Category Archives: Research Spotlight

History of the Center of Biostatistics and Modeling

 by Erin Leone

The Center for Biostatistics and Modeling provides statistical analysis advice and support to FWC staff. We are a service-oriented group that supports FWC’s mission much like the other Information Science and Management sub-sections such as the Center for Spatial Analysis, Information Access, and Research Information Sciences.

Our goal is to ensure the validity and continually improve the currency of the scientific studies produced by FWC. We provide general statistical consulting services, develop efficient data management strategies, design study and sampling plans, perform cutting-edge quantitative analyses, produce high-quality graphics and collaborate with staff on peer-reviewed publications. In addition to these services we offer software and analysis training to help staff feel more comfortable with their statistical knowledge base.

The center’s history began in the Lovett E. Williams Jr. Wildlife Research Laboratory in Gainesville back when it housed the Bureau of Wildlife Research under the Game and Fish Commission. It was the retirement of the lab’s namesake – Lovett William – in 1985 that started it all. Realizing the growing need for statistical support, research staff recommended that the commission replace Lovett’s turkey research position not with another biologist, but rather a statistician.

By the time of the commission’s reorganization and creation of the Fish and Wildlife Research Institute (FWRI) in 2004, the group consisted of two full-time statisticians in Gainesville. Thanks to the support of staff and leadership, the center has grown considerably and now consists of four statisticians and a dedicated database manager.

Staff in the Center for Biostatistics and Modeling have worked on a wide variety of projects touching every division within FWC and every subsection within FWRI.hypoxssdplot_withwaterbody So far this year we have worked on a number of interesting projects including examining the persistence of bald eagle nesting territories, describing determinant growth in American alligators, estimating species richness on habitat restoration sites, estimating black bear abundance using spatial capture-recapture models, and estimated the impact of red tide on sea turtle mortality. Some examples of our work are highlighted in the included figures. If you have questions about study design, analysis, or are in need of statistical assistance or expertise, please feel free to contact any one of us.

Erin Leone, Paul Schueller, Colin Shea

Statewide Wild Turkey Relative Abundance and Distribution Assessment Mapping Application

by Roger Shields and Tyler Pittman

wild turkeyIn 1973, following many years of wild turkey (Meleagris gallopavo) restoration efforts, the Game and Fresh Water Fish Commission (predecessor to the Fish and Wildlife Conservation Commission) conducted the first statewide assessments of the distribution of wild turkeys. In 2001, the FWC Wild Turkey Management Program (WTMP) conducted a similar map-based survey by mail for comparison. After this survey the WTMP included 10-year assessments of the distribution and relative abundances of wild turkey to FWC’s “Strategic Plan for Wild Turkey Management” to further monitor population trends of wild turkeys. We conducted this second assessment in 2011 using a specialized online Geographic Information System (GIS) and created an online mapping application for viewing survey results.

To begin the 2011 assessment, the WTMP and upland game bird staff at the Fish and Wildlife Research Institute (FWRI) worked with an outside vendor to develop an internet-based GIS mapping system. Resource specialists from FWC, other state and federal agencies, and industrial timber companies, as well as members of the Florida Chapter of the National Wild Turkey Federation, the Florida Chapter of The Wildlife Society and antlerless deer permit holders were asked to participate in the survey. Survey participants used personal computers and the online application to select specific cells of a 1 km x 1 km grid covering the state for which they were familiar and enter turkey abundance information. The system provided a hierarchy of underlying maps – statewide aerial photography; statewide rivers and major roads; and county, city and conservation area boundaries – to assist respondents in recording turkey abundance data geographically.

In total, 310 people responded to the survey; unfortunately, survey responses accounted for only about 65 percent of the state, and other means were necessary to acquire information about areas lacking survey data. To this end, WTMP biologists turned to staff at FWRI for additional assistance. Research staff developed a Likelihood-based Moving Window Model that used extrinsic data (including habitat suitability models and movement distance data) and information from known areas to inform estimates for unknown areas. For every cell of the statewide grid, the model evaluated the distance from known locations and pathways through suitable habitat from known turkey locations to determine the possibility (or likelihood) of at least one turkey being able to reach and occupy the cell in question. Staff then conducted field surveys to validate these underlying models and databases to ensure they accurately reflected conditions on the ground.

These efforts culminated in a spatial dataset that represents categories of predicted wild turkey presence throughout the state of Florida. The final step of the project was to build a web-based mapping application that allows for public viewing access to the assessment results.  This online tool is available at:

map of wild turkey distribution estimates
Figure 1. This map displays results of a 2011 statewide wild turkey distribution assessment. Note that this model provides estimates using best available information and these estimates do not replace ground surveys.

However, given that wild turkey are a generalist species and widely distributed, the prediction accuracy may not be exact in all areas. Although this information is useful for depicting the general location of turkeys within Florida, caution should be taken when considering local areas, because the survey resolution (1 km x 1 km) and data are not well suited for small spatial scale application.

Using the spatial data derived from this project, FWC staff will be able to relate current turkey population distributions (based on the 2011 data) to previous assessments, vegetative communities, land ownership, harvest records, and other pertinent information.  This will allow the FWC WTMP to focus management on particular areas of the State that have suitable turkey habitat but low turkey populations.

View an archive of past “Research Spotlight” articles

Habitat Research:
Coastal Wetlands

by Dr. Ryan P. Moyer


Coastal Acidification Studies in Tampa Bay

Women kneeling in field
FWRI coastal wetlands technician Christina Powell conducting coastal acidification sampling of Tampa Bay waters in Old Tampa Bay. Researchers collected samples every three hours over a 30-hour period to measure dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), pH, dissolved oxygen, temperature, salinity, and total nitrogen and phosphorus.

