Population dynamics of Baltic herring since the Viking Age revealed by ancient DNA and genomics

Abstract

The world’s oceans are currently facing major stressors in the form of overexploitation and anthropogenic climate change. The Baltic Sea was home to the first “industrial” fishery ∼800 y ago targeting the Baltic herring, a species that is still economically and culturally important today. Yet, the early origins of marine industries and the long-term ecological consequences of historical and contemporary fisheries remain debated. Here, we study long-term population dynamics of Baltic herring to evaluate the past impacts of humans on the marine environment. We combine modern whole-genome data with ancient DNA (aDNA) to identify the earliest-known long-distance herring trade in the region, illustrating that extensive fish trade began during the Viking Age. We further resolve population structure within the Baltic and observe demographic independence for four local herring stocks over at least 200 generations. It has been suggested that overfishing at Øresund in the 16th century resulted in a demographic shift from autumn-spawning to spring-spawning herring dominance in the Baltic. We show that while the Øresund fishery had a negative impact on the western Baltic herring stock, the demographic shift to spring-spawning dominance did not occur until the 20th century. Instead, demographic reconstructions reveal population trajectories consistent with expected impacts of environmental change and historical reports on shifting fishing targets over time. This study illustrates the joint impact of climate change and human exploitation on marine species as well as the role historical ecology can play in conservation and management policies.

An empirical test of the estimation of historical effective population size using Drosophila melanogaster

“…The method of Santiago et al (2020) has now been applied to different species, particularly in the last year, including insects, such as honeybees (Sang et al, 2022); birds, such as Black Robin (von Seth et al, 2022); fishes, such as turbot, seabream and seabass (Saura et al, 2021), Baltic herring (Atmore et al, 2022), pikeperch (De Los Ríos-Pérez et al, 2022, coho salmon (Martinez et al, 2022), catfish (Coimbra et al, 2023) and sailfish (Ferrette et al, 2023); wild mammals, such as grey wolf (Pacheco et al, 2022), killer whales (Kardos et al, 2023), sika deer (Iijima et al, 2023), scimitar-horned oryx (Humble et al, 2023) and gorilla (Alvarez-Estape et al, 2023); humans (Bird et al, 2023); domestic species, such as pigs (Krupa et al, 2022), cattle (Jin et al, 2022;Magnier et al, 2022), sheep (Djokic et al, 2023;Drzaic et al, 2022), horse (Criscione et al, 2022) and chicken (Gao et al, 2023;Liu et al, 2023); plants, such as walnut (Ding et al, 2022); crustaceans, such as Daphnia (Wersebe & Weider, 2023) and fungi (Singh et al, 2021). As suggested by Santiago et al (2020), the method is generally reliable for about 200 generations in the past, although…”

“…As suggested by Santiago et al (2020), the method is generally reliable for about 200 generations in the past, although the software provides values up to about 600 generations. The analyses carried out to date have followed this recommendation considering between 20 and 200 generations, or up to 300–400 generations in a couple of studies (Atmore et al, 2022; von Seth et al, 2022) and up to 700–800 in others (Gao et al, 2023; Singh et al, 2021). It is important to note that the linkage disequilibrium signal is diminished over time so that very far events cannot generally be inferred with precision.…”

“…In fact, Reid and Pinsky (2022) showed that the method exhibits ≥90% accuracy for detecting fast severe declines in population size, outperforming other methods. The method has also been applied to different sets of data from different species (Atmore et al, 2022;Bird et al, 2023;Ding et al, 2022;Ferrette et al, 2023;Pacheco et al, 2022;von Seth et al, 2022;Wersebe & Weider, 2023). However, the simulation tests have always a restricted scope, as the models simulated cannot encompass all the complexities of real genomes.…”

Ancient DNA sequence quality is independent of fish bone weight

“…While the DNA concentration of each extract was weakly correlated with skeletal element and weight, neither of these were correlated with quality or complexity measures in the end, indicating that the increase in DNA concentration is likely from exogenous sources. Library complexity does not have to be particularly high to sequence a genome to low coverage, and even ultra-low-coverage genomes can yield meaningful biological results (Ferrari et al, 2022;Atmore et al, 2022). However, for some analyses, such as demographic inference, a greater amount of coverage and complexity is required (Schiffels and Wang 2020).…”

Travelling in the omics time machine: uncovering mysteries of the extinct Moroccan trout (Salmo pallaryi)

“…It has given rise to the field of ancient DNA, whereby DNA has been sequenced from a 2,800-year-old domesticated horse (Lippold et al, 2011), a 1,000,000-year-old mammoth (van der Valk et al, 2021) and even ancient environmental DNA from a 2-million-year-old ecosystem in Greenland (Kjaer et al, 2022). These studies have given insights into new lineages, sources of colonization events, adaptive genes and the composition of entire ecosystems, and although many of these studies have focused on large charismatic mammals, there have been a growing number of studies that have been conducted on fish, including 1,100-year-old Baltic herring (Atmore et al, 2022), 2,400-year-old swordfish (Yüncü et al, 2021) and 6,300year-old Atlantic cod (Martínez-Garciá et al, 2022).…”