Realm: Marine
Climate: Temperate
Biome: Temperate shelf and seas ecoregions
Central latitude: 29.188063
Central longitude: -83.279223
Duration: 2 years, from 2009 to 2010

5029 records

319 distinct species

Across the time series Lagodon rhomboides is the most frequently occurring species

Methods

B.1. Site selection and evaluation: We used a design that allowed for an equal distribution of sampling locations. Spatially-balanced sampling approaches have recently been developed (e.g. Stevens and Olsen 2004 Olsen et al. 2012) that incorporate similar characteristics as random or systematic ones but also guarantee that all samples are evenly distributed across the entire sampling frame (rather than clumping commonly associated with random sampling or distinct intervals associated with systematic sampling). Sites were chosen from Floridas Statewide Seagrass polygon Geographic Information System (GIS) data set compiled by the Florida Fish and Wildlife Conservation Commission-Fish and Wildlife Research Institute Center for Spatial Analysis (FWRI 2007). We used an Albers Equal Area projection for the shapefile. The file which includes all seagrass coverage across the state of Florida was clipped to include only the Big Bend region from the St. Marks River in the north and the Anclote River in the south. From the clipped polygon we then chose sites using the spsurvey package (Kincaid et al. 2008) in the R software environment (R Development Core Team 2008). We choose a target of100 sites per sampling season (example R code for 2009 site selection provided in BBSG_2009-2010_site_selection_code_2009.txt; see Class V.C.). However because we anticipated that sampling could not be conducted at some sites (e.g. due to lack of target habitat logistical constraints) we also chose additional oversampling sites to replace target sites that had to be dropped from the initial sample. Based on pilot trials conducted in the northern region of our study area during the summer of 2008 we chose a 23% oversample for our 2009 season (100 target + 23 oversample). Based on the 2009 sampling season we chose a 27% oversample in 2010 (100 target + 27 oversample). Ideally sampling would proceed by first visiting all target sites then visiting the required number of oversampled sites to meet the target sample size. However this two-step approach was not possible in the current study given the large area over which sites were located. We therefore visited all oversampled sites as we visited target ones (i.e. we did not back track). Fourteen of the 123 sites that were drawn in 2009 and nine of the 127 sites from 2010 were dropped due to navigational hazards; typically this meant the site was located in a channel used by other boaters was located in extremely shallow water that we were unable to access or trawl or there was a physical barrier such as an oyster bar or shoal that precluded access. The remaining 109 sites from 2009 and 118 sites from 2010 were evaluated for suitable habitat to conduct sampling.Once we arrived at a site we determined whether it could be sampled based on two criteria: (1) it had to include ? 10% seagrass cover and (2) there were no hazards present to the trawl (e.g. due to presence of rocks or other hard substrate) the research vessel or researchers (e.g. due to navigational hazards such as extremely shallow water) or other boaters (e.g. due to presence of other boats anchored or heavy boat traffic). If the site did not meet both criteria we began piloting the research vessel in a spiraling manner around the site location to determine if there was suitable habitat nearby. We continued the spiraling until we had traveled 250 m from the original site (i.e. 500 m diameter circle 196350 m? area around site). If no suitable habitat was found we dropped the site and noted the reason. Among the sites evaluated 91 of the 109 in 2009 and 79 of the 118 in 2010 met the two criteria. The primary reason for a site not meeting both criteria was that it had less than 10% seagrass cover (usually with no seagrass) often with other floral (e.g. attached macroalgae) faunal (e.g. octocorals sponges) or geological characteristics (e.g. rocky reefs) that may have been incorrectly interpreted as seagrass from the aerial photos used to inform the GIS shapefile.B.2. Faunal data collection: If suitable habitat was found either at the site or from our spiraling search we sampled it using beam trawls (mouth opening: 1.87 meter (m) wide by 0.40 m tall; bag dimensions: 19 mm mesh with 3 mm mesh liner) towed on both sides of a 6.1 m research vessel (Fig. 2). Beam trawls were the most appropriate gear for our study for several reasons. First water depths in the region exceeded 5 m at some sites precluding the use of seine nets and drop traps that are commonly used in shallow seagrass beds (Edgar et al. 2001). Additionally we sought to sample across a large area within each site thus requiring a towed trawl gear. Beam trawls are advantageous over otter trawls (which is a commonly used towed gear in seagrass systems) because they retain a constant sampling volume during tows both across sites (otter trawl sampling volume can be affected by bottom substrate which varies across the study region) and within them (otter trawl sampling volume decreases as the net fills because the doors are pulled inward) (Rozas and Minello 1997 Stallings et al. 2014a). Higher and less variable capture efficiencies in beam trawls have also been attributed to the fixed mouth opening (Kuipers et al. 1992). Indeed Zimmerman et al. (1986) demonstrated that beam trawls had markedly higher capture efficiencies of penaeid shrimp (a common seagrass fauna) than seines and otter trawls. Additionally otter trawls tend to ride on top of submerged aquatic vegetation allowing animals to escape underneath the net resulting in low capture efficiency (Leber and Greening 1986 Stallings et al. 