees and in host tree stands that yielded more eggs per unit of foliage, indicating that ovipositing females responded to some features of host trees and stands. Eggs hatching female or male larvae were similar in size, and the sexes developed at similar rates. A newly described species of parasitoid wasp, Ooencyrtus pitosina (Encyrtidae), emerged from 73.6% of 382 butterfly eggs that failed to hatch in the laboratory (62.7% of 448 eggs overall). Forty‐one other eggs contained dead parasitoid larvae. An additional, unidentified Ooencyrtus wasp species emerged from a single P. godeffroyi egg. No parasitoids were reared from P. godeffroyi larvae or pupae. Of 656 P. godeffroyi eggshells collected in the field and examined in the laboratory, 62.2% showed signs of having been parasitized by O. pitosina. There was no evidence that parasitism rates were density‐dependent. O. pitosina brood sizes ranged from 1 to 5, with the sex ratio skewed toward females (2.40 F:1.00 M). Larger parasitoid broods were associated with slightly larger host eggs, indicating that female wasps may adjust brood size according to host egg size or that fewer wasp larvae are able to complete development in smaller eggs. Techniques used to rear both P. godeffroyi and O. pitosina in the laboratory could be applied to a captive‐rear, wild‐release program, which may facilitate reestablishment of the species in Samoa.

Effective species management and conservation benefit from knowledge of species distribution and status. Surveys to obtain that information often involve replicate sampling, which increases survey effort and costs. We simultaneously modeled species distribution, abundance and spatial correlation, and compared the uncertainty in replicate abundance estimates of the endangered palila (Loxioides bailleui) using hierarchical generalized additive models with a soap film smoother that incorporated random effects for visit. Based on survey coverage and detections, we selected the 2017 point-transect distance sampling survey on Mauna Kea, Hawai'i Island, for our modeling. Our modeling approach allowed us to account for imperfect detections, control the effects of boundary features, and generate visit-specific density surface maps. We found that visit-specific smooths were nearly identical, indicating that little information was gained from a subsequent visit, and that most of the estimator uncertainty was derived from within-visit variability. Scaling back the palila survey to a single visit would halve the survey effort and logistical costs and increase efficiencies in data management and processing. Changing the sampling protocol warrants careful consideration and our findings may help management and regulatory agencies by maximizing efficiency and minimizing costs of surveying protocols, while providing guidelines on how to best collect information critical to species' conservation.

A new species of encyrtid wasp, Ooencyrtus pitosina Polaszek, Noyes & Fusu sp. n., (Hymenoptera: Encyrtidae: Encyrtinae) is described as a gregarious parasitoid in the eggs of the endemic Samoan swallowtail butterfly Papilio godeffroyi (Lepidoptera: Papilionidae) in the Samoan archipelago. It is described here because it is an important natural enemy of this butterfly, and to facilitate identification for future work with this parasitoid and its host.

Invasive fish can profoundly affect communities they invade.In Hawai'i, invasive fishes have become established in many anchialine pools, threatening the persistence of resident invertebrates, including several endangered species.Tools to eradicate invasive fishes from these pools are lacking.This study tested the efficacy of carbon dioxide (CO 2 ) gas diffused into anchialine pool water as a method to eradicate invasive Mozambique tilapia (Oreochromis mossambicus), guppies (Poecilia reticulata), and western mosquitofish (Gambusia affinis).We first conducted aquarium trials to identify how these fishes were affected by elevated CO 2 and the concomitant reduction in pH.We then carried out field trials in pools containing these fish in one pool each at two national historical parks on the Island of Hawai'i during July 2021-January 2022.We also developed environmental DNA (eDNA) protocols to detect fish that may have survived CO 2 treatments.The effect of CO 2 on fish behavior varied among species; at pH 5.3 (CO 2 = 255 mg/L) for tilapia and 5.0 (CO 2 = 488 mg/L) for tilapia, guppies, and mosquitofish, all generally lost their ability to swim, showed slow or no gill movement, and altered their position in the water column.No tilapia survived the trials (n = 4 and 6 individuals at pH 5.3 and 5.0, respectively).In contrast, 41.7% (n = 12) of adult guppies and 66.7% (n = 12) of adult mosquitofish survived treatment at pH 5.0.In the field we were unable to reduce anchialine pool water pH below 5.7.Regardless, we were able to eradicate tilapia from one pool over four sequential treatments.Post-treatment eDNA assessments supported visual surveys, confirming our results.We were not able to eradicate guppies and mosquitofish.Results from this study show that CO 2 can be an effective tool for eradicating invasive tilapia from anchialine pools, and post-treatment eDNA assessments can provide managers with a method for evaluating the success of eradication efforts.

