The Atlantic cod (Gadus morhua) is an iconic fish species with immense historical, cultural, and socio-economic value (Kurlansky, 1997; Kulatska et al., 2025). The catches of cod in both sides of the Atlantic Ocean, including the greater North Sea region, have historically been high (Kulatska et al., 2025). However, the North Sea cod has recently experienced a regime shift, with the southern population remaining in a depleted state and the northern population recovering (Cecapolli et al., 2025). In Norway, cod fisheries management is delimited by location. Cod found south of 62°N and within 12 nautical miles of the coast is considered southern coastal cod and managed as such, which includes the cod in the Oslofjord. Note that the International Council for the Exploration of the Sea (ICES), which provides scientific advice regarding stock health to the North Atlantic region, is not currently involved in the advice of cod in this region, and there is no formal population assessment in place (Stock et al., 2025). The Oslofjord region (Figure 1) is situated in the Skagerrak, adjacent to and closely associated with the North Sea proper.

The monitoring of the cod in the Oslofjord has a long history, in particular linked to a long beach seine data series dating back to 1919 (Stenseth et al., 1999; Knutsen et al., 2018). However, this highly valuable data source mainly covers the first year juveniles (0-group) cod, especially over the last two decades (Knutsen et al., 2022). In addition to the beach seine data, surveys targeting eggs and larval stages have been conducted (Ciannelli et al., 2025), and two stations in the outer fjord have been systematically covered in monitoring of adult fish in winter, mainly since the early 2000s with some data from the 1980s (Roney et al., 2016; Stock et al., 2025). Further, relatively small-scale sampling of cod in the inner fjord has been performed over the last decade, largely in response to the recent poor status of the cod (Haugen et al., 2023; Hylland, 2025).

An important data source for recruited fish is the commercial landings statistics of the Directorate of Fisheries in Norway (www.fiskeridir.no) . Commercial landings have been reported for more than 150 years from e.g., the inner fjord. These records indicate landings between about 10–100 tons per year (Ruud, 1968) for the period 1872-1964, but with some years with very low landings of less than 5 tons and one year with more than 120 tons. Over the last two decades, the commercial cod landings in the Oslofjord have experienced a strong decline (Moland and Strandli, 2021). This decline co-occurs with a generally reduced number of fishing vessels operating in the region, making the link between landings and the biomass of cod less direct. In addition, the reduction in landings is only partially consistent with juvenile (0-group), egg, and larval data (Knutsen et al., 2022), which indicate relatively high abundances in some years. It is well-documented from other regions that mortality in cod is high and variable, also after the 0-group stage (Bogstad et al., 2016). It is therefore important to obtain a better quantification of the changes in cod over time in the Oslofjord, based on recruited fish. Such quantification is topical due to large public interest in the recent negative trends for the Oslofjord cod, and a recent political decision (Regjeringen innfører nullfiskeområder i Oslofjorden, 2025) to introduce full no-take zones in the region from 1^st January 2026 (Figure 1).

In general, the Oslofjord has experienced a range of environmental changes over the past century, including effects from eutrophication, reduced oxygen concentrations, coastal darkening, and increased sea temperatures (Rogers et al., 2011; Moland and Strandli, 2021; Ramon et al., 2025), with anticipated, though with uncertain effects on the cod population. Climate warming is impacting cod recruitment and population dynamics negatively in the North Sea region, as it is close to the southern limit of the cod distribution (Drinkwater, 2005; Sguotti et al., 2022; Ma et al., 2024). Documentation of the quantitative changes that have occurred for cod is a first step in attempting to improve the status of the cod population and of the Oslofjord ecosystem as a whole. Here, we demonstrate how the use of data from the Directorate of Fisheries, based on reported landing tickets for the period 2000-2024, can be used to quantify the catch and construct a proxy for effort to better link the catch data to biomass with estimated uncertainty.

The calculated annual catch in the Oslofjord is shown in Figure 2. The catch increased from around 150,000 kg in 2000 to a peak of around 250,000 kg in 2004 and then dropped steadily to about 84,000 kg in 2010, being stable between 80,000-120,000 kg until 2015, with a further decline to low levels< 5,000 kg in 2024. For the inner fjord, the catches were always below 5,000 kg over the whole period, which is in the lower end of what has been historically reported (Ruud, 1968). The trends are somewhat different for the CPUE for the whole fjord, with more stable CPUE largely between 20–30 kg cod per landing event for the years before 2015 (mean 24 kg), and a sharp decline to around 6 kg per landing event in 2018 and down to around 3 kg per landing after 2020. Analysis of the CPUE from different gear types indicated that the catchability of shrimp trawls was reduced after 2015 (Figure 3). The estimated CPUE after 2019 is based on limited data and should be interpreted with care.

