Monitoring the Responses of Freshwater Ecosystems to Climate Change - Climate Change Impacts on Freshwater Ecosystems - page 104

are assumed to be an important cause of winter mortality, strong negative effects
on the energy budget can be expected (Finstad et al . 2004).
In the river Rhone (France), analysis of long-term fish data revealed that the
variability of fish abundance was correlated with discharge and temperature
during the reproduction period (April-June). Low flows and high temperatures
coincided with high fish abundance. In line with temperature increase, southern,
thermophilic fish species, e.g. chub ( Leuciscus cephalus ), and barbel ( Barbus sp.)
progressively replaced northern, cold-water fish species, e.g. dace ( Leuciscus
leuciscus ) (Daufresne et al . 2004). With increasing water temperature the
incidence of proliferative kidney disease increased in Switzerland, and populations
declined by up to 66% for brown trout (Hari et al . 2006). High temperatures
may induce an upward migration of fish at their thermal limits: e.g. brown trout
(Heggenes & Dokk 2001; Hari et al . 2006), charr ( Salvelinus sp.), and salmon
(Carpenter et al . 1992) or an alteration in their migration behaviour, increasing
the use of thermal refugia in cooler tributaries (chinook salmon, Cole et al . 1991).
The timing of migration and spawning is also closely linked to thermal state
(Salinger & Anderson 2006). Water temperature increase and earlier snowmelt
causes an earlier return of adult salmon and alewives, and also affects the timing
of migration and spawning of brown trout (Huntington et al . 2003). Several
metrics related to the composition of fish communities and key species are suited
as indicators of these changes (Table 5.2).
In a study of climate change and fish biodiversity, a list of river fish species
occurring in 152 European river basins was collected from an intensive literature
survey. In all, the fish database contained 306 fish species. To describe fish
autecology, 21 fish species traits were considered. Species traits were diverse but
most of them concerned reproduction (e.g. fecundity), feeding (e.g. trophic
group) or morphology (e.g. body length). Each trait was coded in several
categories ('modalities') totalling 72 trait modalities. The 306 fish species present
in the database were assessed for these attributes based on the Fish-based
Assessment Method for the Ecological status of European rivers (FAME) database
(Schmutz et al . 2007) or the existing unpublished and published literature and
were completed by expert judgement. The trait composition of fish assemblages
was thus available for each of the 152 basins. First, 9 ecoregions (Illies 1978)
describing a latitudinal gradient were selected. Only the river basins contained
within a single ecoregion were retained. Fish trait composition was then compared
among the different ecoregions (Fig. 5.2). A north-south gradient was observed
for fish body length and traits related to reproduction, but not for traits describing
habitat and feeding. In the southern ecoregions, fish assemblages are characterized
by small fish, early maturation, small eggs with a short incubation period and the
absence of a strategy for egg protection. Those traits are biologically concordant
and counter to strong investment in reproduction, which is one of the main
features of the more northern cold-water species.
Fish trait composition thus appeared to vary strongly along the latitudinal gradient.
To understand more precisely this latitudinal structure, the fish trait composition was
related to environmental variables that are good descriptors of the latitudinal gradient.
These environmental data were extracted from 0.5°× 0.5°grid data (CIESIN 2005)