SEARCH

SEARCH BY CITATION

Keywords:

  • aquatic insects;
  • dispersal;
  • forestry;
  • land use

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information
  • 1
    There have been widespread changes in land use in the uplands of the UK but the implications for dispersal of adult stages of aquatic invertebrates are poorly known. We estimated the lateral dispersal of adult aquatic insects (Plecoptera, Trichoptera, Ephemeroptera) in seven small, upland streams draining catchments under three categories of land use (coniferous plantation forest, cleared forest, moorland).
  • 2
    Malaise traps were set out in transects perpendicular to each stream. More than 29 000 adult insects were taken, distributed among 15 species of stoneflies, 40 species of caddisflies and eight taxa of mayflies. Overall species diversity and equitability were highest in the moorland catchments, and few species were numerous in all catchments.
  • 3
    Nearly all the mayflies were taken in the moorland catchments, where caddisflies were also most abundant. Fewest stoneflies were taken in the forested catchments.
  • 4
    The vast majority of insects were taken either directly over, or very close to, the stream channel. Half the stoneflies were taken within 18 m of the channel, while 90% had travelled less than 60 m. Caddisflies and mayflies travelled even shorter distances. Although there were differences in lateral dispersal between some catchments, there was no overall effect of land use.
  • 5
    The overall sex ratio in stoneflies and mayflies in the riparian zone was close to 1 : 1 and lateral dispersal was similar between the sexes. Male mayflies outnumbered females in the riparian zone and males travelled further from the stream, on average, than females. In catches taken directly over the stream, female stoneflies outnumbered males.
  • 6
    Regardless of land use, the flight of mayflies and caddisflies was concentrated along the stream, rather than perpendicular to it. This was also true for two numerous stoneflies (Amphinemura sulcilcollis and Protonemura meyeri) and for female stoneflies overall.
  • 7
    Synthesis and applications. The stream corridor, including the riparian strip extending 10–20 m on either side of the channel, is the main habitat for adult aquatic insects, and its management may affect the biodiversity of aquatic communities. The stream corridor is also revealed as the main ‘highway’ for adult dispersal. While there is no evidence from this study of an effect on interstream dispersal of land use elsewhere in the catchment, such an effect cannot yet be refuted because rare long-distance dispersal events are difficult to record.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

For many organisms and ecosystems, basic information is still lacking on the population consequences of changes in land use and habitat fragmentation (McCullough 1996). Freshwaters are naturally occurring examples of fragmented habitats scattered through a landscape that is subject to anthropogenic changes. However, as a key component of the population dynamics of aquatic insects, dispersal through the landscape has often been ignored and thus its impact on local ecological processes has been underestimated (Palmer, Allan & Butman 1996; Rundle et al. 2002). No ecosystem is ‘closed’ to its neighbours, however, and the dynamics of populations in catchments depend upon ‘inputs’ and ‘outputs’ as well as on intrinsic processes (Peckarsky, Taylor & Caudill 2000; Hanski 2001; Bohonak & Jenkins 2003; Caudill 2003a).

Aquatic insects often have a larval stage confined to the aquatic environment, while the adult is terrestrial. Thus, while dispersal of the larvae can theoretically occur between populations if their habitats are linked by water, dispersal between unconnected aquatic systems requires overland movement by the adults. Mechanisms of dispersal within streams, such as drift and upstream larval movement, have been studied extensively (Waters 1972; Söderström 1987; Brittain & Eikeland 1988; Mackay 1992; Allan 1995; Jackson, McElravy & Resh 1999; Elliott 2002a; Elliott 2002b). Except for some assessments of the direction of flight by adult insects along the stream corridor, however, our knowledge of the dispersal of adult aquatic insects is still in its infancy (Petersen et al. 1999b; Kopp, Jeschke & Gabriel 2001; Caudill 2003b).

Only recently has more attention been given to the lateral dispersal (i.e. that away from the stream channel) of adult aquatic insects (Jackson & Resh 1989a; Sode & Wiberg-Larsen 1993; Collier & Smith 1995; Kovats, Ciborowski & Corkum 1996; Kuusela & Huusko 1996; Collier & Smith 1998; Griffith, Barrows & Perry 1998; Petersen et al. 1999a; Briers, Cariss & Gee 2002; Miller, Blinn & Keim 2002). Further, little is known about the impact of dispersal on larval distribution, although a few examples suggest that adult flight and female oviposition can initiate marked patchiness in populations and communities in the water itself (Statzner 1978; Harrison & Hildrew 1998; Peckarsky, Taylor & Caudill 2000).

Features that might affect the movement of adult aquatic insects among river catchments are potentially important targets for river basin management. Although the recovery of rivers affected by past environmental impacts has long been an issue in lotic ecology, the role of recolonization via interbasin dispersal is poorly described (Malmqvist et al. 1991; Bohonak & Jenkins 2003). For instance, increased landscape fragmentation might affect the recovery of stream communities from changes in water quality, such as acidification, through an impact on dispersal between streams (Bradley & Ormerod 2002). In this example, dispersal is particularly important in recovery, as acidification has often affected the whole upper reaches of river systems, thereby removing upstream populations from which recolonization could occur. Moreover, the role of catchment afforestation in acidification means that affected streams are sometimes set in extensive blocks of non-native conifers (Ormerod, Donald & Brown 1989). In the UK, more than 10% of the land area is occupied by such non-native trees, but in upland areas the cover increases substantially and sometimes involves whole subcatchments (Ormerod et al. 1993). Although environmentally sound forest management is increasingly being developed, such land-use patterns might have major ramifications for dispersal and connectivity among catchments (Taylor et al. 1993).

