Project Work Plan

Greater Everglades Science Program: Place-Based Studies

Project Work Plan FY 2003

Project Title: THE INFLUENCE OF HYDROLOGY AND ASSOCIATED HABITAT STRUCTURE ON SPATIAL AND TEMPORAL PATTERNS OF APPLE SNAIL ABUNDANCE AND RECRUITMENT
Project start date: 2/2/02 Project end date: 1/31/05
Project Funding: University of West Florida (Excludes USGS indirect costs)
Principal Investigator: Philip C. Darby, Assistant Professor
Email address: pdarby@uwf.edu
Phone: 850-474-2647 Fax: 850-474-2749
Mail address: Department of Biology, University of West Florida, 11000 University Parkway, Pensacola, Fl 32514

Other Investigator(s): Dr. Robert E. Bennetts
Email address: rbennetts@usgs.gov
Phone: 352-378-8181 x 374 Fax: 352-378-4956
Mail address: Florida and Caribbean Science Center, 7920 NW 71st Street, Gainesville, FL 32653

Project Summary: The proposed study, for the first time, will provide data linking snail abundance to hydrology and habitat structure using validated sampling methods. We will also collect data on crayfish and freshwater shrimp abundance in the same habitats. In particular, the study targets the need to predict the impacts of the WCA-3A Decompartmentalization project (CERP Science Objective Number 3007-12) on apple snail abundance. WCA-3A decompartmentalization will alter the hydrology of WCA-3A (frequently used by kites in recent years) and the North East Shark River Slough. In general, the abundance and recruitment data will serve as key Everglades restoration performance measures that can be linked to changes in habitat structure and wildlife habitat use. Given the inextricable link between hydrology and vegetation (previously documented) and apple snails (the objective of this study), our research effort would also support the development of the link between snail kite habitat use and hydrology (CERP Science Objective Number 3007-10). As prey for wetlands wildlife other than snail kites (e.g., limpkins, white ibis, alligators) our data may also prove useful for understanding the affect of hydrology on other species of particular interest in the restoration. Our objectives tie directly to the need for the RECOVER Teams to develop monitoring programs for indicator species that respond to restoration activity.

Project Objectives and Strategy: The overall goal of the project proposed herein is to provide information to help predict the impact of changes in hydrology (and associated vegetation) to apple snail abundance following WCA-3A decompartmentalization. The specific objectives for the proposed project are as follows: 1) estimate apple snail abundance in wet prairie habitats (relatively short hydroperiod) relative to slough habitats (relatively longer hydroperiod) with potential linkages to snail kite foraging habitat selection; 2) estimate snail abundance in sites with foraging snail kites, and incorporate data from this study in to a larger coordinated effort (see below) to link snail kite habitat use and snail availability; and 3) document baseline seasonal apple snail egg production (not yet available in wetlands south of Lake Okeechobee) and assess the potential impacts of drying events on annual egg production.

Potential Impacts and Major Products: The information derived from this project will directly link hydrology to the distribution and abundance of apple snails among different habitat types. This project will provide information needed by managers responsible for achieving restoration goals that include providing suitable conditions for apple snails, snail kites and other snail predators. A six-month progress report (original + 2 copies + electronic version) will be submitted following contract initiation to document success in establishing sampling sites, monitoring protocols, protocols for linking snail sampling data to snail kite distribution, and personnel training. Meta-data files and progress reports (original + 2 copies + electronic version) will be submitted annually thereafter. Manuscripts intended for peer-reviewed journals will be identified in progress reports as data accumulates, and prepared manuscripts will be available for journal review within 18 months of contract termination.

Time Frame:

February 2, 2002 — January 31, 2003

• sample site selection
• develop sampling protocols
• snail, crayfish and freshwater shrimp abundance (established)
• plant community structure (e.g., stem counts)
• egg cluster transects (established)
• linking snail kite foraging success to habitat type + snail density
• preliminary analysis of data collected in 2002

February 2, 2003 — January 31, 2004

• continue sampling in same areas established in 2002
• relocate centers of snail kite activity and link to habitat/snails
• analyze data collected during first two field seasons (include year effect)

February 2, 2004 — January 31, 2005

• finish sampling in same areas established in 2002.
• relocate centers of snail kite activity and link to habitat/snails
• complete data analyses and final report

Collaborators: na
Clients: na

B. WORK PLAN

All tasks in the work plan are linked to the density/distribution data collected using 1-m2 throw traps (including the egg clusters, which are sampled primarily to monitor the timing of the reproductive season). Therefore the work plan is described under one task.

