Predicting coral disease outbreaks in the Pacific Ocean

The goal of the proposed research is to develop better models of coral disease outbreak risk across the western tropical Pacific Ocean and embed these improved forecasts into the NOAA Coral Reef Watch decision support system (DSS). The current DSS was developed based on ~50 km SST data, and it assesses coral disease risk for the Great Barrier Reef and the Hawaiian archipelago using (i) a Seasonal Outlook from winter-time conditions that is (ii) updated in near real-time through the summer period. The ecological models used in the experimental disease risk DSS were derived by assessing the relationship between sea surface temperature (SST) anomalies and a single coral disease, Acropora white syndrome, on the Great Barrier Reef. We propose to improve these disease forecasts by (1) increasing the spatial resolution of SST-based predictions tools to 5 km, through application of NASA/NOAA satellite products, and expanding the application to several different coral diseases, host species and regions; (2) incorporating Short-Term Forecasts (four-month projections) of SST to augment the post-winter risk; (3) incorporating ocean color data from NASA/NOAA satellite products into the models, if robust relationships are found; and (4) undertaking additional coral disease observations, using targeted surveys and innovative bioassays that provide measurements with enhanced sensitivity. The disease forecasting models will be developed for locations in the western tropical Pacific Ocean that have existing long-term (~decade-long) observations of coral disease (i.e., Hawaii, Australia, US affiliated Pacific Islands), but the approach could subsequently be replicated in other coral reef regions. The proposed research addresses several goals of the Earth Science Research Program by developing a system to better predict changes in coral health in the face of ongoing climate change, and enabling researchers, managers and the public to use these forecasting models to better understand and respond to future disease outbreaks. This project specifically addresses the ROSES 2016 research announcement A.46 section 3.1.3, Managing Marine Ecosystems in a Time of Changing Climate through Better Forecasts, by using time series of biological observations (14 years of coral health surveys, physiological data from bioassays) and time series of climate observations (e.g., sea surface temperature anomalies, chlorophyll-a concentrations) to develop an ecological modeling framework that predicts coral disease risk based on ongoing satellite-derived inputs and accounts for uncertainties in climate-disease relationships and uncertainties in the estimates of disease risk.

Bioerosion in the Hawaiian Archipelago

The vulnerability of coral reefs to climate change, eutrophication, sedimentation, and other 603_2012_movie-001human-related threats has created an eruption of studies focusing on the impact of human activities on coral reef health. Historically, coral reef research has focused on the corals themselves and the resource species in which they support. However, many of the less charismatic cryptofauna that live within the cracks and crevices of the reef play a vital role in the reefs ability to bounce back from acute or long lasting disturbances. Many of these cryptic organisms are bioeroders — animals and algae that naturally breakdown the calcium carbonate skeleton formed by corals. Understanding the baseline bioerosion rates on reefs and how these rates may change in the future is important because reefs will only persist if bioerosion rates are slower than coral growth, or accretion rates […]

CAMEO: Multiscale modeling of coral communities

A key challenge in the effective management of marine ecosystems is translating from small-scale studies of distribution and dynamics to the regional scale of management action.  In many marine ecosystems, including the Hawaiian Archipelago, there are extensive survey data of nearshore communities from multiple investigators, representing a huge investment of resources.  Often, these data are underutilized and remain of limited use to managers.  In the Hawaiian Archipelago, at least seven separate entities are engaged in surveys of coral reef communities, with varying degrees of coordination.  The synthesis of these data requires integrated modeling approaches at multiple scales […]

Monitoring invasive species in the Northwest Hawaiian Islands

There are >200 marine introduced species in the main Hawaiian Islands.  Some of these species are highly invasive, spreading from points of introduction (e.g., Pearl Harbor) throughout the state and having substantial impacts on native species and ecosystem function.  In the Papahanaumokua Marine National Monument, fewer than 20 introduced species have been documented.  Of the documented species, most are  centered around Midway, which has been the center of maritime activities in the NWHI, and are most common in anthropogenic habitats, like docks and constructed islands.  I am collaborating with Scott Godwin, the invasive species specialist for the Monument, to design a monitoring protocol for the NWHI that assess the current distribution of known introductions, detect increases in densities or spatial extent of known introductions, and 3) detect new introductions.

Scaling-up Petrolisthes cinctipes dynamics

In marine communities, planktonic larval dispersal decouples local fecundity from local settlement, and there is often large spatial variation in settlement. While both larval supply and post-settlement processes contribute to local dynamics, the interaction between them has received less attention. Spatial theory demonstrates that this interaction between local nonlinearities (post-settlement processes) and spatial heterogeneity (e.g., variation in larval supply) is critical to scaling from local to regional dynamics. To empirically investigate this interaction, I studied Petrolisthes cinctipes, an intertidal porcelain crab that occurs in cobblefields of the northeast Pacific.

For P. cinctipes and its predators, there are three important scales of variation: individual variation in size, rock-to-rock variation in density, and site-to-site variation in density, larval supply, and predation. Combining field surveys of spatial variation with field and laboratory experiments to estimate a rock-scale population model, I predicted regional population dynamics of P. cinctipes using scale transition theory.  Scale transition theory identifies the relative importance of individual, rock, and site-scale variation and covariation and on the regional scale.  Regional population dynamics differed from local dynamics: (i) gregarious settlement enhanced regional recruitment, (ii) regional recruitment was lowered by positive covariance between predators and larval supply, and (iii) intraspecific competition had a stronger impact at the regional scale due to rock-scale aggregation.  Overall, spatial heterogeneity decreased population growth rates at the regional scale compared to the local scale.

Allee effects, conspecific cueing, and gregarious settlement

Population models often assume that organisms disperse or settle without regard to habitat quality, competition, or other environmental parameters.  Clearly, this is not the case, and these behaviors change the spatial variances and covariances that are so critical in spatial population models. In Petrolisthes cinctipes, larvae settled gregariously even though experiments demonstrated strong, size-dependent intraspecific competition in P. cinctipes.  One explanation is that adults provide some direct benefit to settlers, such as decreased predation.  However, an optimal habitats P. cinctipeselection model, indicated that  I demonstrated that such a direct benefit is unlikely to generate the pattern of settlement that I found in  , and that settlers must be responding to adults as an indicator of habitat quality.  The next step in this research is to develop a general framework for considering the covariation between habitat quality and conspecific density in habitat selection models.  Donahue, MJ. 2006.  Conspecific cueing and growth-mortality tradeoffs jointly lead to conspecific attraction.  Oecologia. 149: 33-43.

Predator foraging and prey distributions


Working with Drs. Bob Desharnais and Carlos Robles, we are using a multiple-model approach to investigate how predator foraging behavior influences patterns of size- and spatial- structure in mussel beds. Using geo-referenced images of mussel beds from Barkley Sound, we are testing the predictions of spatial models against the empirical patterns formed in the field. Robles CD, RA Desharnais, C Garza, MJ Donahue. 2010. Landscape patterns in boundary intensity: a case study in mussel beds.  Landscape Ecology. 25: 745-759. Robles CD, RA Desharnais, C Garza, MJ Donahue, and CA Martinez. 2009. Complex equilibria in the maintenance of boundaries: Experiments with mussel beds.  Ecology. 90(4): 985-995.