2014a). Conversely we demonstrated in pilot trials that the beam trawls remained in contact with the seafloor while being towed at varying depths on different substrates (e.g. soft sand compacted sand) and in different seagrass compositions and areal coverage. We towed the trawls from the sides of the boat (instead of astern) to eliminate problems with prop-wash on catch rates in shallow water. The trawls were towed for 2.5 minutes at a standard rate of 1.8?2.0 km/h (ca. 85 m). This tow speed (based on direct underwater observation) was determined to be the most efficient; faster tows caused the net to ride off the bottom and slower tows allowed animals to escape (personal observation during gear trials). The towing direction was determined a priori with a random number generator from 1-360 (i.e. the heading direction in degrees). However the trawl gears operate most efficiently and safely (i.e. avoiding contact with boat motor propeller) when towed downwind so we adjusted accordingly. The locations of the start and end of each tow were marked in decimal degrees using a Global Positioning System (GPS) which allowed us to measure tow distance and calculate the area sampled.All captured animals were identified to the lowest possible taxon and most were counted and measured in size classes. We did not count or measure the sizes of colonial fauna (e.g. ectoprocts ascidians) the tests of tube worms (e.g. sabellariids) or the shells of worm snails (e.g. vermetids) but we noted their presence. Among the major groups of fauna observed we were able to identify 93.4% of the taxa to species for actinopterygian fishes (92 taxa) 85.2% for bivalves (34 taxa) 85.0% for echinoderms (20 taxa) 76.8% for gastropods (95 taxa) and 70.1% for arthropods (87 taxa). Due to the large quantity of animals that had to be measured we used size classes to expedite the process. Based on previous research using trawl gears in seagrass beds (e.g. Stallings et al. 2010) we used six size classes for the current study: (1) 1-25 mm (2) 26-50 mm (3) 51-75 mm (4) 76-100 mm (5) 101-150 mm (6) ? 151 mm. We measured the total length for teleosts carapace width for crabs post-orbital head length for shrimp diameter for echinoderms and longest plane for molluscs. Once animals were identified counted and measured they were then released back into the water. Care was taken to minimize harm to the animals. Some animals were retained to allow for identification in the laboratory where microscopes and detailed taxonomic keys were present (keys used: Schultz 1969 Morris 1973 Hoese and Moore 1977 Lindner 1978 Abbott et al. 1985 Abele and Kim 1986 Robbins et al. 1986 Kaplan 1988 Littler 1989 Kensley and Schotte 1989 Humann and DeLoach 1992 Humann and DeLoach 1994 Abbott et al. 1995 Hendler 1995 Hoese and Moore 1998 McEachran and Fechhelm 1998 Voss 2002 Pomory 2007 Mikkelsen and Bieler 2008; supplemental online resources used: http://www.algaebase.org/ http://www.fishbase.org/ http://www.gastropods.com/ http://www.itis.gov/ http://www.jaxshells.org/ http://www.marinespecies.org/ http://www.sealifebase.org/ ).In addition to fauna the catch in trawl nets often included seagrass blades and drift algae. It was important to remove animals that were attached to the seagrass and entrained in the drift algae. During the 2009 field season we did this by vigorously rinsing all plant materials in 19 liter buckets filled with seawater. The seawater was then drained through a 0.5 mm wire sieve and remaining materials which were mostly fauna mixed with sediments and plant debris were placed in a plastic bag labeled with site location. The sample was frozen and animals were identified and measured in the laboratory.The seawater rinse did well to remove most animals attached to the plant materials. This was readily confirmed by visual inspection of seagrass blades where no animals remained attached after the rinse. However we were concerned at the end of the 2009 field season that a more careful approach was needed to accurately count the number of very small arthropods and molluscs commonly associated with the drift algae. We therefore slightly altered our protocols to separate fauna from drift algae for the 2010 field season. We continued to use a vigorous seawater rinse of seagrass blades in 2010. For the drift algae we would first separate all large fauna (e.g. most fishes large crabs) leaving only the smallest arthropods and molluscs. The total volume of the drift algae captured was then measured using a calibrated 19 liter bucket. In 2009 the median volume of drift algae per tow was 15142 mL. In 2010 we collected a 3785 mL subsample of drift algae which was approximately 25% of this median value. The 3785 mL subsample was taken by haphazardly grabbing small clumps from the algae collected in the trawl. We used this haphazard selection to avoid subsampling a single small section of the algae. The entire subsample was bagged labeled frozen and taken back to the laboratory. In the laboratory the algae was first vigorously rinsed in 19 liter buckets containing seawater and remaining sample sieved as was done during the 2009 field season. The rinsed algae was then carefully inspected for the presence of remaining fauna. As suspected small arthropods and molluscs remained entrained in the algae. These animals were meticulously removed by picking each individual with forceps. Animals were then identified counted and measured. Site abundances for these subsampled animals were calculated as the number of animals counted multiplied by the percent of subsampled algae (total volume of captured drift algae / volume of subsampled drift algae). Unit of abundance = IndCountInt, Unit of biomass = NA

Citation(s)

Stallings, C. D., Mickle, A., Nelson, J. A., McManus, M. G. & Koenig, C. C. (2015) Faunal communities and habitat characteristics of the Big Bend seagrass meadows, 2009–2010. Ecology, 96, 304.