The Dynamics of a Stressed River Fish Community

Water Air Soil Pollut: Focus DOI 10.1007/s11267-006-9073-y

Invasives, Introductions and Acidification: The Dynamics of a Stressed River Fish Community

Bjørn Mejdell Larsen & Odd Terje Sandlund & Hans Mack Berger & Trygve Hesthagen

Received: 16 June 2005 /Accepted: 23 June 2006 # Springer Science + Business Media B.V. 2006

Abstract We describe the development of the fish been a simultaneous major increase in the occurrence community in the acidified and limed river Litleåna in and density of European minnow since 1997. Our southern Norway, and describe how chemical resto- results show that both brown trout and European ration, compensatory introductions of exotics, and minnow increase after liming. Minnow densities are accidental invasion of exotics interact to influence the negatively affected by low pH episodes in the river. population of the naturally occurring brown trout The growth rates of brown trout fry are negatively (Salmo trutta). The river Litleåna is a tributary to the correlated to minnow densities, indicating competi- river Kvina in Vest-Agder County, southern Norway. tion between the species. Brook trout densities have During the years 1996–2004, annual mean pH was decreased since liming started, and during the brown 4.9–5.0 and 6.1–6.4 above and below the liming trout recovery. facility, which was installed in 1994. Originally, brown trout was the only fish species in the river, Keywords brown trout . European minnow. but brook trout (Salvelinus fontinalis) have been brook trout . introduction . acidification . liming intentionally introduced, whereas European minnow (Phoxinus phoxinus) was introduced by accident. Fish densities were recorded by means of electrofishing 1 Introduction annually over the ten year period 1995–2004. Al- though close to extinction before liming was initiated, Pollution and introduced species are among the major brown trout fry densities increased from 1995 to threats to natural biodiversity. In southern Norway, 1999, with subsequent varying densities. There has acidification due to atmospheric deposition (acid rain) has caused local extinction of thousands of inland fish : : stocks. Simultaneously, intentional or unintentional B. M. Larsen (*) O. T. Sandlund T. Hesthagen introduction of non-native fish species has taken place Norwegian Institute for Nature Research, in most water courses. Over the last decade, reduced Tungasletta 2, 7485 Trondheim, Norway levels of acidifying emissions and various mitigation e-mail: [email protected] measures, in particular liming of water courses, has improved living conditions for fish. The introduced H. M. Berger species, however, are still there, and constitutes a Felt-BIO, Flygata 6, major hindrance for a full restoration of the aquatic 7500 Stjørdal, Norway ecosystem. Water Air Soil Pollut: Focus

The development of a river fish community during chemical restoration after acidification has previously been documented (e.g. Larsen & Hesthagen, 2004). The impact of introduced fish species has also been described (e.g., Sandlund & Bongard, 2000; Museth, Borgstrom, Hame, & Holen, 2003). However, the development of a riverine fish community during chemical restoration, with the simultaneous compli- cating presence of non-native fish species has not been described. This paper analyses the development of the fish C 5 LF community in the river Litlåna in southern Nor- 21 way, and discusses how environmental conditions, na compensatory introductions of exotics, and acci- River Litle dental invasion of exotics interact to influence the River Kvina 22 population of the naturally occurring brown trout Lake Galdalsvatnet (Salmo trutta).