The Samoan swallowtail butterfly (Papilio godeffroyi) has become restricted to Tutuila Island, American Samoa. Factors driving its extirpation on other islands may be partly due to the availability and suitability of habitat, given the singular association we observed of P. godeffroyi with its host plant, Micromelum minutum. We expected that as a host plant specialist, P. godeffroyi might respond to variation in the physical traits and habitat context of M. minutum, which could help identify conservation opportunities throughout its present and former range. Because the ecology of P. godeffroyi is largely unknown, we conducted systematic surveys during 15 months in 2013 and 2014 on Tutuila to investigate its reproductive response to the physical and phenological variability of individual host trees and to the structural and geographical variability of the stands in which they occurred. We searched foliage for eggs, larvae, and pupae and monitored seasonal trends in herbivory and production of leaves, flowers, and fruits on 117 host trees in 8 rainforest stands, mostly within the National Park of American Samoa (NPSA), starting in March 2013 and ending in August 2014. We observed high variability in butterfly reproductive output between individual trees and stands, due largely to a strong negative relationship between leaf biomass (at both tree and stand levels) and oviposition density (number of eggs per foliar biomass) and frequency (number of eggs per survey). Distribution patterns of larvae and pupae were less distinct but generally followed trends in oviposition. Our findings indicate that the butterfly is reproducing widely across Tutuila and is likely to respond positively to management to increase the availability of small host trees in low-density stands. Thus, the species is a good candidate for conservation management focusing on habitat.

Effective species management and conservation require knowledge of species distribution and status. We used point-transect distance sampling surveys of the endangered palila (Loxioides bailleui), a honeycreeper currently found only on the Island of Hawai'i, USA, to generate robust estimates of total abundance and simultaneously model the distribution, abundance, and spatial correlation of the species as a density surface model (DSM). Point-transect distance sampling is a widely applied method to estimate bird densities accounting for imperfect detection probability. For the DSM we used a generalized additive model framework and soap film smoothers to control the effects of boundary features. This modeling approach allowed us to account for imperfect detection and propagate detection probability uncertainty. We compared the uncertainty in palila abundance estimates using standard point-transect distance sampling to estimates from the DSM. The DSM, accounting for both distance-sampling-derived detection probability variance and the generalized additive model density estimate variance, did not improve population estimator precision; however, it provided insight into the species' distribution, density, and uncertainty. We also applied excursion sets analysis to objectively identify areas where the species occurs in high densities. The 2017 global population of <2,000 individuals was limited to an excursion area of 1,500 ha. Our findings can help management and regulatory agencies by simultaneously mapping a species' distribution and density, improving survey protocols, and providing information important to species conservation.

Landbird populations on Tinian Island have been periodically surveyed since 1982 to evaluate the status of non-native and native landbirds. We report the results of surveys in 2013 and the observed changes during 31 years in species population trends based on surveys since 1982. A total of 11 native and 3 non-native species were detected during the 2013 survey. Population sizes were estimated using point-transect distance sampling methods, and population trends were assessed using repeated measures analysis of variance for nine forest bird species. In all years, the Rufous Fantail (Rhipidura rufifrons) and Bridled White-eye (Zosterops conspicillatus) were the most abundant species, whereas the White-throated Ground Dove (Pampusana xanthonura) was the least abundant species in 1982, 1996, and 2008, and the Mariana Kingfisher (Todiramphus albicilla) was the least abundant in 2013. The less common species numbered in the low thousands included the Mariana Fruit Dove (Ptilinopus roseicapilla), White-throated Ground Dove, introduced Philippine Collared Dove (Streptopelia dusumieri), Mariana Kingfisher (Todiramphus albicilla), and Micronesian Myzomela (Myzomela rubratra). The Micronesian Starling (Aplonis opaca) and Tinian Monarch (Monarcha takatsukasae) were estimated to number in the tens of thousands. The most abundant species were the Rufous Fantail, numbering more than 100,000, and the Bridled White-eye, numbering more than 400,000. The overall trends in abundance between 1982 and 2013 showed an increase in the Mariana Kingfisher, Micronesian Starling, Rufous Fantail, White-throated Ground Dove, and Philippine Collared Dove, while populations were stable for the Bridled White-eye and Tinian Monarch. Declines were seen for the Mariana Fruit Dove and Micronesian Myzomela. These trends matched previous analyses with the exception that Tinian Monarch abundance showed an increase in the 2013 survey.

AbstractOutbreaks of defoliating insects in low‐diversity tropical forests occur infrequently but provide valuable insights about outbreak ecology in temperate environments and in general. We investigated an extensive outbreak of the endemic koa moth (Scotorythra paludicola), which defoliated endemic koa trees (Acacia koa) over a third of their range on Hawai‘i Island during 2013 and 2014. At Hakalau Forest National Wildlife Refuge, we observed the dynamics of the outbreak and its effects on host trees, nutrient cycling, and insectivorous consumers in reforestation stands of densely planted koa and in natural forest stands of mixed koa and ‘ōhi‘a (Metrosideros polymorpha). Contrary to predictions of the resource concentration hypothesis, caterpillar biomass and defoliation severity were greater in the natural forest sites, where koa density was relatively low. Caterpillars preferentially consumed the most palatable koa foliage type (phyllodes), and koa initially refoliated with the least palatable foliage type (true leaves). Lightly defoliated small trees refoliated more quickly than did heavily defoliated ones but the opposite was true for large trees, which also produced a greater proportion of phyllodes. Mortality was greatest for heavily defoliated small koa. Caterpillar frass caused larger increases in soil nitrogen (N) than phosphorus (P) availability, with the greatest N increases in fine‐textured soils. Foliar N increased in alien grasses under koa canopies compared to grasses away from koa and to native woody understory species. Bird activity was influenced by ‘ōhi‘a flower abundance and the severity of koa defoliation; birds switched to outbreaking caterpillar prey, and they gained weight during the outbreak. Bat foraging times decreased during the outbreak, apparently because they became satiated quickly each night. Parasitoid wasps increased with caterpillar abundance but had little influence on outbreak dynamics. Reducing alien grass cover and increasing tree diversity would likely reduce the impacts of insect outbreaks and similar perturbations to native forests.

