The fall spawning ecotype of the herring, Clupea harengus Linnaeus, 1758, is considered functionally extirpated in the Baltic Proper. Once the dominant ecotype in fisheries landings of herring, few observations of this sub-population have been presented in the 21^st century. This short communication describes the opportunistic observation of fall spawning herring larvae in the costal Baltic Proper in 2023 (n = 61) and 2024 (n = 23). Back-calculated hatch dates indicated that the larvae likely hatched between the months of September and December prior to capture. Warmer water temperatures coincided with hatching extending later in the year. This study highlights the presence of a herring ecotype recently considered largely absent from the Baltic Proper and supports the need for further studies to investigate their current role in the ecosystem and herring fishery.

The effects of environmental changes on contaminant fate in the ecosystem are poorly understood, even in the otherwise well-studied Baltic Sea. This area is considered one of the most polluted in the world and is currently undergoing rapid shifts related to climate change and eutrophication. In this study, we focus on the effects of an altered productivity base and changes in food web structure on contaminant concentrations in the commercially important Baltic herring, which is also a key-species in the ecosystem. In herring of known size and age, collected within the Swedish National Monitoring Program for Contaminants in Marine Biota during the past two to three decades, retrospective analyses of contaminant concentrations and stable isotopes of carbon and nitrogen including amino acid-specific isotope analyses were performed. Partial least squares regression (PLSR) models were applied to dioxins, PCBs, and mercury time series to examine how biological, ecological, and environmental factors (i.e., age, trophic diversity and position, temperature, salinity, proxies of cyanobacterial blooms and ultimate nutrient sources, abundance of relevant benthic fauna as well as biomass and size structure of the zooplankton community) contribute in explaining contaminant concentrations in herring, beyond atmospheric deposition (the main contaminant input in the Central Baltic basin). Our results emphasize that the contaminant burden in Baltic herring is significantly influenced by factors other than atmospheric deposition. Primarily, changes in herring’s trophic ecology, together with nitrogen-fixing cyanobacterial blooms (supporting both growth biodilution and bloom-induced dilution), were linked to dioxin, PCB, and mercury concentrations in fish. Our results support the need to consider all potential ecological synergies and linkages when managing a rapidly changing system such as the Baltic Sea, in order to minimize noxious blooms without compromising the positive impact on contaminant concentrations in fish.

The Baltic herring (Clupea harengus membras) is an ecologically and economically important species of forage fish that plays a central role in the Baltic Sea food web. While elevated concentrations of contaminants within biota have been well documented in this historically polluted sea area, the effects that ecosystem changes such as eutrophication, climate-driven shifts in food web structure, and changes in growth have on contaminant burden remain poorly understood. This thesis investigates the complex effects of changing trophic interactions (Studies I, II, & III) and growth (Study III) on contaminant burden in herring, while providing updated tools to investigate similar questions in the future (Study IV). Study I showed that concentrations of persistent organic pollutants (PCBs and PCDD/Fs) and heavy metals (Hg) in herring are best explained by ecological and trophic changes, rather than decreased emissions. Higher cyanobacterial bloom intensity was associated with decreases in contaminant concentrations. Study II further investigated the relationship between herring and cyanobacteria by using compound specific isotope analysis in amino acids to quantify the amount of diazotrophic-N fixed by cyanobacteria present in juvenile herring over a growth season. Results showed that at peak levels, more than 30% of N in juvenile herring was fixed by cyanobacteria, with levels correlated to cyanobacterial biovolume in the water. Study III analyzed concentrations of PCDD/Fs and PCBs in the same juvenile herring used in study II and investigated the variables driving differences in concentrations. It was observed that contaminant concentration significantly decreased with fish size, with the smallest juveniles having higher contaminant concentrations than adults. General additive models showed that growth rate, assessed from daily growth rings in otoliths, was the variable most responsible for changes in contaminant concentration, with a higher growth rate leading to lower contaminant concentrations. These three studies show that cyanobacteria blooms in the Baltic have possible positive effects including providing N to the production of fish resources thereby potentially improving growth conditions and reducing contaminants through somatic growth dilution. Since non-lethal estimates of in-situ growth rates are difficult, we further developed a bioenergetic model for juvenile herring using ambient zooplankton densities and a functional consumption response, and incorporated water temperature and light period to predict growth (Study IV). The model was shown to better predict natural growth than previous models. The work in this thesis shows that ecosystem changes thought to be detrimental, such as increased cyanobacterial blooms, can have interactions with contaminant burden that are beneficial (Studies I & II), while showing that if the growth of juvenile herring can be improved, contaminant burden can be decreased (Study III). Lastly, a new mathematical method to calculate the important metric of growth rate in juvenile herring is presented, which will allow for future predictions of herring growth in a warming Baltic Sea (Study IV). In conclusion, all four studies support the need to contemplate potential ecological synergies and linkages when managing a rapidly changing system, in order to minimize the potential harmful effects of changes such as eutrophication, without compromising the positive impact on contaminant concentrations in fish.