Given the serious concerns about the status of the Atlantic cod in the Oslofjord, quantitative estimates of cod biomass are acutely needed. Here, we have demonstrated the value of a large commercial landings dataset for estimating how catch per unit effort (CPUE), a proxy for cod biomass (see SI for details), has changed over the past two decades. With these data, we have documented a dramatic decrease in cod landings, with part of the reduction being caused by reduced effort (Figure 2). Importantly, we have quantified a cod biomass proxy in the Oslofjord, with estimates of uncertainty. Our analysis documents a significant decline in the biomass proxy for Atlantic cod. For example, the biomass proxy dropped from a peak in 2004 to below 20% of these peak values in 2018. Biomass further declined after 2019, but due to changes in the management regime this year, the biomass proxy for the later years is presumably less reliable. Nevertheless, a striking feature of the time series of cod biomass is a period of relatively constant biomass followed by a steep decline after 2015. Note that the drop in CPUE is also observed when analyzing subsets of the data with different gear used (Figure 3). A scientific survey comprising one transect conducted four times a year since 2011 shows a comparable trend in abundance indices, with a clear decline after 2015 (Hylland, 2025).

However, what is causing this decline in cod biomass is currently not clear. It would be highly valuable to build models that can account for different mechanisms and drivers behind strong declines in population biomass (cf., Olsen et al., 2024). Such potential drivers include commercial and recreational fishing mortality and climate warming (Rogers et al., 2011; Sguotti et al., 2019), contaminants (Ono et al., 2019), and ecosystem interactions such as predation and food availability (Johannessen et al., 2012). Another potentially important mechanism is related to hypoxic conditions and reduced growth, as documented for Baltic cod (Neuenfeldt et al., 2020; Brander, 2022). Successful models for addressing such questions would require a better understanding of the population structuring, dynamics, and interactions of the sympatric populations in the region (Knutsen et al., 2018).

Importantly, cod populations in nearby areas are also in depleted states. This includes the Southern North Sea cod, which has experienced a collapse since 2015 (ICES, 2024; Cecapolli et al., 2025), mirroring the collapse observed in the Oslofjord region (Figure 2). The cod in both the eastern and western Skagerrak coast has seen a decline over time (Stock et al., 2025), which is observed in the Kattegat cod as well (Kulatska et al., 2025). One exception to this negative trend is the cod in the Öresund, which has shown an upward trajectory in recent years (Timmermann et al., 2023).

The connectivity to the other, adjacent populations of cod should be investigated further. In fjords along the Skagerrak, including the Oslofjord, there are at least two genetically distinct populations of Atlantic cod: a North Sea ecotype typically distributed more offshore but also inside fjords and a fjord ecotype more tightly linked to the fjords (Knutsen et al., 2018; Barth et al., 2019). With respect to connectivity, a small but significant fraction of the offspring from the Southern North Sea and the Viking populations drift into the Skagerrak and potentially into the Oslofjord (Romagnoni et al., 2020). Adult fish from the Kattegat region are known to migrate into the Skagerrak (Righton et al., 2010) and potentially into the Oslofjord (cf., Rogers et al., 2014). Given the strong link to adjacent regions, the status of cod biomass in the Oslofjord should be considered in light of the status of cod in these regions. However, we generally advise against harvesting on collapsed populations.

Despite some limitations in the current study (see SI), we argue that the size of the data set (more than 188,000 unique entries) reduces its impact, and we argue that the overall trends are likely captured by our proxy. We here present an urgently needed quantification of the status of Atlantic cod in the Oslofjord. This information is highly valuable for scientists, decision makers, and the public alike. It may serve as a baseline for the evaluation of future development and potential recovery of cod in the area. For fish stocks with formal assessment, it may be possible to detect both collapses and recovery, and also changes in the drivers of the population dynamics before and after the collapse (Durant et al., 2024). This indicates that even if the biomass recovers after a collapse, the relative role of the underlying drivers may change. With the current data collection in the Oslofjord, such an analysis would be very difficult to perform.

With the current fishing ban in the Oslofjord, alternative data are needed to document and understand a potential future recovery of Atlantic cod. Such data could include regular acoustic surveys with suitable spatiotemporal coverage (MacLennan and Simmonds, 2005) combined with trawl surveys. Another approach that could potentially be valuable both to monitor the future status of cod in the Oslofjord is environmental DNA (eDNA) sampling (Carvalho, 2024), potentially in a citizen science framework (Kvalheim et al., 2024). eDNA methods have recently been shown to reproduce trawl catches in a fjord system, albeit with large uncertainty (Guri et al., 2024). Most likely, a combination of current and emerging methods is needed to document future trajectories of cod biomass, not least to understand why the cod and the fishery have collapsed so drastically in the Oslofjord. A focused approach is equally relevant for other coastal fish populations that are subject to a range of environmental changes. Learning points from the collapse of the cod in the Oslofjord should apply to other stocks in Norway and beyond.