Some research suggests that the riparian vegetation might influence the distribution of adult aquatic insects (Sweeney 1993; Collier, Smith & Baillie 1997; Harrison et al. 1998; Harrison & Harris 2002) but there are few convincing tests of whether the distribution and dispersal of adult aquatic insects differ between different types of land use (Collier & Smith 1995, 1998; Petersen et al. 1999a; Delettre & Morvan 2000; Briers, Cariss & Gee 2002). In densely vegetated landscapes, adult chironomids appeared to be mainly confined to the stream from which they emerged, in contrast to open landscapes where the species assemblage remained similar at different distances from the waterbody (Delettre & Morvan 2000). Petersen et al. (1999a) found that some stonefly species were more numerous in deciduous woodland than on open heath land. Similarly, the abundance and species richness of adult aquatic insects differed among catchments running through different types of forest in New Zealand (Collier & Smith 1995; Collier, Smith & Baillie 1997) and Wales (Briers, Cariss & Gee 2002). However, little is known about what determines the overall distribution of adult aquatic insects, although different factors have been suggested, such as food availability (Harper 1973), predation risk and the likelihood of encountering mates (Stewart 1994). Equally, the complexity of riparian vegetation and variation in physical factors, such as temperature, humidity, wind and light or shade, could affect distribution (Collier & Smith 1995; Petersen et al. 1999a; Delettre & Morvan 2000; Briers, Cariss & Gee 2003).

The present study is the first in which lateral dispersal is compared in replicated catchments with contrasting land use (coniferous forest, cleared forest and open moorland). We investigated whether the dispersal of stoneflies, mayflies and caddisflies differed in catchments with different type of land use and whether the distribution of males and females was similar. Finally, we compared the magnitude of lateral dispersal with that along the stream channel. This information was then used to consider the implications of upland land use on dispersal and connectivity of populations of aquatic insects in streams.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

study sites

The study was carried out on seven tributaries, all draining into the Llyn Brianne reservoir in mid-Wales, UK (latitude 52°8′N, longitude 3°45′W; Fig. 1). In previous research these streams had been designated LI 1–7 (Weatherley, Rutt & Ormerod 1989; Weatherley & Ormerod 1990) and, for comparative purposes, this nomenclature was retained. The streams were all similar in size (mean stream width 1·5–1·8 m) but had catchments of contrasting land use, consisting of coniferous forest (two streams, LI 1 and LI 4), cleared coniferous forest (two streams, LI 2 and LI 3) and open moorland (three streams, LI 5, LI 6 and LI 7). The water chemistry and valley slope also varied among catchments (Table 1). The vegetation at the open moorland sites was dominated by grasses and mosses early in the season (mid-March to mid-June). Thereafter, bracken Pteridium aquilinum (L.) Kuhn dominated, reaching a height of about 1·8 m by the end of the study. At the forested sites there were dense plantations of sitka spruce Picea sitchensis Carriere and lodgepole pine Pinus contorta Douglas ex Loudon and the ground was covered in pine needles and mosses. The cleared catchments were felled in 1996 (LI 2) and 1999 (LI 3) and the ground cover was a mixture of grass tussocks and woody debris. Measurements of stream pH and aluminium were obtained from the Environment Agency (UK).

image

Figure 1. Map of Llyn Brianne, mid-Wales, UK. The numbers indicate the catchments (equivalent to LI 1–7) where transects of Malaise traps were placed.

Download figure to PowerPoint

Table 1.  Seven tributaries draining into the Llyn Brianne reservoir, mid-Wales, UK. Measurements of pH and aluminium were obtained by the Environment Agency (unpublished, public access data)
 CatchmentsLand useCatchment slopeAluminium (mg L−1)1998 annual meanpH
1998 annual mean2001 July
LI 1Nant y BustachForest0·340·395·085·12
LI 2Nant y NannogCleared forest0·240·444·935·53
LI 3Nant y CraflwynCleared forest0·24NDNDND
LI 4Nant CwmbysForest0·200·066·216·51
LI 5(Innominate)Moorland0·450·036·095·63
LI 6Nant y CraflwynMoorland0·350·056·936·97
LI 7Mynydd TrwsnantMoorland0·56NDND6·88

sampling design

Adult stoneflies (Plecoptera), mayflies (Ephemeroptera) and caddisflies (Trichoptera) were caught in Malaise (1937) traps that were placed along a transect, at distances of 0, 15, 30, 45, 60 and 75 m, perpendicular to the stream channel at each of the seven tributaries. The traps were staggered 5 m on each side from the transect in order to reduce interference between them. In addition, one trap was placed directly across each stream 15 m to one side of the transect. The traps were emptied every 8 days in the period 17 March−25 September 2000 and the samples, which contained insects killed in 70% industrial methylated spirit (IMS), brought back to the laboratory for sorting. The insects were identified to species using Hynes (1977) for stoneflies, Elliott & Humpesch (1983) and Harker (1989) for mayflies and Macan (1973) and Malicky (1983) for caddisflies. Males and females were counted separately and subsequently preserved in 70% IMS. Females of the mayflies, and of the caddisflies Wormaldia and Hydropsyche, could not be identified to species. If only one species within each genus was found it was assumed that the females belonged to the same species as the identified males. However, they were identified only to genus if more than one species had been found within that genus. Data on nymphal abundance were obtained by the Catchment Research Group, University of Cardiff (Cardiff, UK), using a standardized 3-min kick-sampling procedure (Weatherley & Ormerod 1987). Samples were taken annually in March–April from 1990 to 2000.

