In this post I want to talk about the latest paper I published with my colleagues. The story starts just outside İzmir.
Every summer from 2023 to 2025, between June and October, we set traps in Azmak Stream in Seferihisar as part of a university research (BAP) project that I led. By the end of the project, 197 blue crabs had been caught in those traps. We measured the carapace width of each one with callipers and weighed it on a precision balance. The smallest was 81 mm wide and weighed about 40 grams; the largest reached 195 mm and 358 grams.
The question we wanted to answer with these measurements looked simple: if we know a blue crab's carapace width, how accurately can we predict its weight?
Why does it matter?
In the Mediterranean the blue crab has two faces. On one side it is an invasive alien that disrupts ecosystems and causes economic losses, above all to fishers. On the other, it is a seafood resource that has been landed and processed along Türkiye's eastern Mediterranean coast for decades and is increasingly export-oriented across the Mediterranean. Whether you are trying to suppress the population or harvest it sustainably, you need the same number: how much crab is out there, or in the catch?
Weighing thousands of crabs one by one in the field is not practical. Measuring width is much easier. To convert width into weight, fisheries science has long used a simple formula: W = a·CW^b. The part I will focus on in this post is the exponent b.
If a crab grew at the same rate in every direction, b would be exactly 3; this is called isometric growth. A b below 3 means the animal gains weight more slowly than expected as it gets wider.
Males and females do not follow the same curve
Both sexes had b values below 3, so both become relatively "lighter" as they grow wider. But the gap between them is not small:
- In males, b is about 2.70
- In females, b is about 2.18
From a fisheries management perspective, I think this is the key message of the study: males and females follow markedly different scaling trajectories. A model that pools both sexes into a single equation can easily erase that difference. If we want reliable biomass estimates and management strategies that work for invasive stocks, we have to account for these sex-specific differences.
When we pooled all the crabs and ran the classical calculation, we got b = 2.22. In our female-dominated sample, that single number almost completely hid the very different growth curve of the males.
Why does female weight decouple from width?
When females reach maturity they moult one last time. After this "terminal" moult, carapace width stays essentially fixed. Weight, however, keeps changing with ovarian development and nutritional condition. Energy shifts from growth to reproduction.
In our sample, 73.5% of the females measured between 130 and 160 mm; most of them were adults that had already completed this final moult. Females of the same width varied considerably in weight depending on the stage of their ovaries. This scatter is not caused by external egg masses: we released egg-bearing females alive without measuring them.
Males, on the other hand, keep moulting and growing after maturity. They invest their energy in muscle and large claws, so their weight tracks their width more closely.
This is not unique to the invasive populations in the Mediterranean. In Chesapeake Bay, in the species' native range, studies as far back as 1949 reported similarly low exponents for females. That suggests to me that the difference we observed is not a by-product of the invasion but part of the blue crab's own biology.
"Outlier" crabs are not errors, they are biology
The usual way of fitting a width–weight relationship assumes that deviations in the data follow a normal distribution, the familiar bell curve. Moulting, irregular feeding and the breeding season add more extreme values to crab data than a bell curve can handle.
So we built four Bayesian models and let them compete. Some used a normal distribution, others a Student's t-distribution, which is more tolerant of extreme values; some accounted for sex differences and some did not. We measured how well each model predicted data it had not seen using cross-validation. The winner was the Student's t model that treats the sexes separately. Instead of suppressing crabs with extreme weights as noise, it treated them as genuine variation driven by reproduction.
What changes in practice?
This difference has concrete consequences for management.
When a single pooled equation is used, biomass estimates are systematically biased: the weight of large females is overestimated and that of large males is underestimated. Any harvest or removal programme based on quotas, size limits or biomass inherits this error.
Because female weight no longer rises with width after the terminal moult, minimum landing size and mesh regulations that ignore this may under-harvest reproductive females. Yet if the goal is to slow population growth, they are exactly the group to target. We therefore recommend using sex-specific conversion parameters both in management programmes and in commercial fisheries.
What we don't know yet
Our sample had only 31 males against 166 females. The Bayesian approach does not hide this imbalance; it shows it openly by widening the uncertainty intervals of the male estimates. Still, the values we report for males should be treated as provisional until they are confirmed with larger samples. Our measurements also come from a single estuary and only from the warm months. The next step is to extend sampling across all seasons and to a more balanced sex ratio.
The full paper is open access: Ceyhan T, Balkı C, Akyol O, Ulaş A, López-Mengual I (2026). Robust Bayesian inference of carapace width–weight scaling in the invasive blue crab (Callinectes sapidus). Journal of the Marine Biological Association of the United Kingdom, 106, e64. doi.org/10.1017/S0025315426101581