23 2 Study Area 24

The river Litlåna is a tributary to the river Kvina in C 8 Vest-Agder County, southern Norway (Fig. 1). The N calcium content in bedrock and sediments is low, and 10 km the capacity to buffer acidifying depositions is poor. Originally, brown trout was the only fish species in Fig. 1 Location of the study area in River Litlåna with electrofishing localities (21–24), the liming facility (LF), and the river. Since 1980, when atmospheric pollution was localities for chemical analysis (C5–C8). The arrows indicate about to wipe out the brown trout populations, the the upper part of the river sections with anadromous salmonids more low pH tolerant brook trout (Salvelinus fontinalis) have been introduced several times. Natural reproduction of brook trout has been recorded. environment for some fish species. These drops in pH European minnow (Phoxinus phoxinus) was detected are usually associated with high precipitation, fol- in the river for the first time in 1997 (Berger, 1999), lowed by high water flows in the river. Calcium levels probably as a result of an accidental introduction in a in Litlåna above the liming facility has over the years tributary at least 10–12 years earlier (cf. Hesthagen & 1996–2004 only occasionally been above 0.5 mg Ca Sandlund, 1997). l−1. Below the liming facility, calcium levels are normally around 2 mg Ca l−1, but with occasional 2.1 Water Chemistry drops to just above 1 mg Ca l−1 (Kaste & Skancke, 2004; personal communications). To chemically restore the river, a liming facility was installed in 1994 (Fig. 1). Subsequently, water quality shows the typical trend of acidified rivers in this part 3 Sampling Methods of Norway, with mean pH at 4.9–5.0 above (station C5; Fig. 1) and 6.1–6.4 below (station C8; Fig. 1) the Sampling was performed annually with a portable liming facility (Fig. 2; Kaste & Skancke, 2004; back-pack electric fishing apparatus at four localities personal communication). Although pH around 6 below the liming facility. Fishing was performed in normally is adequate for brown trout and other August every year over the 10-year period 1995– sensitive fish species, the observed drops in pH 2004. Each locality was fished three times (removal values to below 5.5 (Fig. 2) indicates a precarious method; Bohlin, Hamrin, Heggberget, Rasmussen, & Water Air Soil Pollut: Focus

Fig. 2 pH above and below 8 the liming facility in Above liming Below liming Litlåna, from January 1996 to January 2005 7

6 pH

4 01.01.1996 01.05.1996 01.09.1996 01.01.1997 01.05.1997 01.09.1997 01.01.1998 01.05.1998 01.09.1998 01.01.1999 01.05.1999 01.09.1999 01.01.2000 01.05.2000 01.09.2000 01.01.2001 01.05.2001 01.09.2001 01.01.2002 01.05.2002 01.09.2002 01.01.2003 01.05.2003 01.09.2003 01.01.2004 01.05.2004 01.09.2004 01.01.2005 Date

Salveit, 1989; Zippin, 1958), and all fish was creased to a maximum of 40 fish 100 m−2 in 1999. identified to species. All body lengths were measured The dramatic increase was mainly caused by fry, in the field. which reached a density of 30 fish 100 m−2 that year. The density of older parr showed a more steady increase up to 1999. In 2001, however, brown trout 4 Results density decreased dramatically to only 7 fish 100 m−2. Again, the density differences were more dramatic for Brown trout was close to extinction and very few fry (0+) than for older fish. Since 2001 brown trout anglers took an interest in fishing in Litlåna before densities have increased steadily to nearly 30 fish liming was initiated (Kvinesdal kommune, 1999). 100 m−2. The overall tendency for brown trout over During the first two years of sampling (1995–96), the the period 1995–2004 is a significant density increase 2 total density of brown trout fry (0+) and older parr only for older parr (≥1+) (F1,8=10.36; r =0.56; was at only 6 and 5 fish 100 m−2, respectively P<0.05) (Fig. 3). Brown trout densities vary in a (Fig. 3). Subsequently, brown trout densities in- similar pattern at the four sampling stations.

Fig. 3 Densities of 50 brown trout fry (0+) (ns) Brown trout Fry (0+) Older parr and older parr (≥1+) (F1,8= 10.36; r2=0.56; P<0.05) in Litlåna, 1995–2004. The 40 sampling localities 21–24 were pooled. Fitted regression line is indicated 30

20 Density per 100 m²

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year Water Air Soil Pollut: Focus