Title of Task 1: same as project
Task Funding: see project funding
Task Leaders: Phil Darby and Rob Bennetts
Phone: see P.I. information
FAX: see P.I. information
Task Status (proposed or active): active
Task priority:
Budget and Time Frame for Task 1: requires all 3 years of the contract
Task Personnel: Jennifer DuPree, Jason Liddle, OPS seasonal (TBD), UWF students (TBD)

Task Summary and Objectives: same as for the project

Work to be undertaken during the proposal year and a description of the methods and procedures: We will conduct a 3-year study in Water Conservation Area-3A (WCA3A) and Northeast Shark River Slough (NESRS)(if accessible). If NESRS is not accessible (the case in spring 2003) we will sample additional sites in WCA3A. We will determine snail abundance each year in five randomly selected wet prairie (same 5 sites each year) and five slough habitats (same 5 sites each year) using 1-m² throw traps (n 50-70 throw traps in a 50 x 50 m area) per Darby et al. (1999). The throw trap data will be used to test the hypothesis that apple snail abundance is greater in wet prairie versus slough habitats (as suggested by snail kite habitat use, Rob Bennetts, pers. comm.). Site selection and classification will be based on habitat structure and the best available hydrologic information for the areas. The timing of all sampling will depend, in part, on hydrologic conditions and seasonal patterns in egg cluster production. Most throw trap sampling should be conducted from approximately February through April. Sites that are dry during this period will be sampled following reflooding (surviving snails emerge from aestivation within 24 h). Egg cluster surveys (see below) will provide data to avoid sampling during the period of greatest post reproductive mortality. Protocol modifications may be warranted to document unusual hydrologic events (e.g., extended or atypically timed dry downs or floods).

We will characterize plant community composition (% species coverage for the area), substrate composition, plant stem density (1-m² quadrat samples, n=7) and periphyton abundance in all sampling sites. We will attempt to control for variation in all aforementioned habitat characteristics during sample site selection.

Three wet prairie/slough habitats known to support foraging snail kites (locations provided from our own surveys or by the snail kite monitoring field crew; P.I., Wiley Kitchens, USGS-BRD) will also be sampled for snail abundance using throw traps. Comparisons to the randomly selected sites within a wetland unit (see above) will be made to see if kites forage in sites with relatively higher snail abundance. However, understanding how snail kites respond to apple snail abundance requires comparisons to foraging habitats spread throughout the snail kite's range. Snail kites are highly nomadic, even in generally high-water years (Bennetts and Kitchens 2000), but no snail abundance data has been linked to kite movements. Data from this 3-year study will be examined as part of a coordinated effort [via concurrent studies in the Loxahatchee N.W.R. (USFWS funding in place; P.I., Phil Darby), Kissimmee Chain-of-Lakes (FFWCC funding in place; P.I,. Phil Darby), and potentially other areas (proposal in development by Rob Bennetts, USGS-BRD)] in order to correlate snail abundance with snail kite habitat use. Such coordination enables us to increase our scope of inference across areas with a range of hydrologic regimes and varying substrate types (e.g., sand versus peat) and over a temporal and spatial scale relevant to the nomadic snail kites' foraging range.

Apple snail egg clusters provide critical information about the timing of the annual post-reproductive adult die off (Darby and Percival 2000). A 1m x 2.5m PVC quadrat will be flipped end over end 20 times (CV stabilizes around 25%; Darby, unpublished data) in order to count egg clusters in each transect. Two egg cluster transects will be established in all of the aforementioned throw trap study sites (thus incorporating a hydrology effect) and sampled monthly from February - September each year (n=26 transects each month). Finally, 2 egg cluster transects will be established in each of 3 sites subject to frequent drying events and known to contain apple snails (e.g., portions of NESRS and northwestern WCA3A) in order to quantify recruitment losses related to drying events and/or potential resumption of oviposition after water levels rise following a dry down.

Apple snail density data from an 50 m x 50 m area tends to be distributed as a negative binomial (Darby et al. 1999). A likelihood ratio testing approach (per White and Bennetts 1996, see snail data in Darby et al. 1999) will be used to model apple snail density as a function of community type, substrate type, snail kite presence or absence (or possibly some index of foraging success or extent of kite use, if data were available) and time (+ interactions). Final model selection will be based on comparing Akaike's information criteria (AIC) values from a suite of candidate models (Burnham and Anderson 1998). We will evaluate a suite of candidate models that include each main effect separately as well as all possible combinations of interactions. Egg cluster counts will be analyzed using a generalized linear models approach as described in Darby and Percival (2000). The model for egg cluster production will include temperature, water level, community type, substrate type, and time (+ interactions) as independent variables. Final model selection will be based on an AIC approach as described above.