Luquillo LTER

Chapter 1 (fish community data): In June and July of 2004, eight stations (see figure above) were sampled once per month between 0700 and 1100 hrs. On each sampling event, four experimental 100 x 8 ft nylon sinking gill nets, each of a single mesh size (½, 1, 2, and 3 inches square), were deployed to capture fish at each station. Each net was anchored to the shore and deployed at a 45-degree angle sloped towards the freshwater flow. The 2” and 3” nets were placed on opposite shores at extremes of the sampling station and the ½” and 1” nets were placed on opposite shores between the larger nets. Nets were set for 1.5 hours. Dip nets were used to collect smaller fishes along the shores. All fish were weighed and measured for total and standard lengths. To increase the sample size and capture of crepuscular fishes, additional night sampling (between 1900 and 2300 hours) was carried out at least once per month at all sampling stations. During each sampling event, surface (at 0.25 m) and bottom (at 0.25 m above the substrate) temperature, salinity, dissolved oxygen, and turbidity were recorded from the middle of the channel with a Hydrolab Quanta (Hydrolab Inc.). 1977 fish community data can be found in Corjuo Flores (1980).

Chapter 2 (isotope and gut content data): See Chapter 1 methods for water chemistry collection.

Sample collection
Fish for isotope analysis were collected as described in chapter 1 methods from the stations indicated in the Chapter 2 location map. Isotope sample collected from Espiritu Santo stations 2,5,8 in Chapter 1 are referred to as Espiritu Santo stations 1,2,3, respectively in Chapter 2 (to facilitate comparisons between the two estuaries). Isotope analysis was focused primarily on two pelagic species (Centropomus pectinatus and Bairdiella ronchus) and two benthic species (Diapterus rhombeus and Mugil curema).

In addition to the above fish sampling, juvenile freshwater shrimps (Atya lanipes, xiphocaris elongata, and Macrobrachuim spp.) that were observed in dense congregations beginning their upstream migration were collected with dip nets from littoral areas of both estuaries (note: freshwater shrimp were not observed at all stations).   In addition, adult freshwater shrimp were collected at the lowest elevation water intake site in the Espiritu Santo and Mameyes Rivers (located approximately 0.5 km upstream of the fresh/salt water interface in each case).

Tissue samples for isotope analyses were collected from white muscle of these fishes and tails of shrimps. All samples were rinsed with distilled water, placed in a sealed glass vial, and immediately frozen. Tissue samples were freeze-dried (Virtis Freezemobile 35ES) for at least 36 hours before being ground to a fine power (Spex Certiprep 8000D Mixer Mill). All isotope samples were treated with 0.01% HCl solution to remove carbonates.

Samples of potential organic matter sources were collected from each river and estuary for isotope analysis. To characterize riverine organic matter, replicate samples of upstream riparian leaves and instream leaf litter were collected at the lowest elevation water intake site in the Espiritu Santo and Mameyes rivers. Samples of estuarine basal resources (mangrove leaves, macroalgae, and estuarine biofilm) were collected at each sampling station in the Mameyes and Espiritu Santo estuaries. Fresh mangrove leaves were collected from all stations where mangroves were present (the middle and lower Espiritu Santo and a small area at the lower Mameyes). Macroalgae was collected from woody debris and buoys where present; however, it was not observed at all stations (e.g., station 1 of the Espiritu Santo ). All leaves and macroalgae samples were dried at 40°C for at least three days being ground to a fine power.

Collection of pure samples of phytoplankton and/or benthic mircroalgae is difficult and many studies have been forced to either rely on published values or to exclude phytoplankton and/or benthic microaglae as a potential organic matter source (e.g., Peterson and Howarth 1987; Benstead et al. 2006). We were unable to collected pure sample of phytoplankton or benthic microaglae from the Espiritu Santo and Mameyes estuaries. Instead, biofilm samples were collected at each station by deploying anchored flotation devices that suspended three 12” square tiles in the water column. After one week, the tiles were collected and immediately scraped for biofilm. Biofilm isotope samples were processed as described above for animal tissue.    

Stable isotope analysis

δ13C and δ15N analyses were run at the University of Georgia, Institute of Ecology, Analytical Chemistry Laboratory in a Carlo Erba NA 1500 CHN analyzer (Carlo Erba Instrumentazione, Milan, Italy) coupled to a Finnigan Delta C isotope ratio mass  spectrometer (Thermo Electron Corp., Waltham, USA) operating as a continuous flow system. Reproducibility was monitored using a bovine liver standard. Precision was better than ± 0.2‰ (1 SD). Animal tissue δ34S analyses were run at the Colorado Plateau Stable Isotope Laboratory on a Carlo Erba Model NC2100 elemental analyzer coupled to a Finnigan Delta Plus Advantage isotopic ratio mass spectrometer. Range in measurement error was ± 0.7-1.0‰. Plant and biofilm δ34S samples were analyzed at the Coastal Sciences Laboratory on a VG (Micromass) isotope ratio mass spectrometer. Isotope values are expressed as δ13C, δ15N, or δ34S (with units of ‰) according to the following equation: δ13C, δ15N, or δ34S = [(Rsample/Rstandard) - 1] × 1000 where R = 13C/ 12C, 15N/ 14N or 34S/32S. Reference standards were PeeDee Belemnite carbonate, atmospheric N2, and Canyon Diablo Troilite for δ13C, δ15N, and δ34S, respectively.

Gut content analysis

Guts contents of the four species targeted for isotope analysis (above) plus other common pelagic and omnivorous fishes (Caranx hippos, Caranx latus, Diapterus auratus, Lutjanus jocu, Micropogonias furnieri, Opisthonema oglinum, Pomadasys crocro, Scomberomorus regalis, Selene vomer, Sphyraena barracuda, Strongylura timucu) were analyzed to determine the contribution of freshwater shrimps to fish diet. Stomachs were extracted and the contents were squeezed directly into vials and preserved in 70% ethanol. Because guts of some fishes were very small, the percent contribution of each food item could not be calculated volumetrically. Instead, we placed a petri dish over a grid (2x2 mm) and, using a dissecting scope, calculated the portion of the petri dish covered by each distinct food item (as described in Ley et al. 1994).