data analyses and statistical methods

Analyses of data on adult mayflies and caddisflies were based on records from 17 March to 25 September 2000. Analysis of data on adult stoneflies was based on results from 17 March to 16 August 2000, which covered the majority of the flight periods for all but a few species.

species diversity and equitability

A measurement of species diversity and equitability was estimated for each site using the Shannon index of diversity (H) and equitability (J). H and J were calculated, respectively, as:

  • image
  • image

where Pi is the proportion of total individuals in the ith species and S is species richness (Begon, Harper & Townsend 1990).

abundance, lateral distribution and land use

Abundance and lateral dispersal were investigated by analysis of the distribution of insects caught in each of the seven transects of Malaise traps placed perpendicular to the stream channels. Data analyses were performed on the most common species, although few species occurred at all catchments. In order to compare the dispersal pattern between catchments of differing land use, it was therefore necessary to pool the data at the level of insect order, and stoneflies, mayflies and caddisflies were examined as separate groups. This might conceal differences in dispersal among the species within orders, and the data should therefore be interpreted with some caution. However, there was no analytical alternative to the strategy adopted and, where single common species were analysed separately, few deviations from the more general pattern were revealed. For caddisflies and mayflies, few individuals were caught at LI 1 and LI 1–4, respectively, and consequently statistical analyses could not be performed on data from these catchments. As the data consisted of counts of insects at increasing distance from the stream, two models for lateral distribution were fitted (the negative exponential, y = ae–bx, and an inverse power function, y = ax–b, where y= numbers caught at distance x from the stream channel) by maximum likelihood using generalized linear modelling (GLM) in genstat for Windows (NAG 1997), assuming a Poisson distribution likelihood with a logarithmic link function but allowing for over-dispersion of counts (Bullock & Clarke 2000; McCullagh & Nelder 1989). By using this method we circumvented the problem of zero counts, which exists in fitting models by least square regression on log- or log–log-transformed data (Bullock & Clarke 2000). A comparison of the mean residual deviance between the models generated on the basis of the exponential and power functions showed that the latter gave a consistently better fit to the observed data, and this model was consequently used.

For each of the common species, and the three insect orders, one GLM was first fitted to all data (model a: constant + distance), the result of which indicated whether the catch declined with distance from the stream channel. The distance from the stream was then used as the covariate in further analysis. By adding terms to the GLM (model b: model a + catchment; model c: model b + distance × catchment), additional effects of differences in abundance between catchments and the interaction between distance and catchment were examined. A significant result of the latter indicated that dispersal differed between the individual catchments. The comparisons were made on the basis of the difference in residual deviance between the models. As the residual difference had a χ2 distribution, it was possible to evaluate whether a more complex model should be preferred to a simpler one (McConway, Jones & Taylor 1999). Examination of the effect of land use was performed in a similar way: the catchment term in the models was replaced with land use and comparisons were made between the three types [model d: model a + land (use); model e: model d + distance × land (use)]. Pairwise comparison was made between the estimated parameters by means of t-statistics (McConway, Jones & Taylor 1999) to find which differed from which. Adjustments for multiple comparisons were made by sequential Bonferroni procedure (Holm 1979), in which the P-values were ranked and the smallest value were tested at the 0·05/c significance level, the next at 0·05/(c − 1), etc., where c equals the number of tests carried out for each model. This procedure provides more power for the individual tests and is recommended in favour of the conventional Bonferroni procedure (Quinn & Keough 2002). The regression coefficients (b) of the fitted models were interpreted as ‘dispersal potential’ (i.e. the rate at which the catch declined with distance from the stream) and the relationship between this and the valley slope of the catchments was tested using Spearman's rank correlation.

Estimates of the distance not exceeded by 50% (i.e. median distance) and 90% of individuals in each taxon within the study area were obtained by integrating the model within the distance at which the traps were placed (x). Hence, the integrals obtained were an estimate of local distribution rather than an absolute estimate of dispersal range.

male and female distribution

Comparison of the sex ratio was performed on the catch in the: (i) Malaise traps placed in the riparian zone (i.e. all traps excluding that placed directly across the stream channel); (ii) trap placed across the stream channel; and (iii) first Malaise trap placed on the stream bank (i.e. that nearest the stream). For the analysis of (i) a GLM was first fitted to all data for each taxonomic group (model a: constant + catchment + distance + catchment × distance). The data were then split according to sex and, by adding terms [model b: model a + (sex); model c: model b + distance × sex], any additional effects of sex on abundance and lateral dispersal were examined. A significant result of the latter would then indicate that dispersal differed between the sexes. For the analyses of (ii) and (iii), the catches of males and females were compared with a paired t-test. The analyses were performed at the level of insect order and for those individual species that occurred in at least three catchments (the minimum number of replicates required for a paired t-test). Only catchments in which the sum of the catches of males and females was more than 10 specimens were included in the analyses. This was done to increase the power of the tests (Sokal & Rohlf 1995). Adjustments for multiple comparisons were again made by sequential Bonferroni procedure (Holm 1979), where c equals number of tests carried out on species within each taxa. The paired t-tests were performed in minitab on logarithmically transformed data.

