Counting the cetaceans of an entire sea, at the same time, with the same method: that is what the Mediterranean is going to attempt in the summer of 2026. The operation is called the ACCOBAMS Survey Initiative, it has reached its second edition, and it is one of the few exercises of this kind in the world.
Why a simultaneous census changes everything
Most of what is known about cetaceans comes from local monitoring: one team, one area, one season, over years. This work is irreplaceable for understanding behaviour or following individuals, as our monitoring of the Bay of Biscay shows. But it lends itself poorly to addition.
Two neighbouring teams do not necessarily use the same protocol, do not cover the same area, and do not correct in the same way for the animals they did not see. Adding their results produces a figure whose value nobody knows. And if the counts are spread out in time, an animal may be counted twice, or not at all.
A synoptic census answers this difficulty: same protocol, same summer, continuous coverage of the basin. The result is not only more precise, it is of another nature, because it allows comparisons between sub-regions and, from one edition to the next, a genuine measure of trend.
What the first edition established
The first ACCOBAMS Survey Initiative took place between 2018 and 2023, covering the Mediterranean and the Black Sea. It combined three approaches: aerial surveys, sightings from vessels, and passive acoustic listening, indispensable for deep-diving species such as the sperm whale, which is heard far more often than it is seen.
It produced the first abundance estimates at the scale of the basin, and a few structuring findings. The fin whale is the only baleen whale regularly present in the Mediterranean, with a population of the order of twenty thousand individuals. Four species account for more than nine encounters out of ten: common bottlenose dolphin and striped dolphin among the small cetaceans, fin whale and sperm whale among the large ones.
How abundance is estimated from an aircraft
The method is called distance sampling, and it deserves to be described, because it explains both the power and the limits of these campaigns.
The aircraft follows predefined flight lines. The observers record each group detected and its perpendicular distance from the line. It is known by construction that animals far from the track are seen less often than those directly below it: the distribution of the observed distances therefore makes it possible to reconstruct a detection function, that is, the probability of seeing an animal as a function of its distance.
From this function the area effectively surveyed is derived, hence a density, hence a population for the whole area. Two corrections are then added. The first arises because some animals are diving as the aircraft passes and cannot be seen, however attentive the observers: this is availability bias, particularly heavy for species that dive for long periods. The second arises from animals present at the surface but missed, which are estimated by having two independent teams work on board and comparing their detections.
It is the rigour of these corrections that distinguishes an abundance estimate from a simple count, and it is also what makes the application of an identical protocol from one campaign to the next indispensable.
The particular role of acoustics
The sperm whale illustrates why listening complements visual observation. It spends most of its time in deep dives, and it emits powerful clicks detectable several kilometres away. A towed hydrophone therefore hears a great deal more than an observer sees.
Converting these detections into a population remains delicate, however: it is necessary to know how many animals produce the sounds heard, at what distance, and how often they fall silent. Acoustics is excellent for establishing presence and distribution, and harder to convert into numbers. Our page on passive acoustic monitoring sets out these networks.
What the second edition adds
Launched in June 2025 and planned to run until 2028, ASI-II broadens the object. The synoptic census of the summer of 2026 will also cover sea turtles, seabirds and pelagic fish, and will include an assessment of human pressures, among them floating litter.
| First edition | 2018 to 2023, Mediterranean and Black Sea, first abundance estimates at the scale of the basin |
|---|---|
| Second edition | launched in June 2025, running until 2028 |
| Main synoptic campaign | summer 2026, completed in 2027 |
| Methods | aerial surveys, observation from vessels, passive acoustics |
| Broadening | turtles, seabirds, pelagic fish, litter, noise, ship strikes |
What is at stake in this second campaign is particular: it is the one that will turn the first into a starting point. An isolated estimate describes a state; two comparable estimates describe a trajectory, and it is the trajectory that governs management decisions.
What these figures are used for in practice
Three main uses, all of them regulatory.
First, setting thresholds. Determining whether a level of bycatch or of ship strikes is sustainable presupposes knowing the size of the population concerned. Without a denominator, no threshold has any meaning.
Next, delimiting areas. Protected areas, traffic restrictions and exclusion zones are drawn from density maps, as our reading of the French distribution maps shows.
Finally, evaluating the measures already taken. Without a repeated census, nobody can say whether a marine protected area produces an effect. Our pages on marine protected areas and sanctuaries set out these arrangements.
The limits of the exercise
A synoptic census remains a snapshot. It gives the state of one summer, not the variability between seasons, and cetaceans redistribute themselves over the course of the year. The estimates also carry confidence intervals that are sometimes wide, particularly for rare or inconspicuous species: a population estimated at a few hundred individuals may be estimated with a margin of the same order.
Finally, the comparison between two editions is valid only if the protocols have remained identical. Any methodological improvement, however desirable it may be, complicates the reading of the differences. This is the usual dilemma of long series, which we also encounter on our data pages.
Sources: ACCOBAMS secretariat, presentation of the ACCOBAMS Survey Initiative II and reports of the scientific committee; published results of the first edition, 2018-2023.
Did you find this article useful?
Add L'info des Cétacés to your preferred sources on Google. Our stories will be suggested to you more often, and the choice can be undone as easily as it is done.
Follow us on GoogleThis preference is saved in your Google account and can be changed at any time. It is neither a certification nor a recommendation from Google.
Leave a review