Long-Term Seasonal and Interannual Patterns of Marine Mammal Strandings in Subtropical Western South Atlantic

Abstract

Understanding temporal patterns of marine mammal occurrence is useful for establishing conservation strategies. We used a 38 yr-long dataset spanning 1976 to 2013 to describe temporal patterns and trends in marine mammal strandings along a subtropical stretch of the east coast of South America. This region is influenced by a transitional zone between tropical and temperate waters and is considered an important fishing ground off Brazil. Generalized Additive Models were used to evaluate the temporal stranding patterns of the most frequently stranded species. Forty species were documented in 12,540 stranding events. Franciscana (n = 4,574), South American fur seal, (n = 3,419), South American sea lion (n = 2,049), bottlenose dolphins (n = 293) and subantarctic fur seal (n = 219) were the most frequently stranded marine mammals. The seasonality of strandings of franciscana and bottlenose dolphin coincided with periods of higher fishing effort and strandings of South American and subantarctic fur seals with post-reproductive dispersal. For South American sea lion the seasonality of strandings is associated with both fishing effort and post-reproductive dispersal. Some clear seasonal patterns were associated with occurrence of cold- (e.g. subantarctic fur seal) and warm-water (e.g. rough-toothed dolphin) species in winter and summer, respectively. Inter-annual increases in stranding rate were observed for franciscana and South American fur seal and these are likely related to increased fishing effort and population growth, respectively. For subantarctic fur seal the stranding rate showed a slight decline while for bottlenose dolphin it remained steady. No significant year to year variation in stranding rate was observed for South American sea lion. The slight decrease in frequency of temperate/polar marine mammals and the increased occurrence of subtropical/tropical species since the late 1990s might be associated with environmental changes linked to climate change. This long-term study indicates that temporal stranding patterns of marine mammals might be explained by either fishing-related or environmental factors.

Fig 1. Study area.

Area 1 = 87 km; Area 2 = 48 km; Area 3 = 63 km; Area 4 = 70; Area 5 = 87 km.

Fig 2. Marine mammal stranding rate by year (A) and month (B) from 1976 to 2013 (records of unidentified species are included).

Gray line in (B) represents stranding rate by month after removing records of species with high mortality, migratory whales, neonates of South American fur seal and unidentified marine mammals. Y-axes in (B) are on a square-root scale for ease-of-read purpose. Only stranding records from areas 2 and 3 were included.

Fig 3. Temporal trends in franciscana strandings.

Estimated smooth function (solid line) with 95% confidence interval (dashed lines) for the fitted GAM by year (A) and month (B) from 1992 to 2013. Y-axis = fitted function with estimated degrees of freedom in parentheses.

Fig 4. Stranding rate of frequent species by year from 1976 to 2013.

Only stranding records from areas 2 and 3 were included.

Fig 5. Mean stranding rate of frequent species by month from 1976 to 2013.

Black line represents the stranding rate of carcasses in which the cause of death was attributable to bycatch. Y-axes in the left graphics are on a square-root scale for ease-of-read purpose. Only stranding records from areas 2 and 3 were included.

Fig 6. Temporal trends in common bottlenose dolphin strandings.

Estimated smooth function (solid line) with 95% confidence interval (dashed lines) for the fitted GAM by year (A) and month (B) from 1992 to 2013. Y-axis = fitted function with estimated degrees of freedom in parentheses.

Fig 7. Temporal trends in South American sea lion strandings.

Estimated smooth function (solid line) with 95% confidence interval (dashed lines) for the fitted GAM by year (A) and month (B) from 1992 to 2013. Y-axis = fitted function with estimated degrees of freedom in parentheses.

Fig 8. Temporal trends in South American fur seal strandings.

Estimated smooth function (solid line) with 95% confidence interval (dashed lines) for the fitted GAM by year and month from 1992 to 2013. Neonates (A and B) and juveniles/adults (C and D). Y-axis = fitted function with estimated degrees of freedom in parentheses.

Fig 9. Temporal trends in subantarctic fur seal strandings.

Estimated smooth function (solid line) with 95% confidence interval (dashed lines) for the fitted GAM by year (A) and month (B) from 1992 to 2013. Y-axis = fitted function with estimated degrees of freedom in parentheses.

Fig 10. Stranding rates of false killer whale, sperm whale and rough-toothed (A) and baleen whales (B) by month from 1976 to 2013.

Stranding records from all areas (1–5) were included.

Fig 11. Stranding rates of tropical/subtropical and temperate/polar species by year (A) and month (B) from 1976 to 2013.

Tropical/subtropical species: pantropical spotted dolphin, Stenella attenuata, Atlantic spotted dolphin, Stenella frontalis, rough-toothed dolphin, Fraser’s dolphin, Lagenodelphis hosei and Bryde’s whale. Temperate/polar species: Burmeister’s porpoise, Phocoena spinipinnis, spectacled porpoise, Phocoena dioptrica, Peale’s dolphin, Lagenorhynchus australis, Commerson’s dolphin, Cephalorhynchus commersonii, southern elephant seal, Mironga leonina, crabeater seal, Lobodon carcinophaga, Antarctic fur seal, Arctocephalus gazella, Gray’s Beaked whale Mesoplodon grayi, strap-toothed whale, Mesoplodon Layardii, and Arnoux’s beaked whale, Berardius arnuxii. Stranding records from all areas (1–5) were included.