USGS2025
The Birds, Not Mosquitoes Monitoring and Support Science Working Group detailed methods for monitoring the population response of Hawaiian forest birds during implementation of the Incompatible Insect Technique (IIT) on the islands of Maui and Kauaʻi. The group prioritized methods for measuring the influence of mosquito suppression on populations within IIT treatment and control areas and identified focal species for IIT efficacy monitoring in birds. Three primary metrics were established to assess the impact of IIT on vulnerable species: population demography, density, and geographic range. Each metric can be evaluated using multiple methods. This report reviews those methods, with emphasis on approaches supported by pre-IIT baseline data and compatible with a before-after control-impact (BACI) study design for evaluating population responses over time. Focal avian species were selected based on population size estimates, fecundity, and disease susceptibility. We identified ʻākohekohe (Palmeria dolei), ʻiʻiwi (Drepanis coccinea), Maui ʻalauahio (Paroreomyza montana), Hawaiʻi ʻamakihi (Chlorodrepanis virens), Kauaʻi ʻamakihi (Chlorodrepanis stejnegeri), Kauaʻi ʻelepaio (Chasiempis sclateri), and ʻanianiau (Magumma parva) as focal species for monitoring population level response to disease suppression. Populations of kiwikiu (Pseudonestor xanthophrys), ʻakikiki (Oreomystis bairdi), akekeʻe (Loxops caeruleirostris), and the ʻiʻiwi population on Kauaʻi may be too small (e.g., <100 individuals) to effectively monitor, and it is unlikely that sufficient data can be collected from these birds to show IIT efficacy in a relatively short time frame (i.e., 5–10 years). Despite the logistical challenges to IIT implementation, there is potential to maintain disease-free status in individual populations of birds. Indeed, the continued existence of these critically endangered species in the wild within or near IIT treatment areas could be considered an accomplishment of IIT, given the current predictions for their extinction in the wild within 5–10 years. Demographic monitoring methods, including territory mapping, nest monitoring, mist-netting, and mark-recapture studies, provide direct evidence of survivorship and reproductive output. When combined with disease surveillance, these approaches could provide the most robust evidence of increased survivorship and productivity resulting from avian malaria suppression via IIT. However, demographic studies require several years of monitoring to achieve statistically robust BACI comparisons of survivorship and are more difficult to implement relative to other approaches. Given that these field efforts are labor-intensive and heavily reliant on personnel availability and funding, demographic monitoring could be conducted when adequate resources permit. On both Maui and Kauaʻi, passive acoustic monitoring (PAM) was identified as a priority method for monitoring the range, occupancy, and relative abundance of focal species. Autonomous recording units (ARUs) can record bird vocalizations in remote areas for several months. Innovative machine learning techniques permit rapid and semi-autonomous identification of most endemic honeycreepers on each island, maximizing sampling efficiencies and minimizing data processing costs. We predict mosquito suppression could support expansion of focal species into areas where disease transmission is currently excluding these species and expect acoustic monitoring data of focal species to reflect these spatial patterns. Additionally, the relative occupancy and call densities can be monitored temporally and spatially to assess the efficacy of IIT for supporting positive growth in vulnerable bird species. It is not yet clear if PAM is more effective than other methods, such as distance sampling, for detecting trends in the densities of rare species. However, the increased detections resulting from the larger sample size per observation point using ARUs will likely improve accuracy in detecting changes in species’ ranges. Collection of during and after treatment data within the BACI design could help to provide critical information to track avian population response, recovery, and potential range expansion related to IIT efforts. Point-transect distance sampling (point-counts) was prioritized as a method for monitoring population densities of focal species. Extensive historical sampling across focal species’ ranges provides a robust baseline for detecting change. These counts provide updated population densities and can be used to assess the distribution of focal species within IIT treatment areas. However, detecting subtle population changes with traditional distance sampling requires intensive spatial and temporal effort and may be less effective for rare species. To improve resolution, density surface modeling can integrate multiple data sources (e.g., point-counts, PAM, spot-mapping, and resightings) to estimate species-specific densities at finer spatial scales, including within and outside IIT treatment areas. This integrated modeling approach allows for detailed comparisons and may reveal early signs of recovery, including recolonization of formerly occupied sites. A coordinated monitoring strategy can allow managers to evaluate the success of mosquito suppression as a conservation intervention and support adaptive management in the face of emerging challenges.