35321.4.1.1 Feeding
After studying the stomach contents of 200 individuals
caught from April 1992 to December 1992, Jamet ( 1995 )
concluded that feed is numerically dominated by the
year- round consumption of Daphnia longispina. The biggest
contributors to weight gain are seasonally available food
opportunities, essentially Asellus aquaticus from April to
September and, char eggs from October to December, but
chironomid larvae, larval stonefl ies, and Cyclops abyssorum
are also consumed. May and August marks the period of
peak feeding activity.
21.4.1.2 Reproduction
Tracking shows that both males and females increase Gonado
Somatic Index ( GSI ) in August (Jamet 1995 ) and again in
September. In October, GSI decreases in males but increases
again in females. Female reproduction begins in October–
November, when GSI starts gradually decreasing until
December. Males appear to be present on spawning grounds
in Lake Pavin earlier than the females—a phenomenon fre-
quently noted in salmonids with the youngest individuals
fi rst to show when spawners start to gather (Gillet 2001 ). The
presence of males and females in differing proportions fol-
lowing the reproductive period has already been reported in
scientifi c literature (Dussart 1955 ).
Arctic char as a species displays very broad diversity in its
breeding habits. Rubin and Buttiker ( 1992 ), Dussart ( 1955 )
and Gillet ( 2001 ) report this diversity both in the breeding
season, in substrata ranging from blocks to fi ne sediment, and
in the depth of reproduction in the littoral zone, from 0.5 m
down to bottom depths of nearly 100 m. One criterion for
selection is the presence of non-silted substrates (Gillet 2001 ).
Lake Pavin has two well-known reproduction zones. For
years, successive lake farmers have fi shed these two zones to
increase the trout production and produce fry by artifi cial
fertilization in the basement of the hotel restaurant. Many
diving forays over the decades have confi rmed the presence
of these two main char breeding grounds, and reproductive
activity has been fi lmed a number of times (D Chassain CAP
1993–2003).
These egg-laying sites are located at the following coordi-
nates: X 0643.232, Y 2055.553, and X 0643.369, Y 2055.384;
(Extended Lambert II). These two sites correspond to talus
zones consisting of blocks ranging in size from a few dozen
centimeters to a meter and situated at a depth of between 10
and 18 m. Eggs are laid in rock crevices that have been
cleared of all fi ne sediment. While the physical characteris-
tics of the char breeding grounds are different, all authors
agree that the reproductive sites remain well-frequented
regardless of changing lake conditions (rising and subsiding
water levels).
Echo sounding surveys were carried out in December
1995 (Fig. 21.6 ). The screenshot below shows detections
performed while passing at a 90° angle over a supposed char
breeding ground. The equipment used was a SIMRAD
EY500 (70 kHz, 14°) echo sounder. Analysis of recorded
echo readings shows a high density of fi sh that the echoes
were able to distinguish as individuals. The focal concentra-
tion of adult fi sh is situated at a depth of between 8 and 20 m
and confi rms the observations reported from dives. However,
the survey was unable to fi nd any other major sites than those
already known and so does not exclude the presence of
smaller-sized char.
Spawning and its success are largely dependent on the
quality of the substrates, especially the oxygenation condi-
tions (Guillard et al 1992 ). A minimum concentration of
8 mg l −1 is essential for the entire duration from incubation to
resorption of the yolk sac, or a period of 450 °days.21.4.1.3 Growth
Growth has been analyzed by several methods.
The study of otoliths conducted by Audinot and Jamet
( 1998 ) cannot serve to plot a reliable growth curve. The
small number of individuals observed and uncertainties in
measurements in conjunction with inadequate equipment
preclude any attempt to propose a growth model here.
Similarly, a study of growth using a statistical method was
conducted as an initial approach (Bhatthacharya 1967 ), but
the results remain questionable because despite a relatively
large number of individuals captured, the nets out-selected
the youngest and oldest individuals, which interfered with
their distribution by size class. A growth model established
using this method would lead to over- or under-estimating
the K and Lα parameters of the Von Bertalanffy model.
Analysis of scales allows us to obtain a coherent growth
curve. In 2002, during monitoring on catches conducted by
the FDPPMA 63 , 232 individuals taken from the catches
were analyzed. The growth curve obtained from their real and
back-calculated ages is globally comparable to that obtained
from statistical analysis. Compared to other lakes, Arctic char
growth in Pavin is quite slow, especially from the second year
(Fig. 21.7 ) (Rubin 1990 ; Gerdeaux 2001a ; Zanella 2003 ), but
on a par with that of Lake Chauvet (63) located a few miles
from Lake Pavin, for which we carried out a study in 2009.21.4.2 Arctic Char Population TrackingGiven its economic importance and also its sensitivity to the
quality of its living environment, the numerical evolution of
the Arctic char population was tracked from 1995 to 2005. In
the 1990s, shrinking catch rates for Arctic char led managers
to scale back commercial fi shing and stock the lake with fi n-
gerlings to supplement natural reproduction. These fi nger-
lings are produced from broodstock caught in the lake and
raised in captivity until release.21 History of the Fish Fauna of Lake Pavin: A Population Heavily Infl uenced by Man?