Tommi Malinen

Comparison of gillnet and trawl in diurnal fish community sampling

The fish community in a shallow, eutrophic lake basin in southern Finland was sampled diurnally with gillnets and trawl. The differences in species number, relative abundances and length distributions were considered. The fish density estimations differed notably depending on the gear and diurnal period. The most abundant species in the trawl catch, smelt, was almost totally missing from the gillnet catch. The proportions of perch, roach, white bream and asp were higher in the gillnet catch. Gillnets regularly underestimated the proportion of small (<10 cm) individuals in size distributions. The trawl probably underestimated the density of piscivores. In the two gears, diurnal changes, in both fish catch and species distribution, were considerable. Both trawl and gillnets are needed to get more reliable figure of fish communities in shallow eutrophic lakes.

Direct and indirect mechanisms behind successful biomanipulation

Lake Vesijärvi is a relatively large (length 25 km; total area 110 km2), shallow (mean depth 6 m), but stratified lake in southern Finland. The Enonselkä basin (26 km2), surrounded by the city of Lahti, received its sewage effluent, and changed from a clear water basin with flourishing fisheries from the 1940–50s to one of the most eutrophic lake systems in Finland thereafter. In 1976, the sewage effluent was diverted, resulting in a temporary recovery of water quality. However, in the 1980s, massive surface scums of cyanobacteria degraded the water quality and arrested the recovery of the lake. A restoration strategy providing an ecologically sound basis for the management of the lake was initiated in 1987. This strategy involved biomanipulation (mass removal of coarse fish) together with conventional pollution control measures on discharges to the lake. Biomanipulation was chosen instead of much more expensive chemical and/or technical methods, such as chemical treatment or dredging of the profundal sediment. The large-scale biomanipulation trial was carried out in the Enonselkä basin during 1989–93. Following the mass removal of coarse fish (1000 metric tons of fish; mainly roach and smelt), the biomass of cyanobacteria collapsed concomitantly with a decline of total phosphorus concentration from 45 to 35 mg P m−3, and with an increase of Secchi depth from 1 m to 3.5 m. These observed improvements in the water quality were matched with a large decline in roach-mediated phosphorus movement from littoral to pelagial, from 100 mg P m−2 in 1989 to 15 mg P m−2 in 1993. Year-to-year variation within the littoral communities, and in the recruitment of fish, could in this way cause large oscillations in the whole ecosystem. The involvement of local people (fishermen, farmers etc.) in controlling non-point nutrient loading and fish stock development, is of prime importance for the long term success of lake restoration.

Predation by brown trout (Salmo trutta) along a diversifying prey community gradient

Predation has a fundamental role in aquatic ecosystems, but the relative importance of factors governing prey selection by predators remains controversial. In this study, we contrast five lakes of a subarctic watershed to explore how prey community characteristics affect prey selection and growth rate of the common top predator, brown trout (Salmo trutta). The lakes constitute a distinct gradient of different coregonid prey fish, ranging from monomorphic common whitefish (Coregonus lavaretus) to polymorphic whitefish co-occurring with vendace (Coregonus albula). The brown trout was a morph–species- and size-specific pelagic predator, selecting the small-sized, pelagic whitefish morph or vendace over the benthic whitefish morphs. In all lakes, the average prey size increased with predator size, but small-sized prey were also included in the diet of large predators. The selection of small-sized, pelagic prey fish appeared to be a favourable foraging strategy for the brown trout, yielding higher growth rates...

A metalimnetic oxygen minimum indirectly contributing to the low biomass of cladocerans in Lake Hiidenvesi - a diurnal study on the refuge effect

The diurnal vertical migrations of smelt (Osmerus eperlanus), larvae of phantom midge (Chaoborus flavicans) and cladoceran zooplankton in eutrophic Lake Hiidenvesi were studied in order to clarify the factors behind the low zooplankton biomass. In the study area, an oxygen minimum occurred in the metalimnion in the 10–15 m depth. No diurnal fluctuations in the position of the minimum were observed. Cladocerans inhabited the epilimnion throughout the study period and their vertical movements were restricted to above the thermocline and above the oxygen minimum. C. flavicansconducted a diurnal migration. During the day, the majority of the population inhabited the 12 – 15 m depth just in the oxygen minimum, while during darkness they were found in the uppermost 8 m. Smelts started ascending towards the water surface before sunset and reached the uppermost 3 m around 23:00. During daytime, the majority of smelts inhabited the depth of 7–9 m, where the water temperature was unfavourably high for them (18 °C). Smelts thus probably avoided the steep oxygen gradient in the metalimnion, whereas Chaoborusused the oxygen minimum as a refuge against predation. Those smelts that were found in the same water layers as Chaoborusused the larvae as their main prey. The metalimnetic oxygen minimum thus seemed to favour the coexistence of vertebrate and invertebrate predators, leading to a depression of cladoceran zooplankton.

