Stable isotopes in bivalves as indicators of nutrient source in coastal waters in the Bocas del Toro Archipelago, Panama

Abstract

To examine N-isotope ratios ((15)N/(14)N) in tissues and shell organic matrix of bivalves as a proxy for natural and anthropogenic nutrient fluxes in coastal environments, Pinctada imbricata, Isognomon alatus, and Brachidontes exustusbivalves were live-collected and analyzed from eight sites in Bocas del Toro, Panama. Sites represent a variety of coastal environments, including more urbanized, uninhabited, riverine, and oceanic sites. Growth under differing environmental conditions is confirmed by δ (18)O values, with open ocean Escudo de Veraguas shells yielding the highest average δ (18)O (-1.0‰) value and freshwater endmember Rio Guarumo the lowest (-1.7‰). At all sites there is no single dominant source of organic matter contributing to bivalve δ (15)N and δ (13)C values. Bivalve δ (15)N and δ (13)C values likely represent a mixture of mangrove and seagrass N and C, although terrestrial sources cannot be ruled out. Despite hydrographic differences between end-members, we see minimal δ (15)N and δ (13)C difference between bivalves from the river-influenced Rio Guarumo site and those from the oceanic Escudo de Veraguas site, with no evidence for N from open-ocean phytoplankton in the latter. Populated sites yield relative (15)N enrichments suggestive of anthropogenic nutrient input, but low δ (15)N values overall make this interpretation equivocal. Lastly, δ (15)N values of tissue and shell organic matrix correlate significantly for pterioideans P. imbricata and I. alatus. Thus for these species, N isotope studies of historical and fossil shells should provide records of ecology of past environments.

Keywords: Anthropocene; Carbon isotopes; Caribbean; Eutrophication; Molluscs; Nitrogen isotopes.

Figure 1. Map of the Bocas del Toro Archipelago, Panama.

Almirante Bay and Chiriquí Lagoon watersheds are represented by green shading. Reef extent is represented by orange shading and mangrove extent is represented by light green shading. Sample sites are labeled with black circles.

Figure 2. Boxplots of muscle, mantle, stomach, and gill tissue δ¹⁵N and δ¹³C values including all locations for P. imbricata, I. alatus, and B. exustus.

Figure 3. Location averages for muscle and mantle tissue δ¹⁵N versus δ¹³C for each respective location.

Error bars shown are ±2 SE of replicate samples for each site.

Figure 4. Boxplots displaying variability between different species at study sites where two or more species were present.

Figure 5. Location averages for muscle and mantle tissue δ¹⁵N versus δ¹³C for each respective location.

Error bars shown are ±2 SE of replicate samples for each site.

Figure 6. δ¹⁵N comparison for muscle and mantle tissues and corresponding species average shell value for (A) P. imbricata, I. alatus, and (B) combined (compared to average shell values for both species).

Brachidontes exustus were not available in enough abundance for statistical analysis. Each point represents the average value of specimens at one location.

Figure 7. Average shell δ¹³C versus δ¹⁸O values shell carbonate values for each location.

Average shell δ¹³C versus δ¹⁸O values shell carbonate values for each location including (A) Pinctada imbricata, (B) Isognomon alatus, and (C) Brachidontes exustus (n = 73). Location symbols are the same for remaining figures unless stated otherwise. Ti, tissue; Sh, shell. Error bars shown are ±2 SE of replicate samples for each site.

Figure 8. Location averages for δ¹⁵N versus δ¹³C including all species, compared with values for potential N and C sources.

Error bars shown are ±2 SE. Symbols are defined in Fig. 2. Banana leaf, R. mangle, terrestrial grass and T. testudinum δ¹⁵N and δ¹³C values are from Almirante Bay, Panama (Hilbun, 2009). The gray-shaded rounded rectangle indicates potential δ¹⁵N values of sewage (up to 20‰; Lapointe, Barile & Matzie, 2004; McClelland, Valiela & Michener, 1997). The green-shaded rounded rectangle estimates the forest leaf range in δ¹⁵N values in legumes and non-legumes from Barro Colorado Island, Panama in 2007 (1.5 to 3.0‰; Hietz et al., 2011). The yellow rounded rectangle represents δ¹⁵N values of open ocean POM (4–7‰; Cifuentes et al., 1996; Montes et al., 2013). The blue-shaded rectangle represents coastal POM values from the Bocas del Toro Archipelago (−21.2 to −19.6‰; Seemann et al., 2014). The vertical solid green line denotes grassland δ¹³C values and the vertical dashed brown line denotes forest floor litter δ¹³C values from Barro Colorado Island, Panama (Schwendenmann & Pendall, 2006). The represents the δ¹⁵N value of terrestrial organic matter entering a coastal ecosystem in Brazil (Corbisier et al., 2006). The horizontal purple dashed line represents δ¹⁵N values of T. testudinum shoots from Almirante Bay and horizontal red solid line represents δ¹⁵N values of T. testudinum in Chiriqui Lagoon (Carruthers et al., 2005).