Nitrogen-fixing cyanobacterial blooms worldwide have been shown to support production in primary consumers such as zooplankton and benthic invertebrates, but there have been few estimates on how much of the cyanobacterial nitrogen ends up in secondary consumers, such as fish. Using compound specific analysis of δ¹⁵N in amino acids (CSIA-AA), we have investigated if cyanobacterially fixed (diazotrophic) nitrogen (N) can be traced in higher trophic level organisms such as zooplankton and mysids, and observed in juvenile Atlantic herring (Clupea harengus). At the eutrophied station, the biovolume of N fixing cyanobacteria was higher than at the reference station and the autotrophic population was comprised of a higher proportion of cyanobacteria palatable to zooplankton. The δ¹⁵N ‰ of source amino acids in juvenile herring significantly decreased over the season at the eutrophied station. Bayesian mixing models were used to calculate the percent of diazotrophically-fixed N found in herring muscle tissue in relation to pre-bloom N, and results showed that herring at the eutrophied station had a higher percent (~34 %) of diazotrophically-fixed N in their muscle tissue than fish at the reference station (~18%). A significant relationship was found between the percentage diazotrophically-fixed N in herring muscle tissue and the biovolume of N fixing cyanobacteria in the environment two months prior to the capture date of the fish. This study adds to the mounting evidence that cyanobacteria blooms, a global occurrence, support production at higher trophic levels, including fish.

Cyanobacterial blooms are intensifying worldwide due to eutrophication and climate change, increasing cyanotoxin exposure to aquatic organisms. This study investigated the physiological, biochemical, and behavioural impacts of cyanobacterial blooms on the three-spined stickleback (Gasterosteus aculeatus), a widespread mesopredatory fish. Adult sticklebacks were exposed for two weeks to naturally collected bloom material dominated by toxic Nodularia spumigena, non-toxic Aphanizomenon sp., or a 50:50 mix. We measured toxin accumulation (NOD_eq), hepatic enzymatic activities (ethoxyresorufin-O-deethylase [EROD], glutathione S-transferases [GSTs], glutathione reductase [GR], and catalase [CAT]), and escape swimming performance (centre-of-mass velocity, angular velocity, distance, and duration) in a multiparametric endpoints approach. Sub-lethal toxin levels in muscle tissue ranged from 0.006 to 0.077 µg g⁻¹ d.w. Results showed that fish exposed to toxic-dominated treatments showed significantly elevated EROD activity (up to 200 % increase), moderate increases in GR and GSTs, and reduced CAT activity compared to controls. Notably, distance travelled during escape responses was reduced by ∼50 % in the high-toxicity treatment and showed an inverse correlation with EROD activity, suggesting a trade-off between detoxification effort and swimming performance. Overall, our results demonstrate that EROD is a sensitive biomarker for cyanotoxin exposure in fish under natural bloom conditions. This finding highlights the need to consider natural cyanotoxin effects when interpreting environmental assessments, particularly given the projected increase in bloom frequency and severity under future climate scenarios.

Contaminant concentrations in fish are not only determined by concentrations in the environment, but also affected by interactions between food web structure and physiological factors. Understanding these interactions and identifying which variables best explain contaminant concentrations is needed to predict how contaminant burden may affected by a changing climate. In the Baltic Sea, monitoring of persistent organic pollutants (POPs) in herring (Clupea harengus membras), an ecologically and commercially important species, has been historically performed on adult fish. Juvenile fish, on the other hand, undergo rapid changes in physiology and diet as they mature into adults, and therefor may offer a better insight into interactions between trophic structure, physiology, and contaminants. In this study, juvenile herring (<1 yr old) from two stations in the Baltic Sea were seasonally sampled over a summer and analyzed for dioxins and dioxin-like compounds (polychlorinated dibenzo-p-dioxins and dibenzofurans - PCDD/Fs, and dioxin like polychlorinated biphenyls - dl-PCBs, respectively). It was found that a significant decrease in concentration was observed with fish size, with smallest juveniles having higher levels than 4-12 times higher levels than the largest juveniles, and higher levels than adults. Generalized additive models showed that juvenile growth rate, obtained through otolith analysis, alone best explained POP concentrations while trophic position, calculated from nitrogen isotopes in amino acids, did not improve the models. This study is the first to present PCDD/F and dl-PCB concentrations in juvenile Baltic herring and to investigate physiological vs food web structure effects on the contaminant concentrations.