distribution close to the stream channel

In order to study the direction of movement of adult aquatic insects, catches in the Malaise traps placed directly across the stream channel were compared with the catches in the first trap along each transect, i.e. that positioned directly on the stream bank. The traps across the streams caught insects moving along the stream channel, whereas the traps on the stream bank caught insects moving perpendicularly to the stream. If the prevailing flight direction was along the stream channel, most insects would be caught in the trap placed over the stream. If flight was dominated by insects moving perpendicularly, however, more insects should be trapped on the stream bank. No difference between the catches would indicate a lack of differentiation between longitudinal and lateral flight direction close to the streams. The data were analysed using a paired t-test, with the catchments as replicates. The analyses were performed at both the order and species level and with the same criteria applied as above.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

species composition, richness and diversity

A total of more than 22 500 stoneflies, 4430 caddisflies and 2230 mayflies was caught, the total catch comprising 15 and 40 species of stoneflies and caddisflies, respectively, and eight taxa of mayflies (Table 2). Overall species diversity and equitability were highest in the moorland catchments, but varied between taxa and catchments (Table 3). Only a few species were numerous in all catchments and, in most cases, just a few species dominated the catch. Thus, Amphinemura sulcicollis, Leuctra hippopus and Leuctra nigra were among the five most common stoneflies taken at LI 1–4, where they accounted for 74–84% of all plecopterans. Leuctra inermis and Siphonoperla torrentium dominated the catches at LI 6–7, where they accounted for 52–57% of the stoneflies. At LI 5 the catches were dominated by Nemoura cinerea, although the nymphs of this species are not confined to running water (Hynes 1977) and it is likely that many of the individuals caught in this catchment did not originate in the stream. Indeed, nymphs of Nemoura cinerea were not found in the stream (Table 2). If Nemoura cinerea was excluded from the analysis, the species composition at LI 5 resembled that at LI 6–7 (Table 2).

Table 2.  Total catch of adult stoneflies (Plecoptera), mayflies (Ephemeroptera) and caddisflies (Trichoptera). Bold figures denote species or taxa found in kick samples
Land use CatchmentsForestedClear forestedMoorlandTotal
LI 1LI 4LI 2LI 3LI 5LI 6LI 7
Stoneflies
Amphinemura sulcicollis164 117 369 845 263 318 588 2 664
Brachyptera risi  1   3   1  12  57  76 236   386
Chloroperla tripunctata  0   0   0   0  20  56 197   273
Isoperla grammatica  4   1   8  28 156 235 322   754
Leuctra hippopus 41 119 598 227 128 400 286 1 799
Leuctra inermis  4  8  68 130 55717982365 4 930
Leuctra moselyi  0  64   1   3  78 440 589 1 175
Leuctra nigra 7910422019 953  86 227 100 4 506
Nemoura cambrica  0   0   1   0   0   1   0     2
Nemoura cinerea  9  18 151  691239  63  83 1 632
Nemoura erratica  0  55   0   4   8  13  17    97
Nemurella pictetii 27  49 146 116  54  60 228   680
Protonemura meyeri 18   2 188 303 115 304 374 1 304
Perlodes microcephala  0   0   0   1   2  43  25    71
Siphonoperla torrentium  5  47  29  27 1581145 894 2 305
Sum35215253579271829215179630422 578
Caddisflies
Adicella reducta  0   0   2   2   5  44  22    75
Beraea maurus  0   0   0   0   0  16  30    46
Crunoecia irrorata  0   6   0   0   0   0   0     6
Diplectrona felix  0  75   1   0   2   6  89   173
Drusus annulatus  0   0   0   1  23  71 118   213
Glossosoma conformis  0   0   1   0   0   0  96    97
Halesus radiatus  1   0  10  12  14   2   2    41
Hydropsyche siltalai  1   0   0   0  23  56   5    85
Limnephilus centralis  1   1   3   0  20   1   1    27
Limnephilus luridus  0   0   4   8  94   6   2   114
Limnephilus sparsus  0   8  14   6 137  23  21   209
Micropterna lateralis  4  82  23  21 399  22  31   582
Micropterna sequax  1   5   0   0   0   0   0     6
Odontocerum albicorne  0   1   0   0   0  47  13    61
Philopotamus montanus  0   3   1   2   5 618 374 1 003
Plectrocnemia conspersa 12  25  59 135 395  17  43   686
Plectrocnemia geniculata  1  10   7  20   6   9  23    76
Potamophylax cingulatus  1  27   2  38   8   5   5    86
Rhyacophila dorsalis  1   0   4   0   3 102  78   188
Rhyacophila obliterata  0   0   0   0   1  50  74   125
Silo pallipes  0   0   0   2  13  96 158   269
Stenophylax permistus  3   5   0   2   0   1   1    12
Wormaldia occipitalis  0  45   0   0   2  72  82   201
Others  1   7   4   7  13  15  11    58
Sum 27 300 135 256116312791279 4 439
Mayflies
Baetis sp.  0  38  10   5 306 428 729 1 516
Ecdyonurus sp.  0   0   0   0   0   4   3     7
Ephemerella ignita  0   0   0   1   0  99 166   266
Heptagenia lateralis  0   0   0   0   0 167 159   326
Leptophlebia marginata  0   0   0   0   0   1   1     2
Paraleptophlebia submarginata  0   0   0   0   0   0   2     2
Rhithrogena semicolorata  1   0   0   2   1  44  58   106
Siphlonurus lacustris  0  10   0   0   2   0   0    12
Sum  1  48  10   8 309 7431118 2 237
Total38018733724298243937201870129 254
Table 3.  Species richness, Shannon diversity and equitability index for stoneflies, mayflies and caddisflies and all three taxa pooled
Land useForestedCleared forestMoorland
CatchmentLI 1LI 4LI 2LI 3LI 5LI 6LI 7
Species richnessStoneflies10121213141514
Caddisflies10181619242524
Mayflies 1 2 1 3 3 6 7
All three taxa21322935414645
Diversity (H)Stoneflies 1·56 1·23 1·41 1·69 1·88 2·00 2·06
Caddisflies 1·74 1·95 1·88 1·52 1·74 2·02 2·39
Mayflies 0 0·51 0 0·90 0·04 1·12 0·99
All three taxa 1·84 1·90 1·60 2·00 2·53 2·69 2·75
Equitability (J)Stoneflies 0·68 0·50 0·57 0·66 0·71 0·74 0·78
Caddisflies 0·76 0·67 0·68 0·51 0·55 0·63 0·75
Mayflies 0·74 0·82 0·04 0·62 0·51
All three taxa 0·60 0·56 0·47 0·55 0·68 0·70 0·72