The coastal wetlands research group at FWRI has recently begun pilot field studies to understand the potential buffering effects that seagrass meadows may have on acidified coastal waters. The project was initiated in May 2014, when the coastal wetlands team joined a team of collaborators from the U.S. Geological Survey, ESA Associates, and the Tampa Bay Estuary Program to conduct diurnal sampling at two locations in Tampa Bay. This work uses cutting-edge autonomous seafloor sensors with discrete measurements of water column chemistry to understand the hydrodynamics and biogeochemistry of seagrass habitats in Old Tampa Bay offshore of Rocky Point, and in Lower Tampa Bay offshore of Fort DeSoto. Parameters measured over a continuous 30-hour period include dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), pH, dissolved oxygen, temperature, salinity, current speed and direction, and total nitrogen and phosphorus.

Coastal waters experience much higher diurnal geochemical variability than open ocean waters due to the high rates of photosynthesis and respiration that occur in coastal estuaries such as Tampa Bay. As the concentration of atmospheric carbon dioxide (CO2) increases, some excess CO2 is transferred to surface waters, thereby lowering pH and resulting in a phenomenon known as ocean acidification. Since coastal waters have higher rates of respiration (produces CO2) and photosynthesis (consumes CO2), they may serve as a natural buffer area for organisms to adapt to acidified waters. Further, seagrass acreage in Tampa Bay has been slowly recovering since the 1980s and is now the highest it has been since the 1950s. This increase in seagrass means more photosynthesis, and less CO2 in Tampa Bay waters – a trend that is opposite of the expected acidification trend (lower pH) that has been observed in most other coastal and open ocean areas. Thus, Tampa Bay may provide insights towards utilizing strategic habitat restoration to minimize the impacts of ocean acidification in coastal waters. The coastal wetlands research group at FWRI, along with our partners, is trying to understand this unique geochemical occurrence in Tampa Bay, and hope to develop a Bay-wide coastal acidification monitoring program for Tampa Bay in collaboration with the Environmental Planning Commission of Hillsborough County.


Vegetation Monitoring at Clam Bayou

Woman on boat
FWC intern Alexandra Wilcox (left) and coastal wetlands technician Kara Radabaugh (right) measure salt marsh soil pH at Clam Bayou in Gulfport, FL, while coastal eetlands technician Amanda Chappel (center) records the data.

The FWRI coastal wetlands research team also continues to monitor water quality and vegetation in mangrove and salt marsh habitats of Clam Bayou in Gulfport, FL. This work compares commonly used monitoring techniques in tidal coastal wetlands in support of the Coastal Habitats Integrated mapping and Monitoring Program (CHIMMP). The project was initiated in July 2014, when water quality monitoring activities began as part of a consortium with researchers from the University of South Florida College of Marine Science, the U.S. Geological Survey, and YSI Inc. Then in December 2014, permanent vegetation monitoring plots were established to compare several different field-based monitoring techniques. Water quality monitoring activities have continued monthly, and vegetation monitoring has been conducted every four months. Three salt marsh transects and three mangrove plots are visited quarterly. The following parameters are measured: sediment accretion, soil moisture, soil pH, porewater salinity, tidal creek salinity, temperature, dissolved oxygen, and pH, vegetation density, canopy height, canopy density, vegetative species identification, and mangrove diameter. To date, the monitoring efforts have documented a major shift towards mangrove vegetation in salt marsh habitats at Clam Bayou. Information gained from this study will help make recommendations for the development of future coastal wetland monitoring programs statewide. Coastal wetlands are particularly vulnerable to natural and anthropogenic impacts on local hydrology and water quality, which in turn determine the suitability of the habitat for numerous fish and bird species. The wetlands found at Clam Bayou are especially vulnerable due to their proximity to urbanization and altered hydrology.


View an archive of past “Research Spotlight” articles

Scanning the Ocean Floor for Fish Habitat

by Eric Weather

Maximizing the Efficiency of Reef Fish Surveys Through IncorporaThe Gulf of Mexico benthic habitat can be summarized as a mosaic of sand with limestone rock outcroppings (reefs).  These reefs facilitate the settlement and growth of a wide variety of benthic organisms including corals, sponges and crustaceans.  They also support a diverse fish community from tiny blennies and gobies to commercially and recreationally important groupers and snappers.  The Fisheries-Independent Monitoring (FIM) program in cooperation with the National Marine Fisheries Service (NMFS) conducts annual fisheries monitoring on these reef habitats to provide data for single species and ecosystem-based management initiatives.  In 2009, the FIM program implemented the use of side-scan sonar to identify and classify the benthic habitats in the eastern Gulf of Mexico between 10 m and 110 m deep.  These sonars are towed behind a survey vessel and are equipped with two side-facing transducers that generate and receive sound signals.  When a signal is generated it propagates through the water column and reflects off the seafloor and then back to the transducer.  The intensity at which the return signal is received is interpreted into an image that is deciphered by a trained survey technician.  The configuration of the transducers on the sonar allows the sound signal to ensonify a very wide swath of the seafloor (up to 300 m) at one time.  As the survey vessel moves through the water, a streaming image of the seafloor is generated, as depicted in the video.  When rocky outcrops are identified by a survey technician they are given a habitat category based on their structure and complexity and become the basis for the NMFS/FIM reef-fish survey.  Additionally, these data are used by regional geologists, cartographers and other biologists to help answer a wide variety of research related questions regarding the benthic habitats of the eastern Gulf of Mexico.