Brook trout was first introduced to several small minnow at the upper sampling locality (by 1999), lakes in Litlåna in the beginning of the 1980s (Berger, densities of minnow has been 12–75 times higher 1999), and later spread to the main stem of Litlåna. than age-0 brown trout at this locality. Downstream The purpose of this introduction was to compensate migration of minnow has occurred in conjunction for the loss of acid-sensitive brown trout by establish- with the population expansion at the upper locality. ing a population of the acid-tolerant brook trout. Minnows were detected at locality 22, which is When the monitoring programme started in 1995, the 6.5 km downstream from locality 21, in 2002. The mean total density of brook trout was 9 fish 100 m−2, initial density at this locality was 11 fish 100 m−2, but and the population density reached a maximum in already one year later densities were above 200 fish 1998, at 11 fish 100 m−2 (Fig. 4). Brook trout has 100 m−2. In 2004, the first minnows were recorded at mainly been found at the three upper sampling locality 24, which is 8 km below the outlet of the stations, in particular at the station upstream of Lake 4.5 km long lake Galdalsvatn. This indicates a Galdalsvatn. Spawning has also been observed at this downstream migration rate of 19.0 km over seven locality. At the river stations, more than 90% of the years, i.e. approximately 2.7 km year−1. The first brook trout was fry (0+) during these years, indicating individuals of minnow to be caught at any locality a solid recruitment level. There is a significant were always large and adult fish. Thus, the adult decrease in the number of brook trout fry over the minnows (60–80 mm in length) are the pioneers of 2 ten year period 1995–2004 (F1,8=18.26; r =0.70; downstream migration, and become the first coloniz- P<0.005) (Fig. 4), and densities of brook trout have ers of new habitats in the river. In the established steadily decreased to less than 0.6 fish 100 m−2, i.e. minnow populations, recruitment is variable among the species is now nearly extinct in the river. years. The proportion of fish <40 mm (age ≤2+) was European minnow has most likely been introduced 42–44% in 1998 and 2003. In 2001 and 2002 no to a tributary to Litlåna in the 1980s. In the main small minnows were recorded, while the proportion in stem of Litlåna, our electrofishing surveys first the remaining years was 2–5%. detected the species at the upstream sampling station The length distribution of brown trout fry is nearly in 1997, in low densities (0.6 fish 100 m−2) (locality identical to minnows older than two years (Fig. 6). 21; Fig. 5). A swift population increase over the There was a significant reduction in trout fry lengths 2 following three years brought the minnow density at during 1995–2004 (F1,8=15.56; r =0.62; P<0.005). this sampling locality to a maximum of 162 fish There was a non-significant decrease with brown 100 m−2 in 2000. A temporary decrease in 2001 was trout density (P=0.12), but a significant decrease in followed by a further increase up to 537 fish 100 m−2 trout fry lengths with increasing minnow density 2 in 2003. During the years since the establishment of (F1,8=8.44; r =0.45; P<0.05) (Fig. 7).

Fig. 4 Densities of 25 brook trout fry (0+)(F1,8= 2 Brook trout Fry (0+) Older parr 18.26; r =0.70; P<0.005) and older parr (≥1+) (ns) in Litlåna, 1995–2004. The 20 sampling localities 21–24 were pooled. Fitted regression line is indicated 15

10 Density per 100 m² 5

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year Water Air Soil Pollut: Focus

Fig. 5 Densities of Europe- 600 an minnow in Litlåna, localities 21, 22 and 24 European minnow during 1995–2004 500 Locality 24 Locality 22 Locality 21

400

300

200 Density per 100 m²

100

0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year

We may assume that minnow became well estab- and the total densities of brown trout and brook trout, lished at locality 21 since 1999, and at locality 22 respectively. since 2003 (Fig. 5). During the post-establishment years, our data indicate that minnow densities increase with increasing minimum pH-values 5 Discussion (recorded during the preceding 10 months) up to approximately 5.7 (Fig. 8). At higher pH values, Our results from River Litlåna demonstrate that minnow densities appear not to be affected. This chemical restoration of acidified rivers by liming suggests that episodes with low pH values cause some favours brown trout over the intentionally introduced mortality in the minnow population. brook trout, while liming also benefits the acciden- The densities of brown trout and brook trout were tally introduced European minnow. not significantly related to water pH, nor was there It appears that minnow is very sensitive to water any significant correlation between minnow densities quality, and that densities decrease whenever there is