The caddisfly Plectrocnemia conspersa was among the five most common caddisflies species at LI 1–5, whereas Philopotamus montanus dominated the catches at LI 6 and LI 7. Micropterna lateralis was also numerous in several catchments, although larvae of this species are able to live in temporary streams and ditches (Wallace, Wallace & Philipson 1990) and, again, it is likely that many of the adults caught did not originate from the stream. The same is true of three species of the genus Limnephilus, as their larvae are inhabitants of temporary pools or ponds (Wallace, Wallace & Philipson 1990). Of the mayflies, Baetis sp. was most common and found primarily at LI 5–7.

abundance, lateral distribution and land use

The majority of insects in this study were caught either over or close to the channel (Table 4) and, in general, the number of individuals caught declined with distance from the stream (model a: constant + distance; Table 5 and Fig. 2a), although there were occasional exceptions. For instance, a relatively high number of stoneflies was caught in the trap placed 30 m at LI 2 compared with traps placed at the same distance in the other catchments, suggesting an additional source of adults near this point (Fig. 2b). Similarly, species with larval stages not confined to running waters (Nemoura cinerea, Micropterna lateralis and Limnephilus sp.) had different distribution patterns (Table 4 and Fig. 2c). Further analyses on abundance and lateral dispersal were carried out excluding data from LI 2 and the species mentioned above from the remaining catchments.

Table 4.  Spatial distribution of the catch of adult stoneflies (Plecoptera), caddisflies (Trichoptera) and mayflies (Ephemeroptera). St, catch in the Malaise traps placed across the stream channel: Ba, catch in the first Malaise trap placed on the stream bank: 15–75, catch in five Malaise traps placed 15–75 m away from the stream channel. The table includes only species or taxa of which more than 200 specimens were caught
Land use Catchment Trap positionForestedCleared forestMoorland
LI 1LI 4LI 2LI 3LI 5LI 6LI 7
StBa15–75StBa15–75StBa15–75StBa15–75StBa15–75StBa15–75StBa15–75
Stoneflies
Amphinemura sulcicollis125372103 11 3 191 38140459243143147101 15 193 9233 356182 50
Brachyptera risi  1 00  2  0 1   1  0  0  3  9  0 20 26 11  21 47 8 119 84 33
Chloroperla tripunctata  0 00  0  0 0   0  0  0  0  0  0  6 12  2  35 20 1 137 38 22
Isoperla grammatica  4 00  1  0 0   2  3  3 14 14  0 70 78  8 132 96 7 119105 98
Leuctra hippopus 27 95 93 22 4 284178136 93104 30 40 25 63 10720192  28100158
Leuctra inermis  3 01  6  0 2  23  9 36 86 34 10194280 831079627921219695451
Leuctra moselyi  0 00 60  4 0   1  0  0  2  0  1 64 11  3 24617519 523 53 13
Leuctra nigra 44323746230661213681125378446129 42 27 17  5914226  28 42 30
Nemoura cinerea  6 21  7  4 7  16 19116 19 18 32550225464   7  947   6  5 72
Nemurella pictetii 22 50 33 13 3  61 36 49 32 22 62 19 21 14  20 2317  12 79137
Protonemura meyeri 18 00  1  1 0  82 40 66171 91 41 54 33 28 229 5025 293 44 37
Siphonoperla torrentium  5 00 44  3 0  14  8  7  5 21  1100 44 14 73638524 447356 91
Caddisflies
Drusus annulatus  0 00  0  0 0   0  0  0  1  0  0 18  4  1  56 14 1  97 20  1
Limnephilus sparsus  0 00  0  2 6   1  0 13  0  1  5 35 35 67   0  320   0  0 21
Micropterna lateralis  2 11 13  960  11  6  6 12  6  3199134 66   6  610   9  7 15
Philopotamus montanus  0 00  3  0 0   1  0  0  2  0  0  3  0  2 547 71 0 336 36  2
Plectrocnemia conspersa  8 22 11 10 4  41 11  7 88 41  6312 78  5  11  5 1  39  3  1
Rhyacophila dorsalis  1 00  0  0 0   4  0  0  0  0  0  3  0  0  69 33 0  63 14  1
Silo pallipes  0 00  0  0 0   0  0  0  2  0  0  8  4  1  59 35 2 104 45  9
Wormaldia occipitalis  0 00 32  9 4   0  0  0  0  0  0  2  0  0  66  6 0  73  7  2
Mayflies
Baetis sp.  0 00 38  0 0  10  0  0  5  0  0229 68  9 388 3010 683 21 25
Ephemerella ignita  0 00  0  0 0   0  0  0  0  1  0  0  0  0  95  4 0 159  5  2
Heptagenia lateralis  0 00  0  0 0   0  0  0  0  0  0  0  0  0 147 20 0 140  8 11
Table 5.  Abundance, lateral distribution and land use. Results of accumulated analysis of deviance (approximate P-values); significant results in bold. Full summaries of analyses are given in Appendices a–d (see Supplementary material). Model a: constant + distance; model b: model a + catchment; model c: model b + distance × catchment; model d: model a + land use; model e: model d + distance × land use
 Data included from catchments (LI)Effect of catchmentEffect of land use
134567Model aModel bModel cModel dModel e
Stoneflies (Plecoptera)
All stoneflies++++++< 0·001< 0·001   0·004< 0·0010·912
Amphinemura sulcicollis++++++< 0·001< 0·001   0·029< 0·0010·055
Leuctra hippopus++++++< 0·001< 0·001< 0·001< 0·0010·091
Leuctra inermis + +++< 0·001< 0·001   0·001< 0·0010·921
Leuctra nigra++++++< 0·001< 0·001   0·394< 0·0010·978
Nemurella pictetii +++++< 0·001< 0·001   0·282   0·0570·509
Protonemura meyeri + +++< 0·001   0·053   0·446   0·0060·191
Siphonoperla torrentium   +++< 0·001< 0·001   0·109NANA
Caddisflies (Trichoptera)
All caddisflies +++++< 0·001< 0·001   0·068< 0·0010·334
Mayflies (Ephemeroptera)
All females   +++< 0·001   0·213   0·143NANA
All males   +++< 0·001   0·674   0·082NANA
image

