Does the River Continuum Concept apply on a tropical island? Longitudinal variation in a Puerto Rican stream

Short name: 


Data set ID: 



We examined whether a tropical stream in Puerto Rico matched predictions of the River Continuum Concept (RCC) for macroinvertebrate functional feeding groups (FFGs). Sampling sites for macroinvertebrates, basal resources, and fishes ranged from headwaters to within 2.5 km of the fourth-order estuary. In a comparison to a model temperate system where RCC predictions generally held, we used catchment area as a measure of stream size in order to examine truncated RCC predictions (i.e., cut off to correspond to the largest stream size sampled in Puerto Rico). Despite dominance of generalist freshwater shrimps, which use more than one feeding mode, RCC predictions held for scrapers, shredders, and predators. Collector-filterers showed a trend opposite that predicted by the RCC, but patterns in basal resources suggest that this is consistent with the central RCC theme: longitudinal distributions of FFGs follow longitudinal patterns in basal resources. Alternatively, the filterer pattern may be explained by fish predation affecting distributions of filter-feeding shrimp. Our results indicate that the RCC generally applies to running waters on tropical islands. However, additional theoretical and field studies across a broad array of stream types should examine whether the RCC needs to be refined to reflect the potential influence of top-down trophic controls on FFG distributions.

Date Range: 
2001-06-01 00:00:00 to 2001-08-01 00:00:00

Publication Date: 

2011-06-07 00:00:00



Additional Project roles: 

Name: Eda Melendez-Colom Role: Data Manager
Name: Catherine Pringle Role: Associated Researcher


At each site, we established a reach that was 10 times the channel width and visually identified major habitats. Proportional availability of major habitats was determined using occurrence frequency of mesohabitats (riffles, pools, and runs, based on measured lengths) combined with occurrence frequency of substrates (based on 54 random measurements in each major mesohabitat type). All measurements and samples were taken during base flow from June through August 2001. Epilithic chlorophyll a was sampled using a suction device modified from Loeb (1981). We collected a pooled sample of four individual and randomly located Loeb samples in each of three pools and three riffles at each site. Chlorophyll a samples were brought back to the lab on ice, filtered, frozen, and analyzed fluorometrically according to standard methods (American Public Health Association 1985).

To measure non-shrimp benthic invertebrate densities and standing stocks of very fine (VF, <250 m) and medium fine (MF, 250 m – 1 mm) benthic inorganic matter (BIM) and VF, MF, and coarse (C, >1 mm) benthic organic matter (BOM), three randomly located samples per major habitat were collected and processed using methods modified from Lugthart and Wallace (1992) and Grubaugh et al. (1996). Macrophytes in runs, as well as sand habitats in pools, were sampled using a core of known area. Pools with boulders and interstitial cobble and gravel were sampled using the "benthic block net" method described by Greathouse and Pringle (2005); the benthic block net apparatus functions similarly to a core. Bedrock–boulder cascades were sampled by scrub-brushing invertebrates and benthic matter from a known area into a hand net or into a Surber sampler with foam rubber tied around the frame to create a seal with the substrate. Cobble–coarse gravel riffles were sampled with an unmodified Surber net. In cobble runs, a sample consisted of a single cobble placed into a downstream D net, scrubbed on the bank to remove invertebrates and benthic matter, and measured to determine surface area. Because hand, Surber, and D nets were 250 m mesh, we further modified methods of Lugthart and Wallace (1992) and Grubaugh et al. (1996) by using the Loeb sampler to collect VFBOM and VFBIM in Surber and hand-net samples, and we used the Loeb sampler on nearby cobbles to estimate VFBOM and VFBIM in D-net samples. An additional modification is that we did not separate macrophytes from coarse detritus when measuring standing stocks of CBOM; however, macrophytes only occurred in run and riffle samples at the lowest-elevation site (Hydrilla) and in a single riffle sample at site 3 (filamentous algae). Invertebrate samples were preserved in the field in ethanol.

Samples intended for estimating VFBOM/BIM (from Loeb, core, and benthic block net sampling) were sieved to remove material >250 m before filtration onto pre-ashed, preweighed glass fiber filters (Whatman GF/F, 0.7 m). Nondecapod invertebrate samples were separated into MF and C fractions using nested 1 mm and 250 m sieves before sorting for macroinvertebrates. After sorting, we subsampled the MFBOM/BIM for filtration onto a pre-ashed, pre-weighed glass fiber filter. CBOM and filters of VFBOM/BIM and MFBOM/BIM were dried for 24 h at 50 °C, weighed, ashed for 3 h at 500 °C, and reweighed to determine ash-free dry mass (AFDM) for VFBOM, MFBOM, and CBOM and inorganic dry mass for VFBIM and MFBIM. Non-decapod invertebrates were identified to the lowest practicable level (2005 NRC Canada 136 Can. J. Fish. Aquat. Sci. Vol. 63, 2006).

We obtained quantitative estimates of shrimp and crab biomass and fish abundance by electroshocking and snorkeling. A minimum of one pool or run habitat and one riffle habitat per site were electroshocked using depletion or removal sampling methods (White et al. 1982; Fièvet et al. 1996, 1999). Block and dip nets were 0.635 cm mesh. Captured shrimps and crabs were identified to genus or species and measured to the nearest millimetre for postorbital carapace length (Hobbs 1991). Electroshocking was ineffective for capturing gobies (Sicydium and Awaous) and sleepers (Eleotridae) in pools, as has been observed by other researchers for Puerto Rico (E. García, US Department of Agriculture Forest Service, Southern Region, Atlanta, GA 30309, USA, and N. Hemphill, Trinity County Resource Conservation District, Weaverville, CA 96093, USA, personal communication). Thus, in one pool or run per site, we snorkeled a known area, counting all gobies and sleepers observed. Visibility was similar across sites. Carapace length– mass regression equations for shrimps and crabs, determined from individuals preserved in ~10% formaldehyde, were used to determine shrimp and crab biomass. In determining biomass of macroinvertebrate functional feeding groups, we apportioned biomass of shrimp taxa equivalently to each known functional feeding group membership.



Time Period: