- Top of page
- Materials and methods
Food acquisition provides the necessary energy for life maintenance, growth and reproduction of living organisms and is instrumental in the ecology and evolution of animal populations. Food availability may, however, be restricted, resulting in competitive interactions for limited resources (e.g. Keddy 1989). As individuals within the same species have highly similar resource requirements, the intensity of intraspecific competition is assumedly severe and strongly related to the population density. Intraspecific competition is furthermore expected to result in reduced resource acquisition and diminished individual growth and development rates, which in turn may lead to increased mortality rates. In mammals and birds with highly fixed and pre-determined individual growth and development trajectories manifested in, e.g. characteristic adult sizes, severe resource limitation and starvation may directly influence survival (Piatt & Pelt 1997; Mduma, Sinclair & Hilborn 1999; Keymer et al. 2001; Camphuysen et al. 2002; Reid & Forcada 2005). Fish and other organisms with a high plasticity in somatic growth and other life-history traits, may in contrast have a higher capacity to buffer acute mortality effects of food shortage and hunger, even being able to survive substantial time periods with feeding rates below the maintenance level (Jobling 1994; Fishelson 1997; van Dijk, Hardewig & Hölker 2005). Fish populations may therefore persist at exceedingly variable levels of food availability and intraspecific competition, facilitating studies of food acquisition and resource competition under highly contrasting scenarios.
The growth performance of most fish species is indeterminate and flexible, and may be affected by several biotic and abiotic factors (Werner 1986; Wootton 1998). An inverse relationship between population density and fish growth has been recognized for many years, and this relationship is usually assumed to be related to density dependence of food availability (Beverton & Holt 1957; Backiel & Le Cren 1978; Jenkins et al. 1999; Post, Parkinson & Johnston 1999; Lorenzen & Enberg 2002). Experimental studies and bioenergetic considerations further emphasize a close connection between food consumption and individual growth (Brett & Groves 1979; Jobling 1994; Lucas 1996). Only a few field studies have, however, been carried out comparing food acquisition and growth rates of fish (e.g. Hayward & Margraf 1987; Boisclair & Leggett 1989), and conflicting conclusions have also emerged from these studies (Boisclair & Leggett 1990; Hayward 1990; Hewett, Kraft & Johnson 1991; Hewett & Kraft 1993). Furthermore, in situ studies of the relationship between population density and quantitative food consumption are rare both in fish and other animal populations, and studies addressing population density, food acquisition and somatic growth rates simultaneously are to our knowledge missing.
In temperate and subarctic lakes, high-density fish populations with stunted growth rates commonly occur and are recognized as a large problem for the fish management (Burrough & Kennedy 1979; Amundsen 1988; Donald & Alger 1989; Langeland & Jonsson 1990; Ylikarjula, Heino & Dieckmann 1999; Amundsen et al. 2002; Aday et al. 2005). Density reductions have been proposed as a potential means to alleviate stunting in these overcrowded populations (e.g. Amundsen 1988; Donald & Alger 1989; Langeland & Jonsson 1990). In this respect, a large-scale stock depletion experiment was carried out from 1984 to 1989 in the stunted Arctic charr Salvelinus alpinus (L.) population of the subarctic lake Takvatn, northern Norway, resulting in highly improved growth performance of the remaining Arctic charr (Amundsen, Klemetsen & Grotnes 1993; Klemetsen et al. 2002). In the present study, analyses of the food consumption rates of the charr were carried out during the ice-free season on six annual occasions prior to, under, and after the density manipulation experiment, covering a total time span of two decades (1980–99) with large variations in fish density. It has previously been documented that the stunted charr population in the lake had low food consumption rates, suggesting a restricted food supply at high population densities (Amundsen & Klemetsen 1988; Amundsen 1989). Addressing the long-term effects of the stock manipulation experiment, it was therefore hypothesized that population density has a major influence on food acquisition rates of individual fish through intraspecific competition. Furthermore, it was hypothesized that individual growth rates are highly dependent on the quantitative food consumption, resulting in a negative relationship between population density and somatic growth performance. A co-occurring fish species, brown trout Salmo trutta L., experienced competitive release and increased population density after the density depletion of Arctic charr (Klemetsen et al. 2002), and potential additive effects of increased trout density on the feeding and growth rates of charr were also considered.
- Top of page
- Materials and methods
Resource limitations, i.e. resource demands in excess of the immediate supply, are considered a pre-requisite for competitive interactions to occur (Keddy 1989), but are not necessarily easy to document under natural conditions. Food resource limitations and density-dependent intraspecific competition have furthermore been keystone parameters in the development of quantitative population ecology (e.g. Sinclair 1989), ever since the introduction of the logistic growth model (Pearl & Read 1920). Nevertheless, intraspecific competition is generally underrepresented in competition studies (Gurevitch et al. 1992), and surprisingly few studies have addressed the density-dependent effects of resource limitation and intraspecific competition on food acquisition and growth within animal populations even though these aspects are implicit in most population regulation models (Sinclair 1989, 2003; Bayliss & Choquenot 2002; Krebs 2002; Sibly and Hone 2002). In the present contribution, we demonstrate resource limitation and large density-dependent impacts on the per capita feeding and growth rates of the Arctic charr population under study. Apparently, the fish suffered from severe food restrictions and intraspecific competition at high population density, resulting in low food consumption and somatic growth rates of the individuals. A similar conclusion was also reached in studies of a dense population of European perch Perca fluviatilis (Persson 1983, 1987). The whole-lake experimental manipulation in Takvatn strongly reduced the charr population density and concurrently there was a large increase in the food consumption rates. Evidently, the reduced population size led to released intraspecific competition and enhanced feeding conditions for the remaining charr. Over the total study period, the feeding rates of the Arctic charr changed in firm correspondence with the variations in fish density, demonstrating a strong inverse relationship between population density and food consumption rates. Over the same time period, there were also large variations in the growth rates of charr that were closely correlated to the changes in food consumption rates and hence showed a similar negative correlation to the population density. The study thus demonstrates that the food consumption rates of the Arctic charr represent the connective link between fish density and individual growth rates, apparently being highly influenced by intraspecific competition for limited resources.
Stunted growth in overcrowded fish populations has long been attributed to food shortage (e.g. Alm 1946; Mittelbach 1983; Donald & Alger 1989; Heath & Roff 1996), but there has been a lack of quantitative evidence. Prior to the fish removal experiment in Takvatn, the charr population was severely stunted (Amundsen & Klemetsen 1988; Amundsen et al. 1993). The density reduction of the population led both to a strong increase in food consumption rates and large enhancements of the somatic growth rates. Thus, our results strongly support the assumption that stunted growth is related to food shortage. Stunting in fish has also been explained by genetic differences in growth abilities (Murnyak, Murnyak & Wolgast 1984), but for the Takvatn charr the stunted growth was evidently not due to genetic constraints, a conclusion that has also been reached for other freshwater fish species (Heath & Roff 1987). A recent study of guppies Poecilia reticulata does, however, demonstrate that a long-term subjection to low food resource levels may favour the evolution of slow growth (Arendt & Reznick 2005), suggesting that slow somatic growth performance eventually may become a fixed genetic trait in fish populations that remain stunted for a multitude of generations.
The strong negative relationship between individual growth rates and population density of Arctic charr was best described by a negative power curve. Our results are consistent with Jenkins et al.'s (1999) findings of a negative power relationship between mean fish size and density of brown trout in streams. Similar negative power trajectories have also been found to describe the growth-density relationships in other studies of stream-living salmonids (Imre, Grant & Cunjak 2005; Lobón-Cerviá 2005). These findings are consistent with comparable observations of a curvilinear relationship between growth rate and population density both in lotic and lentic freshwater fish populations (e.g. Crisp 1993; Pierce, Tomcko & Margenau 2003), in marine fish (Lorenzen & Enberg 2002) and in other poikilotherms (Van Buskirk & Smith 1991; Werner 1994), suggesting that this may represent a general ecological phenomenon. An important consequence of this pattern is that responses in growth rates to changes in population density will tend to be strong at low densities, whereas density-dependent growth may be difficult to perceive and even remain undetected at high-density levels (Jenkins et al. 1999; Lobón-Cerviá 2005). The negative power relationship between growth rate and population density may have important practical consequences with respect to the management and potential rehabilitation of stunted fish populations. At the high densities typically observed in stunted fish populations (e.g. Alm 1946; Burrough & Kennedy 1979; Amundsen 1988; Donald & Alger 1989; Langeland & Jonsson 1990; this study), a very large density perturbation is needed to achieve any significant impacts on the mean individual growth rates. Rehabilitation of overcrowded and stunted fish populations through fish removal may therefore be a difficult task (Langeland & Jonsson 1990; Amundsen et al. 2002). An important and unprecedented finding of the present study is that also the food consumption rates exhibited a negative power relationship with population density. This suggests that the negative power trajectories frequently observed between growth rate and population density are in fact a direct result of density-dependent impacts on the food intake rates. Thus, it is apparently not the somatic growth rates per se that have a negative power relationship with population density, but rather the quantitative food acquisition of the individuals.
Whereas numerous studies have examined the relationship between fish density and individual growth rates (e.g. Backiel & Le Cren 1978; Tonn, Holopainen & Paszkowski 1994; Jenkins et al. 1999; Post et al. 1999; Lorenzen & Enberg 2002; references therein), only a few have examined the relations between food consumption and growth. In a series of studies on yellow perch Perca flavescens, Boisclair & Leggett (1989) found a lack of correlation between field estimates of food consumption and growth rates, and concluded that the quantity of food consumed by fish only played a secondary role in explaining among-population variability in fish growth. Hayward & Margraf (1987), on the other hand, related growth differences between yellow perch in two separate lake basins of Lake Erie primarily to interbasin differences in food supply and feeding rates. Their conclusion was questioned by Boisclair & Leggett (1989), and this initiated a debate on the relationship between growth and consumption rates (Boisclair & Leggett 1990; Hayward 1990; Hewett et al. 1991; Hewett & Kraft 1993). The dissension demonstrates that the in situ relationship between growth and food consumption may not be as straightforward as could be assumed from energy budget considerations and laboratory studies (e.g. Brett & Groves 1979; Lucas 1996). However, the strong positive relationship observed between food consumption and somatic growth rates in the present study seems unambiguous, and confirms that the food consumption is an important link between population density and somatic growth. Apparently the density reductions of the Arctic charr in Takvatn released the intraspecific competition for food in the fish population, leading to increased food intake and a subsequent improvement of the growth rates of the charr.
Impacts of fish removal as observed in the Arctic charr population in Takvatn have also been shown at the community level. After density reduction of roach Rutilus rutilus, both food consumption and growth rates of the sympatric perch population increased (Persson 1986; Persson et al. 1993). Similarly, Hayes, Taylor & Schneider (1992) showed that removal of white suckers Catostomus commersoni increased both feeding and growth rates of yellow perch. Effects at the community level were apparently also present in the Takvatn experiment. Arctic charr and brown trout are two salmonid species that are known to exhibit intense resource competition (Nilsson 1967; Svärdson 1976; Forseth et al. 2003; Klemetsen et al. 2003). Brown trout became very rare in Takvatn after overexploitation and the introduction of Arctic charr (Klemetsen et al. 2002). Similar developments have been observed in many other Scandinavian lakes where Arctic charr has been stocked (Svärdson 1976). After the density reduction of Arctic charr in Takvatn there was a distinct increase in the density of brown trout (Klemetsen et al. 2002), apparently as a result of a density compensation following competitive release. The inclusion of brown trout abundances in the present density considerations provided slightly stronger correlations and curve fits for the relationships of food consumption and growth rate with population density, suggesting that the increased abundance of brown trout also affected the feeding and growth of the Arctic charr. Thus, both inter- and intraspecfic competitive interactions were apparently influencing the food acquisition rates of the Arctic charr, but intraspecific competition was definitely the key determinant under the observed fish community composition.
Abiotic factors like seasonal temperature developments may also influence feeding and somatic growth rates of fish (Wootton 1998). Some interannual variations in water temperature were seen over the study period in Takvatn (Table 2). Nevertheless, both food consumption and individual growth rates of the Arctic charr exhibited strong negative correlations with population density, providing a strong case for the conclusion that density-dependent intraspecific competition was the major determinant for these parameters. Density-dependent intraspecific competition has been suggested to be an essential contributor to population regulation and stabilization through effects on reproductive output and mortality rates (Sinclair et al. 1985; Sinclair 1989). Increased mortality rates may occur indirectly as a result of increased vulnerability to, e.g. predation at smaller body sizes, or directly as a result of decreased life sustainability at low resource availabilities. The latter development is particularly likely to be seen in birds and mammals that are largely unable to adjust their growth and life-history performances as a response to increased resource limitations (see, e.g. Piatt & Pelt 1997; Camphuysen et al. 2002; Reid & Forcada 2005). For fish that usually have highly flexible growth patterns, increased resource limitations due to increased population densities are in contrast more likely to be followed by decreased growth rates and smaller body sizes, potentially leading to increased mortality from size-selective predation. However, if predation rates are low, the potential for population regulation through such indirect effects of density-dependent intraspecific competition may be absent or low. In temperate and Arctic lacustrine fish populations the main predators are usually other piscivorous fish species or intraspecific predators, i.e. cannibals (Popova 1978; Keast 1985; Amundsen 1994b; Mittelbach & Persson 1998; Amundsen et al. 2003; Byström 2006). For the Takvatn charr population it has been suggested that the development of overcrowding and stunting was related to overexploitation of large-sized piscivorous brown trout and cannibalistic Arctic charr from extensive use of large mesh-sized gillnets (Amundsen et al. 1993; Klemetsen et al. 2002). In the absence of predators, the response to an increasing population density may be restricted to reductions in the somatic growth rates of the Arctic charr without any detrimental effects on survival. Hence, there will be little or no regulation of the population growth and overpopulation and stunting may be a likely endpoint in the population development under such circumstances. Large flexibility in growth performance is a characteristic trait for most temperate and Arctic freshwater fish species and may represent a useful adaptation to harsh and unpredictable environmental conditions. Apparently this flexibility is also a significant element in the prevalent occurrence of overcrowding and stunting in freshwater fish populations in these regions.
Several important conclusions can be inferred from the long-term effects of the density manipulation experiment in Takvatn. First, resource limitation appeared to be a significant issue over the whole range of population densities observed, leading to competition for the available food resources. Secondly, density-dependent competitive interactions negatively influenced the per capita food consumption rates of the Arctic charr. Thirdly, there was a strong and positive correlation between quantitative food intake and somatic growth, resulting in a negative correlation between population density and individual growth rates. The findings also confirm that reduced food acquisition due to severe resource limitations at high population densities is the ultimate cause for stunting in Arctic charr. Furthermore, the food consumption is shown to represent the connective link between fish density and individual growth rates, apparently being highly influenced by intraspecific competition for limited resources. Moderate effects of interspecific competition from brown trout could also be revealed. Finally, both food consumption and somatic growth rates exhibited negative power relationships with population density, suggesting that large perturbations may be necessary to induce improved feeding conditions and higher somatic growth rates in high-density fish populations.