Fig. 6 Length distribution 20 European minnow Brown trout Brook trout of brown trout, brook trout and European minnow sam- pled in Litlåna during 1995– 2004 15

10 Percentage, %

0 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99 100-104 105-109 110-114 115-119 120-124 125-129 130-134 135-139 140-144 145-149 150-154 155-159 160- Length, mm Water Air Soil Pollut: Focus

Fig. 7 The relation between 80 densities of European minnow and the body length of brown trout fry (0+) in Litlåna, 1995–2004 2 70 (F1,8=8.44; r =0.45; P< 0.05). Fitted regression line is indicated

50 Length brown trout fry, mm Length brown

40 0 25 50 75 100 125 150 175 200 Number of European minnow per 100 m² any low-pH episode in the river. Neither brown trout minnows overlap almost completely in length sup- nor brook trout appear to be that clearly influenced by ports the notion that the two species may compete for pH in the post liming period. However, the number of food in relatively small streams where habitat overlap sampling years may be too low to demonstrate any between the two species may be significant. correlation in this regard. The brook trout has been introduced in many We would have expected brown trout fry densities Scandinavian water courses, both in good quality to be negatively correlated to minnow densities. This waters and in rivers subjected to acidification. The is not the case, but possible impacts of minnow on intention has been to establish the species as a target for brown trout are indicated by the decreasing growth of anglers, either in addition to the native brown trout, or brown trout fry with increasing minnow densities. The to replace the locally extinct brown trout populations. mechanism behind this impact in streams may be This obviously complicates the process to restore assumed to be competition for the same food sources native fish communities through chemical restoration (Hesthagen, Hegge, & Skurdal, 1992). The fact that of water quality. However, the development in Litlåna brown trout fry and the major part of the 2+ and older seems to support the assumption that brook trout loses

Fig. 8 Minnow density in 2.8 August at locality 21 (1998– 2004) and locality 22 (2003–2004) as a function 2.6 of the minimum pH recorded the preceding 2 10 months (F2,5=4.07; r = 0.62; P=0.09). The selected 2.4 years are subsequent to the “boom” phase years in population establishment. Fitted 2.2 regression line is indicated Log (minnow density) Log (minnow 2