Figure 2. Lateral distribution of adult aquatic insects. In general, the numbers caught declined with distance from the stream, exemplified by (a) the distribution of Amphinemura sulcicollis at LI 6. Exceptions to this general pattern of lateral distribution were (b) Amphinemura sulcicollis at LI 2 and (c) Nemoura cinerea (circles) and Micropterna lateralis (triangles) at LI 5.

Download figure to PowerPoint

abundance

Abundance differed significantly between the catchments for both stoneflies and caddisflies (model b; Table 5). For stoneflies, abundance differed between all pairs of catchments except between LI 6 and LI 7, where no difference was found (t(24) = 0·08, P= 0·935). The abundance of caddisflies also differed between most pairs of catchments, except LI 6 and LI 7 (t(20) = 1·27, P= 0·217), LI 3 and LI 4 (t(20) = −0·72, P= 0·477) and LI 3 and LI 5 (t(20) = 2·01, P= 0·058). There was a clear difference in the abundance of mayflies, with very few caught in LI 1–4, leaving 97% of the total catch in LI 5–7 (Table 2). The abundance of male and female mayflies differed (see Male and female distribution) but the abundance did not differ among LI 5–7 within each sex (model b; Table 5).

When examined across land uses, a significant difference was found in the abundance of both stoneflies and caddisflies (model d; Table 5). For stoneflies, significantly fewer individuals were caught in the forested catchments than in the cleared and moorland catchment (forested vs. cleared, t(30) = 3·47, P= 0·002; forested vs. moorland, t(30) = 4·75, P= 0·001) but there was no difference between cleared forest and moorland (t(30) = 1·30, P= 0·203). For caddisflies, more individuals were caught in the moorland catchments than at either of the other two categories (moorland vs. forested, t(24) = 3·81, P= 0·001; moorland vs. cleared forest, t(24)= 3·36, P= 0·003), while there was no difference between forested and cleared catchments (t(24) = 0·62, P= 0·544). The abundance of mayflies was closely related to land use, and only a few individuals were caught outside the moorland catchments (Tables 2 and 4).

For the stoneflies, when examined at the level of species, the abundance resembled that of the whole order with a few exceptions, and differences in total numbers caught were found between many catchments (model b; Table 5). Leuctra nigra was most abundant in the cleared catchments (forested vs. cleared, t(30) = 4·21, P= 0·001; cleared vs. moorland, t(30) = 5·83, P= 0·001), whereas Leuctra inermis was most abundant in the moorland catchments (moorland vs. cleared, t(20) = 2·71, P= 0·013). Only a few individuals of Leuctra inermis were caught in forested catchments. Amphinemura sulcicollis, Leuctra hippopus and Protonemura meyeri were also less abundant in the forested catchments than in the others. Nemurella pictetii was more numerous at LI 7 than at any other catchments and, hence, its abundance was not related simply to land use.