Testing how plant restoration influences animal taxonomic and functional diversity can shift restoration projects beyond mainly plant community considerations. We incorporated multi-trophic interactions into restoration by describing an ongoing functional trait-based restoration experiment in Hawaiian lowland tropical wet forest (Liko Nā Pilina Experiment), where litter arthropods are examined from a functional perspective thereby linking plants and higher trophic levels. We hypothesized that (1) communities with greater plant functional trait diversity would have cascading effects through food webs, increasing animal diversity and network complexity, and (2) increases in animal species and network complexity would be stronger for restoration efforts in plant communities with more complementary functional traits than those with more redundant traits. We examined experimental treatments of planted communities with the same species richness but with different plant functional trait profiles based on (1) rates of expected carbon turnover (slow or moderate), and (2) the similarity of their functional trait measurements (redundant or complementary), as determined by functional dispersion calculations. Initial data on arthropod communities and leaf litter decomposition rates revealed linkages between plant functional traits and arthropod community diversity. Overall, we argue that a more comprehensive evaluation of restoration accounts for both functional diversity and the multi-trophic nature of animal and plant communities. Developing restoration projects based on plant functional traits that influence both plant and invertebrate species provides a new paradigm, and the incorporation of both native and non-native (but non-invasive) plants shows promise in restoring ecosystem function in disturbed lowland tropical forests.

ABSTRACT In Hawaii and other oceanic islands with few native land mammals, black rats (Rattus rattus) are among the most damaging invasive vertebrate species to native forest bird populations and habitats, due to their arboreal behavior and generalist foraging habits and habitat use. We evaluated the nesting response of Hawaii Elepaio (Chasiempis sandwichensis; Monarchidae), a generalist insectivore, to the removal of black rats using rodenticide in a before-after-control-impact study in high- and low-elevation mesic montane habitat recovering from long-term damage from introduced ungulates and weeds. We monitored nesting success and rat abundance during 2015–2016 before applying rodenticide bait in 2017 to remove rats from two 700 × 700 m treatment plots that were paired with 2 nontreatment plots of the same size. Rat abundance was reduced by 90% during treatment, with combined variables treatment and elevation best explaining the change using GLM methods and AIC model selection. The daily survival rate (DSR) of nests (n = 191) was greater on treated plots after rodenticide application (mean ± SE = 0.980 ± 0.004 treatment; 0.964 ± 0.004 nontreatment), modeled nest success increased from 29% to 50%, and apparent nest success (number of successful nests per total nests) increased from 37% to 52%. The most informative model for predicting DSR included the effect of treatment. Predation by rats was documented at 3 of 16 nests using video surveillance, and we observed additional evidence of rat predation during in-person nest monitoring. Rats targeted adults on the nest and sometimes removed intact eggs, leaving little trace of their activity. Our results demonstrate that reducing rat predation can immediately improve the nesting success of even a common bird species in habitat with a long history of forest restoration. Sustained predator control may be critical to accelerating the recovery of native forest bird communities.

Biological invasions of rodents and other species have been especially problematic on tropical islands. Invasive Rattus rattus consumption of Hibiscadelphus giffardianus (Malvaceae; common Hawaiian name hau kuahiwi) fruit and seeds has been hypothesized to be the most-limiting factor inhibiting the critically endangered tree, but this has not been experimentally tested, and little is known about other factors affecting seed dispersal, germination, and seedling establishment. Thus, we do not know if rat removal is sufficient to increase hau kuahiwi recruitment. This study aims to evaluate the effect of rat population control on the ability of hau kuahiwi to retain fruit and establish seedlings. We compared hau kuahiwi fruiting and seedling recruitment in a stand treated to reduce rat abundance and a neighbouring control stand. Fruit retention increased following treatment but seedling establishment did not. Although rat control improves the ability of hau kuahiwi to retain fruit, other, presently unknown inhibitors to seed dispersal, germination, and/or seedling development remain. Seed and seedling predation by other species, competition from numerous invasive plant species, unsuitable climate, and/or other factors may be primary inhibitors in the absence of rats, but we emphasize that progressive isolation of these factors at individual hau kuahiwi life stages may be necessary to identify the remaining threats to the conservation of this critically endangered plant.

..............................................................................................................................