1.8 5 5.2 5.4 5.6 5.8 6 6.2 pH Water Air Soil Pollut: Focus out in competition with brown trout under reasonable European minnow (Phoxinus phoxinus)intheRiver water quality conditions. In Scandinavian streams, it is Litleåna (Kvina watershed) in the County of Vest-Agder, Southern Norway). NINA Oppdragsmelding 580, 29 pp. commonly observed that brook trout is restricted to the (In Norwegian with English summary). uppermost source sections when coexisting with Bohlin, T., Hamrin, S., Heggberget, T. G., Rasmussen, G., & brown trout (Grande, 1982; Nyman, 1970). Saltveit, S. J. (1989). Electrofishing-theory and practice European minnow is swiftly becoming the most with special emphasis on salmonids. Hydrobiologia, 173, 9–43. common fish species in Norwegian inland waters Grande, M. (1982). Introduction and the present status of brook (Hesthagen & Sandlund, 1997; Taugbøl, Hesthagen, charr (Salvelinus fontinalis Mitchill) in Norway, EIFAC Museth, Dervo, & Andersen, 2002). Introductions Symposium “Stock enhancement in the management of occur through accidental releases by careless anglers, freshwater fisheries”. Budapest. Hesthagen, T., Hegge, O., & Skurdal, J. (1992). Food transfer of water among water bodies through choice and vertical distribution of European min- hydropower or water supply tunnels, etc. (Sandlund now, Phoxinus phoxinus, and young native and stocked & Bongard, 2000). Because the species is very brown trout, Salmo trutta, in the littoral zone of a sensitive to low pH, it will not establish strong subalpine lake. Nordic Journal of Freshwater Research, 67,72–76. populations in acid waters (Almer, 1972; Hultberg, Hesthagen, T., & Sandlund, O. T. (1997). Endringer i 1977). Our data also indicates that minnow survival is utbredelse av ørekyte i Norge: årsaker og effekter sensitive to acid episodes. However, chemical restora- (Changes in the distribution of European minnow, Phox- tion of acid waters will favour the minnow, and when inus phoxinus, in Norway: causes and effects). NINA Fagrapport 13, 16 pp. (In Norwegian with English the species live in sympatry, the intended restoration of summary). brown trout populations may face somewhat unex- Hultberg, H. (1977). Thermally stratified acid water in late pected problems. Minnow has a two-tiered role in winter – a key factor inducing self-accellerating processes relation to brown trout. Brown trout fry and minnow which increase acidification. Water, Air and Soil Pollution, 7, 279–294. may compete for food when they share habitats, which Kaste, Ø., & Skancke, L. B. (2004). Kvinavassdraget. 2. may most often occur in the nursery streams of brown Vannkjemi’ in direktoratet for naturforvaltning. Kalking i trout. Larger brown trout, however, may prey on vann og vassdrag. Effektkontroll av større prosjekter 2003, minnows, and thereby increase their growth rates, if DN-Notat 2004-2, pp. 103–104. (In Norwegian). Kvinesdal kommune (1999). Vassdragsplan for Kvinavassdra- their habitats overlap (Museth et al., 2003; Taugbøl et get. Kvinesdal kommune, Report, no pagination. (In al., 2002). In Litlåna, brown trout may very rarely Norwegian). reach a body size which enables it to prey on minnows. Larsen, B. M., & Hesthagen, T. (2004). Laks i kalkede vassdrag In summary, both brown trout and minnows are i Norge. Status og forventninger (Atlantic salmon in limed Norwegian rivers. Present status and expectations). NINA favoured by liming in acidified rivers and streams. Fagrapport 81, 25 pp. (In Norwegian with English Minnow densities have a negative impact on the summary). growth rates of brown trout fry, and may therefore Museth, J., Borgstrøm, R., Hame, T., & Holen, L. A. (2003). cause brown trout restoration not to fully reach its Predation by brown trout: a major mortality factor for sexually mature European minnows. Journal of Fish potential. Brook trout densities decrease as pH values Biology, 62, 692–705. are brought back to around 6, allowing brown trout Nyman, O. L. (1970). Ecological interactions of brown trout, and minnow densities to increase. However, there was Salmo trutta L., and brook trout, Salvelinus fontinalis no significant correlation between minnow densities (Mitchill) in a stream. Canadian Field Naturalist, 84, 343–350. and the densities of brown trout and brook trout, Sandlund, O. T., & Bongard, T. (2000). The freshwater respectively over our 10 years of sampling. environment. In: I. R. Weidema (Ed.), Introduced species in the Nordic countries. Nordisk ministerråd, København, NORD 2000, 13, 87–122. Taugbøl, T., Hesthagen, T., Museth, J., Dervo, B., & References Andersen, O. (2002). Effekter av ørekyteintroduksjoner og utfiskingstiltak-en vurdering av kunnskapsgrunnlaget Almer, B. (1972). Försurningens inverkan på fiskbestånd i (Effects of European minnow introductions and fish västkustsjöar (The impact of acidification on the fish stocks removal actions-an evaluation of existing knowledge). in Swedish west coast lakes). Information från Sötvatten- NINA Oppdragsmelding 753, 31 pp. (In Norwegian with slaboratoriet, Drottningholm 12–1972, 47 pp. (In Swedish). English summary). Berger, H. M. (1999). Ørekyte (Phoxinus phoxinus) i Litleåna I Zippin, C. (1958). The removal method of population estima- Kvinavassdraget i Vest-Agder 1998 (Distribution of tion. Journal of WildlifeManagement, 22,82–90.