The supremacy of invertebrate predators over fish: factors behind the unconventional seasonal dynamics of cladocerans in Lake Hiidenvesi

In the eutrophic Lake Hiidenvesi (30.3 km 2 ), the early summer biomass maximum of cladoceran zooplankton is missing and the highest biomass is found unusually in mid summer. The contribution of the phantom midge Chaoborus flavicans to the unconventional seasonal dynamics of cladocerans was studied. The seasonal development of the C. flavicans population and the zooplankton community was studied together with food consumption by chaoborids and the most common vertebrate predator, smelt. The population density of Chaoborus peaked in June and collapsed in mid summer. Thereafter, another peak occurred at the end of August. The maximum density of chaoborids in the water column varied between 5,500 and 8.700 inc./m 2 . In early summer, the most common food items of Chaoborus were bosminids while later in summer the share of daphnids increased. Food consumption by the Chaoborus population peaked at 26-46 μg Cl -1 d -1 in June and clearly exceeded the production rate of cladocerans. Cladoceran consumption by smelt population reached its maximum, 1.13 μg Cl -1 d -1 , in early July. Cladocerans showed a 20-40 μg C/l biomass peak in July-August, coincidently with the minimum density of chaoborids. When predation pressure decreased, the mean size of cladocerans increased at 10-15 m, the depth occupied by chaoborids during daytime. The results suggest that in early summer the high predation pressure prevents cladocerans from reaching a sufficient body length to defend them against Chaoborus and that the timing of the cladoceran biomass peak is determined by the relaxation of chaoborid predation.

Implications of rudd (Scardinius erythrophthalmus) herbivory on submerged macrophytes in a shallow eutrophic lake

The role of rudd (Scardinius erythrophthalmus) herbivory was studied in Kirkkojärvi, a shallow and turbid basin in Lake Hiidenvesi, Finland. The submerged species dominating in the rudd diets were Potamogeton obtusifolius, Ranunculus circinatus, Sparganium emersum, bryophytes, and filamentous algae. Plant consumption estimated with bioenergetics modelling increased with fish age, being highest in late summer concomitant with the macrophyte biomass peak. Depending on the age structure, a rudd biomass of 20 kg ha−1 consumes 18–23 kg of macrophytes ha−1 a−1, while a rudd biomass of 100 kg ha−1 results in plant consumption of 92–115 kg ha−1 a−1. Although, rudd seemed to feed rather unselectively on suitable-sized and edible plants, some species abundant in the littoral, such as Myriophyllum verticillatum and pleustophytic Ceratophyllum demersum, were not found in rudd guts, indicating selective plant consumption. In Kirkkojärvi, selective grazing by rudd and increased turbidity and high nutrient levels partly caused by bottom dwelling cyprinid fish, may promote the inedible and pleustophytic macrophytes, which have increased in Kirkkojärvi during the past decades.

Species introduction promotes hybridization and introgression in Coregonus: is there sign of selection against hybrids?

