Aarhus University Seal

Cormorant vs. salmon

Cormorant vs. salmon in the River Skjern: Extent of predation and testing of mitigation methods

Cormorants feed on fish in both marine and freshwater aquatic systems, often creating conflicts between fisheries and conservation interests. In order to find sustainable solutions, it is important to have a sound knowledge base for making decisions about possible management actions. Do cormorants threaten fish stocks? How do various actions e.g. scaring, impact or limit cormorant colonies?


To answer the following questions:

Predation of juvenile salmons

  1. Has predation of post-production migrating salmon juveniles been significantly reduced after several years of adjustment in the colonies and the recently launched scaring of cormorants?
  2. At what time during spring are salmon juveniles most vulnerable? Is the survival higher among salmon juveniles who migrate early than among salmon juveniles who migrate late? Knowing this may be important for when it is particularly important to ensure effective scaring.
  3. How does predation in the river (autumn / winter) impact production of smolt?

Cormorant population, preference for food and foraging place

  1. How is cormorant predation on salmon juveniles distributed on location - in the creek, respectively, in the fjord?
  2. How big are the populations of breeding and non-breeding cormorants in the fjord through the spring, and where are they located?
  3. Which particular fish species comprises the diet of cormorants during April-May?

Scaring of cormorants

  1. Which monitoring system needs to be developed to optimize likelihood of cormorants to be discovered when they attempt to predate salmon juveniles?
  2. What is cormorants’ short- and long-term reaction to repeated scaring?
  3. Where and how should scaring be planned to optimize impact?
  4. What are the impacts of scaring om day and night roost located along the rivers?
  5. Can the cormorants feeding in the estuary of the River Skjern and at the Hvide Sande Sluice effectively be scared away?

Project description

The project is carried out during the years 2016-2018. The project includes three field seasons. In the first season (2016/17) implemented regulation of cormorant colonies in the form of oiling of eggs, and performed scaring during the spring, fall and winter.

Counting birds

The number of nests in the colonies and the number of cormorants in the fjord are counted to give an estimate of the average number of birds present in smolt-period (April-May). Additional counts are carried out in August-September.


In 2017, an intensive scaring program will be launched during the smolt period in the mouth of the River Skjern and at the sluice in Hvide Sande. In the fall, winter and spring scaring will be conducted in the Skjern River and its tributaries. The scaring in the river system will include scaring at the day- and night-roost located along the Skjern River and its tributaries. These activities will be combined with monitoring of cormorants’ behavior and reactions to human traffic and scaring.

Salmon juveniles

Smolt traps will be established at the beginning of April in 2016 and 2017. The traps are inspected/cleaned /emptied daily. During the smolt migration 75 smolt will be taken out for radio-tagging. Tagged salmon juveniles are released directly into the river after awakening from anesthesia. Data logger stations will be established next to cormorant colonies and at the mouth of Skjern Å to distinguish between fish eaten in the river and fish eaten in the fjord. This investigation lasts very shortly (a few weeks) and provide very precise data on cormorant predation. Besides radio labelling smolt will be equipped with a PIT-tag. This can provide a basis for comparison of predation on radio-tagged and PIT-tagged fish. A program to monitor radio transmitter sound and to scan for PIT-tags signals will be established in the appropriate colonies and roosting areas.

Salmon fry

To determine cormorant predation on salmon fry before they become smolts and migrate the following two studies are conducted:

In the first study two river stretches are selected (each of a length of 100 m) for non-scaring, and two similar stretches are managed with intensive scaring. The density of salmon fry is monitored by electric fishing during August and again in March.

In the second study four river stretches are selected. On all four locations, the density of fish including salmon fry is registered. In autumn, a visible nylon-net is placed across two of the stretches, and the two other are left unprotected and with no attempts of scaring. In the spring, all four stretches are re-monitored with respect to fish density and the cover are removed. This experiment is repeated by altering the net cover between locations. Wildlife monitoring cameras will be set up on all four river stretches to continuously monitor the presence of potential predators.