lateral distribution

Examined at the level of order, there was a significant interaction for stoneflies between catchment and distance (model c; Table 5), indicating that lateral distribution differed among catchments. Such differences were found between LI 3 and LI 6 (t(24) = 2·40, P= 0·024) and LI 6 and LI 7 (t(24) = 4·39, P= 0·001), suggesting differences in lateral distribution among these catchments, with a steeper decline in the catches at LI 6 than at the other two (Fig. 3). There was no relationship between the lateral distribution of stoneflies (b in Table 6) and the valley slope of the catchments (Catchment slope in Table 1) (Spearman's rank correlation, stoneflies, n= 6, r= 0·371, P= 0·468). No difference was found between lateral distribution in the remaining catchments, and a common model could therefore be applied to the lateral distribution of stoneflies, excluding LI 6 (Table 6 and Appendix b in Supplementary material).

image

Figure 3. Lateral distribution of stoneflies at LI 6 and 7. Catches in Malaise traps at LI 6 (solid symbols) and LI 7 (open symbols). Models fitted for lateral distribution: the solid line and dashed line represent models for the catch at LI 6 and 7, respectively. See Table 6 for parameters of the models.

Download figure to PowerPoint

Table 6.  Parameter estimates of inverse power functions (y = axb) used as a model for the lateral distribution of adult stoneflies, caddisflies and mayflies. Models were fitted to each taxon based on an analysis of deviance (Appendix b in Supplementary material)
Taxa CatchmentIndividual models for each catchmentsCommon models for each taxon
abab
Stoneflies LI 1  84·6901·084  64·2640·663
LI 31013·3330·716 960·0640·663
LI 4 292·6570·786 260·6040·663
LI 5 669·8140·723 629·5470·663
LI 61864·9700·932  
LI 71846·4130·6001990·2190·663
Caddisflies LI 3    67·2220·973
LI 4    60·3400·973
LI 5   119·8210·973
LI 6   186·9800·973
LI 7   166·3340·973
MayfliesMalesLI 5    37·0400·684
LI 6    37·0400·684
LI 7    37·0400·684
FemalesLI 5    17·9391·123
LI 6    17·9391·123
LI 7    17·9391·123

The lateral distribution of stoneflies was not affected systematically by land use, as indicated by the lack of a significant interaction term between land use and distance from the channel (model e; Table 5). This was the case whether LI 6 was included in the data analysis or not (Appendix d in Supplementary material). When examined at the level of species, lateral distribution differed among catchments for Amphinemura sulcicollis, Leuctra hippopus and Leuctra inermis (model c; Table 5) but this was not related systematically to land use (model e; Table 5). The lateral distribution of the remaining species examined did not differ (models c and e; Table 5).

The lateral distribution of caddisflies was not affected either by catchment (model c; Table 5) or by land use (model e; Table 5). For mayflies, a significant difference was found between the lateral distribution of males and females (see Male and female distribution), but there was no difference in the lateral distribution between LI 5–7 when the sexes were analysed separately (model c; Table 5). The mayflies were numerous only at the moorland catchments, and hence a statistical comparison between land uses was not possible although the difference was obvious.

It was estimated that half of the stoneflies went less than 18 m from the stream, while 90% travelled less than 60 m; the same distances were estimated for the mayfly males. The caddisflies and the mayfly females travelled shorter distances (Fig. 4). Half of the caddisflies, male mayflies and female mayflies travelled less than 9, 17 and 7 m from the stream, respectively. The equivalent distances for 90% of the populations were 50, 59 and 44 m, respectively.

image

Figure 4. Lateral distribution at LI 5 of (a) stoneflies, (b) caddisflies and (c) mayfly males (solid line closed symbols) and females (dashed lines and open symbols). Symbols represent the total catch of insects taken in Malaise traps, and the lines are the models fitted (see Table 6 for parameters of the models). Distances not exceeded by 50% and 90% of the individuals are indicated by arrows.

Download figure to PowerPoint

male and female distribution

The overall sex ratio in stoneflies and caddisflies did not differ from 1 : 1 in the catches in the riparian zone (i.e. all traps excluding only that directly across the stream channel). Thus, there were no differences between the abundance of male and female stoneflies or caddisflies after the overall effects of catchment and distance had been taken into account (model b; Table 7). Neither was there any difference in the dispersal of the two sexes (model c; Table 7). Male mayflies were more abundant than females in the riparian zone (model b; Table 7) and a significant difference was found between the lateral distribution of males and females. While male mayflies were caught in relatively high numbers in the traps at 15 m from the stream channel, only a few females were caught at any distance from the stream.

Table 7.  Distribution of stoneflies, caddisflies and mayflies split up between genders in relation to catchments. Accumulated analysis of deviance: the numbers in the deviance column give the incremental variability (as ‘deviance’) attributable to the factors in the model added sequentially. Approximate P indicates whether the model represents a significant improvement on the previous model (McConway, Jones & Taylor 1999)
Modelsd.f.DevianceMean devianceDeviance ratioApprox. P
Stoneflies
Model a: constant + distance + catchment + distance × catchment1117823·3901620·308176·24< 0·001
Model b: model a + gender 1    2·766   2·766  0·30    0·585
Model c: model b + distance × gender 1   22·866  22·866  2·49    0·120
Residual58  533·232   9·194  
Total7118382·260 258·905  
Caddisflies
Model a: constant + distance + catchment + distance × catchment 9 1820·125 202·236109·20< 0·001
Model b: model a + gender 1    0·518   0·518  0·28    0·599
Model c: model b + distance × gender 1    4·970   4·970  2·68    0·108
Residual48   88·899   1·852  
Total59 1914·512  32·449  
Mayflies
Model a: constant + distance + catchment + distance × catchment 5  405·935  81·187 28·65< 0·001
Model b: model a + gender 1   47·633  47·633 16·81< 0·001
Model c: model b + distance × gender 1   12·438  12·438  4·39    0·045
Residual28   79·358   2·834  
Total35  545·364  15·582  

