John W. Jones¹ and William Loftus^2
1US Geological Survey Eastern Geographic Science Center, Reston, VA, USA
²US Geological Survey Florida Integrated Science Center-Water and Restoration Studies, Homestead, FL, USA

Aquatic animals in southern Florida wetlands must cope with seasonal environmental changes, often by moving among habitats to find refuge from drying. The highly eroded, karst landscape of the Rocky Glades in Everglades National Park once offered dry-season refuge to aquatic animals in solution holes that were conduits to groundwater before water levels were lowered by drainage. Today, the refuge quality of many solution holes has been reduced, but this habitat is the focus of scientific study for restoration. However, we need the ability to extrapolate ecological data from intensive, site-based studies to the greater landscape before the consequences on aquatic animals of manipulating groundwater levels can be forecasted. Such extrapolation requires an estimate the density of holes on the landscape, and an estimate of their depth distributions. Topographic-data planes collected for the Rocky Glades before this pilot study are too coarse for models of solution-hole habitat use by animals. Standard land-surveying techniques are too time intensive and cost prohibitive to use in determining solution-hole densities and depth distributions at the landscape level. Even if standard survey techniques were cost effective, their widespread use might compromise the wilderness features of the study area. Remote sensing is the only feasible way to acquire needed solution-hole data.

The goal of our research was to develop and evaluate cost-effective methods for deriving useful information on solution-hole distribution, density, and depth in test areas of the Rocky Glades region. For this pilot study, several different airborne-image data sets were collected, including color-infrared, natural color, and Light Detection And Ranging (LIDAR). Machine-based, image-processing techniques were emphasized as the ultimate endpoint because they provide objective, systematic, and replicable methods of surface-feature mapping and are generally lower cost, less labor-intensive, and less subjective then visual-interpretation techniques. Once we had identified sets of solution holes and characterized them to the greatest extent allowed by the imagery and techniques, we checked the derived characteristics for specific solution holes against field observations on solution-hole location, size, depth, and vegetative/geologic structure. We found that correct timing and adequate spatial resolution of optical remote sensing image collection are critical and fusion of these and LIDAR data are needed to fully characterize solution-hole habitat. Finally, we fused these multi-dimensional data and created tools to allow interactive solution-hole habitat visualization and remotely sensed data exploration.

This work was funded by the USGS/NPS Park-oriented Support Program, by USGS Greater Everglades Priority Ecosystem Science and Land Remote Sensing Program funding to Jones, and USGS base funding to Loftus. The NPS CESI program also provided support.

Contact Information: John W. Jones, USGS, 521 National Center, Reston, VA, 20192 USA, Phone: 703-648-5543, Fax: 703-648-4165, Email: jwjones@usgs.gov

(This abstract is from the 2006 Greater Everglades Ecosystem Restoration Conference.)

Back to Project Homepage