Modalités spatio-temporelles de la dispersion d'alevins de saumon atlantique (Salmo salar L.) à l'émergence Temporal and spatial aspects of the dispersai patterns of Atlantic salmon (Salmo salar L.) fry at emergence

Résumé

Nous avons étudié la dispersion (vitesse, durée, distance) d'alevins de saumon atlantique à l'émergence dans un ruisseau expérimental.

Le rythme journalier de dévalaison suit étroitement le rythme d'émergence des alevins, ce qui montre une bonne corrélation entre les deux activités, tout au moins en début de période. Les histogrammes de capture des alevins échantillonnés tous les 10 m présentent des caractéristiques de forme similaires (durée, asymétrie et aplatissement). Jusqu'au pic des captures, près de 50 % des alevins dévalent en 5 jours. Ils se dispersent à partir de la frayère sur 50 m en 3 ou 4 nuits et forment une cohorte homogène d'après leur rythme de dévalaison et leur taille. Après le mode, les captures sont plus étalées dans le temps (environ 10 jours) et montrent plus de variabilité en fonction de la distance. Les alevins résidents ne sont pas distribués uniformément dans les cinq biefs : les densités, de même que les poids moyens, tendent à augmenter vers l'aval. Plus de 50 % de la population d'alevins survivants s'est établi dans les 50 m en aval de la frayère. Les résultats révèlent deux vagues de dévalants. La première vague de dispersion, aussitôt après l'émergence, est rapide et importante. Elle ne dépendrait pas directement de la compétition territoriale et de la densité, mais permettrait d'éviter des densités localement trop élevées et d'utiliser plus efficacement les zones productives situées en aval de la frayère. La deuxième vague d'alevins dévalants correspondrait aux émergents tardifs et aux poissons soumis aux effets de la compétition territoriale.

Abstract

We have analysed the dispersal patterns (rate, duration, extent) of Atlantic salmon fry at emergence. The rate and duration of movement, and the distance travelled were measured in an experimental stream, located near S-Pée-sur-Nivelle, in SW France. A batch of 8 850 eyed eggs, from the grilse wild stock of the Nivelle River, was buried in the gravel substrate at the upstream end of a series of 5 sections, each 10 m long by 3 m wide. Shortly before emergence, drift nets equipped with fry traps were installed at the downstream end of each section. The nets sampled about 1/10 of the flow, except for those nets at the downstream end of the last section which collected all downstream moving fry. The traps were visited every morning and the fry enumerated. At the time of peak movements, samples were collected for length-weight measurements. At the end of the dispersal period, fry which had settled in the different sections (residents) were captured with electro-fishing gear and measured.

The pattern of downstream movement of fry in a set of drift nets was closely related to the pattern of emergence from an artificial redd upstream of the nets. Hence, emergence and downstream dispersion were well synchronized, at least during the first part of the dispersal from the redd. The time-frequency histograms of fry sampled every 10 m showed the same pattern and general shape (duration, skewness and kurtosis). Until the peak of captures, nearly 50 % of all downstream moving fry were caught within 5 days in each section. Dispersion from the redd over 50 m occurred within 3-4 nights. During this first period, the fry exhibited similar characteristics with respect to activity patterns and sizes. After the modal day of capture, catches were more evenly spread over time (about 10 days) and showed greater variability in relation to the distance travelled from the redd.

Resident fry were not uniformly distributed in the 5 sections : densities, as well as average weights, increased from upstream to downstream. Over 50 % of the surviving fry (75.3 % of planted eggs) settled within 50 m downstream from the redd.

Our results showed two waves of downstream dispersion. The first dispersal wave, occurring soon after emergence, was swift and implied large numbers of fry. It was not the result of territorial competition or density, since it occurred before the onset of aggressive behaviour. Rather, this first wave appeared as process to avoid the formation of clumps and allow for a more efficient use of the more productive zones, generally located downstream from the redd. The second wave of downstream moving fry corresponded to late emerging fry and to those fry which, 10-12 days after emergence, were displaced by territorial competition.

Keywords:

• Salmo salar,
• fry,
• emergence,
• dispersal