Bridging archaeology and marine conservation in the Neotropics

Abstract

Anthropogenic impacts on tropical and subtropical coastal environments are increasing at an alarming rate, compromising ecosystem functions, structures and services. Understanding the scale of marine population decline and diversity loss requires a long-term perspective that incorporates information from a range of sources. The Southern Atlantic Ocean represents a major gap in our understanding of pre-industrial marine species composition. Here we begin to fill this gap by performing an extensive review of the published data on Middle and Late Holocene marine fish remains along the southern coast of Brazil. This region preserves archaeological sites that are unique archives of past socio-ecological systems and pre-European biological diversity. We assessed snapshots of species compositions and relative abundances spanning the last 9500 years, and modelled differences in species' functional traits between archaeological and modern fisheries. We found evidence for both generalist and specialist fishing practices in pre-European times, with large body size and body mass caught regularly over hundreds of years. Comparison with modern catches revealed a significant decline in these functional traits, possibly associated with overfishing and escalating human impacts in recent times.

Fig 1

(A) Archaeological sites with fish remains, and (B) archaeological sites with fish remains in Babitonga Bay. Maps generated using ArcGIS 10.7 ([42]), CGIAR Consortium for Spatial Information ([43]) and NASA/JPL-Caltech (adapted from https://www.jpl.nasa.gov/images/pia03388-south-america-shaded-relief-and-colored-height).

Fig 2

Boxplot showing the distribution of A) NISP/m³ versus mesh size and B) SR/NISP versus mesh size.

Fig 3

Correlation between SR and NISP/m³ for A) the entire dataset (sources A and B), B) faunal remains recovered with 2–3 mm mesh size and identified using both bones and otoliths, and C) the same data in B after removing a single outlier (site of Rio do Meio) with the highest SR and NISP/m³.

Fig 4

(A) Relative abundance of bony fish taxa above 1% in pre-European assemblages, and (B) relative frequency of bony fish taxa above 10%.

Fig 5

Relative abundance of Micropogonias furnieri as a function of (A) the median radiocarbon age of sites (modelled cal BP), and (B) latitude in pre-European assemblages. Green band represents the 95% confidence interval.

Fig 6

A) Relative abundance and B) relative frequency of cartilaginous fish taxa in pre-European assemblages.

Fig 7

Species Richness as a function of A) the latitude and B) the median radiocarbon dates of the sites; C) weighted average trophic level per latitude and D) chronology; E) relative abundance of Elasmobranchii remains in relation to latitude and F) chronology of the sites. The bands represent the 95% confidence interval.

Fig 8. Latitudinal changes in species richness.

Maps generated using ArcGIS 10.7 ([42]) and CGIAR Consortium for Spatial Information ([43]).

Fig 9

Fish traits exploited between 4500–1150 cal BP, 1050–600 cal BP and the modern period (1994–2015) for A) maximum body sizes, B) maximum body masses and C) trophic levels. Box plots represent observed values from fisheries catches across 4500–1150 cal BP, 1050–600 cal BP and the modern period (AD 1994–2015). Vertical black line denotes the median observed values, and black points are data outliers. Grey violin plots on the back represent the density distribution of null model values from random samples of the total species pool. P-values for pairwise comparisons of observed traits versus randomised data are shown below the boxes. Asterisks mark significant differences between time periods as indicated by Tukey HSD.

Fig 10. Fish traits and functional groups across time periods.

A) The proportion of fish species within different body size categories, B) trophic groups, and C) functional entities between pre-European (4500–1150 cal BP and 1050–600 cal BP) and modern (AD 1994–2015) periods. In C, functional entities represent the combination of fish trophic groups and body size classes, and darker tones represent larger body size categories within the same trophic group. Two-tailed Binomial tests reveal significant (p < 0.05) differences in the proportions of 15.1–30 cm species between the 4500–1150 cal BP and modern (AD 1994–2015) periods.