Species introductions are considered one of the major drivers of biodiversity loss via ecological interactions and genetic admixture with local fauna. We examined two well‐recognized fish species, native whitefish (Coregonus lavaretus) and introduced vendace (Coregonus albula), as well as their morphological hybrids in a single lake to test for selection against hybrids and backcrosses in the wild. A representative random subsample of 693 individuals (27.8%) was taken from the total catch of coregonids. This subsample was examined with the aim to select c. 50 individuals of pure whitefish (n = 52), pure vendace (n = 55) and putative hybrid (n = 19) for genetic analyses. The subsequent microsatellites and mitochondrial (mt) DNA analyses provided compelling evidence of hybridization and introgression. Of the 126 fish examined, four were found to be F1, 14 backcrosses to whitefish and seven backcrosses to vendace. The estimates of historical gene flow suggested higher rates from introduced vendace into native whitefish than vice versa, whereas estimates of contemporary gene flow were equal. Mitochondrial introgression was skewed, with 18 backcrosses having vendace mtDNA and only three with whitefish mtDNA. Hybrids and backcrosses had intermediate morphology and niche utilization compared with parental species. No evidence of selection against hybrids or backcrosses was apparent, as both hybrid and backcross growth rates and fecundities were high. Hybrids (F1) were only detected in 2 year‐classes, suggesting temporal variability in mating between vendace and whitefish. However, our data show that hybrids reached sexual maturity and reproduced actively, with backcrosses recorded from six consecutive year‐classes, whereas no F2 individuals were found. The results indicate widespread introgression, as 10.8% of coregonids were estimated to be backcrosses.

Density and habitat shifts of a roach (Rutilus rutilus) stock assessed within one season by cohort analysis, depletion methods and echosounding

Abstract The stock estimates of roach obtained by cohort analysis, depletion methods and echosounding are compared. The study was carried out in summer 1993, the last year of the 5 year mass removal of coarse fish in L. Vesijarvi in southern Finland. Additionally, the habitat shifts by the roach were explored based on the differences in the density estimates given by cohort analysis and echosounding. The estimates on the density of the stock in the beginning of the fishing season given by cohort analysis (1276–1423 fish ha −1 , depending on the natural mortality rate) and the depletion method (1235–1383 fish ha −1 ) were in agreement. The echosounding estimates in the early summer (500 fish ha −1 ) were much lower than those given by cohort analysis, whereas in the late summer echosounding gave higher estimates than cohort analysis. The differences were due to the habitat shifts by the roach. The differences in the stock estimates given by the different assessment methods can be used especially in studying the timing of the migrations.

Fish stocks of Lake Vesijärvi: from nuisance to flourishing fishery in 15 years

Original goals ofthe biomanipulation project ofLake Vesijãrvi, southem Finland, were to eliminate blooms of cyanobacteria and re-establish a sustainable fishery (KETO & SAMMALKORPI 1988). From 1989-93, b1ooms of cyanobacteria were eliminated with mass remova1 of 1018metric tons coarse fish, main1y roach (52%), Rutilus rutilus (L.) and smelt (27%), Osmerus eperlanus (L.), from the Enonselkã basin (2600 ha, HORPPILA & PELTONEN 1994, HORPPILA et al. 1996, 1998). Fishing with seine and fyke nets to remove 60-86 metric tons of coarse fish annually from 1994-96 prevented new increase o f roach stock (PELTONEN et al. 1999a).

Optimal sampling and traditional versus model-based data analysis in acoustic fish stock assessment in Lake Vesijärvi

Abstract Optimal sampling strategy and data analysis methods were studied for pelagic fish stock assessment in eutrophic Lake Vesijarvi in southern Finland based on 10 surveys conducted in 1993. The systematic sampling scheme was applied with parallel transects. The precision of the fish density estimates in unstratified and post-stratified sampling were compared with each other and with the precision that could be obtained by an optimum effort allocation. Because the frequency distribution of fish density in the primary sampling units is typically skewed, a model-based estimation was also applied in which the distribution is approximated with a known distribution. The best normalizing transformation was found with Boxs and Coxs test. These model-based estimates were compared with the estimates derived from non-transformed data. Post-stratified sampling by depth areas produced more precise estimates than unstratified sampling (reduction in the variances about 40%), reduced the bias caused by spatial autocorrelation and eliminated the bias caused by unproportional sampling of depth areas. The most precise estimates (reduction in the variances compared with post-stratification about 50%) would be given by optimum allocation, which suggests that more effort should be allocated to the deep areas. The optimal fraction of sample size to deep areas increased in the course of the summer (about 20% in June and over 50% in August). The model-based estimation produced a slight gain in precision with the post-stratified data. But, surprisingly, with unstratified data and with optimum allocation more precise estimates were given by initial untransformed data. In conclusion, we do not recommend the model-based estimation (including logarithmic transformation) in lakes if a suitable survey time is restricted by fish behaviour (diel migrations, schooling behaviour), because of difficulties in obtaining adequate sample size to determine appropriate transformation.