When analysed at the level of order, female stoneflies comprised the majority in the catches from traps placed across the stream channel, whereas the equivalent catches of mayflies and caddisflies did not differ from a 1 : 1 sex ratio (Table 8 and Fig. 5). No difference in sex ratio was found in the catches in the traps placed on the stream bank for any of the three taxa (Table 8 and Fig. 5). When analysed at species level, a female bias remained in the catches over the stream for some of the most common stoneflies (Amphinemura sulcicollis and Leuctra nigra), even after adjustments for multiple comparisons had been made (Table 8). The sex ratio for the remaining species varied among catchments. Of the caddisflies and mayflies, data analysis was possible at the species level only for Plectrocnemia conspersa, Wormaldia occipitalis and Baetis sp., where no bias in sex ratio was found.

Table 8.  Test of sex ratio in catches from Malaise traps placed across the stream channel (Stream) and those directly on the stream bank (Bank), results of paired t-tests; significant results in bold. An asterisk* indicates significance after a sequential Bonferroni procedure was performed on the tests of individual species of stoneflies, with the stream and bank tested separately (i.e. c= 10); n= sample size (number of catchments included in the analyses); –= test was not possible as too few individuals were caught. See also Fig. 5
TaxaStreamBank
ntP-valuentP-value
Stoneflies
All stoneflies7   5·380·0027   1·420·204
Amphinemura sulcicollis7   5·650·001*7   2·540·044
Brachyptera risi3   1·210·3503   0·690·562
Isoperla grammatica4−0·240·8254−1·390·260
Leuctra hippopus7   2·950·0266   0·540·615
Leuctra inermis5   2·590·0614−1·640·200
Leuctra moselyi4   0·500·6493−3·710·066
Leuctra nigra7   4·600·004*7   2·660·038
Nemurella pictetii6   0·810·4486−0·320·760
Protonemura meyeri6   3·250·0205−3·020·039
Siphonoperla torrentium5   1·440·2204   0·010·996
Caddisflies
All caddisflies7   0·850·4297−0·280·788
Plectrocnemia conspersa6   0·780·4724−0·470·673
Wormaldia occipitalis3   2·200·159
Mayflies
All Mayflies4   0·040·9723−1·640·242
Baetis sp.4   0·050·9633−1·350·309
image

Figure 5. Sex ratio in the catches of aquatic insects taken in Malaise traps across the stream channel (Stream) or in the first trap on the bank (Bank). Black and hatched bars represent females and males, respectively, and vertical bars indicate standard errors: (a) stoneflies, (b) caddisflies and (c) mayflies. See also Table 8 for analyses of sex ratio.

Download figure to PowerPoint

On the stream bank the pattern for single species resembled that of the whole order in that none of the species showed significant deviation from a 1 : 1 sex ratio after adjustment of multiple correction had been made for the species of stoneflies (Table 8).

direction of movement close to the stream channel

When analysed at the level of order, more caddisflies and mayflies of both sexes were caught in the Malaise traps placed across the stream channel than in the first trap on the stream bank (Table 9 and Fig. 5), indicating more flight activity along the channel than laterally. For the stoneflies, only the females were more likely to be taken in flight along the channel than on the stream bank, whereas the males were equally likely to be caught over the stream and on the stream bank (Table 9 and Fig. 5). Analysed at the species level the pattern resembled that of the whole order for mayflies and caddisflies. For the stoneflies only the catches of females of Amphinemura sulcicollis and Protonemura meyeri were significant higher in the traps placed over the stream channel than in those on the bank, after the sequential Bonferroni procedure had been applied (Table 9). For the remaining species examined, no significant difference was found between the catches over the stream and in the first trap on the stream bank.

Table 9.  A comparison of the catch in traps placed across the stream with that in traps placed on the stream bank. A significant result of a paired t-test indicates that more individuals were caught over the stream channel than on the stream bank, and thus were flying along the channel rather than perpendicular to it. An asterisk* indicates significance after a sequential Bonferroni procedure was performed. Female and male stoneflies were tested separately (i.e. c= 9); n= sample size (number of catchments included in the analyses)
Taxa nt-valueP-value
Stoneflies
All stoneflies 7   3·84   0·009
All females 7   6·66   0·001
All males 7   1·71   0·138
Females ofAmphinemura sulcicollis7   5·69   0·001*
Brachyptera risi3−0·34   0·767
Isoperla grammatica3   8·97   0·012
Leuctra hippopus7   1·39   0·214
Leuctra inermis5   4·27   0·013
Leuctra nigra7   1·13   0·300
Nemurella pictetii6   0·72   0·502
Protonemura meyeri6   6·04   0·004*
Siphonoperla torrentium4   2·19   0·116
Males ofAmphinemura sulcicollis7   2·02   0·089
Brachyptera risi3−0·98   0·431
Isoperla grammatica4−1·58   0·213
Leuctra hippopus7−1·36   0·222
Leuctra inermis5   0·28   0·799
Leuctra nigra7−0·32   0·763
Nemurella pictetii6−0·07   0·950
Protonemura meyeri6   1·65   0·174
Siphonoperla torrentium6   0·61   0·569
Caddisflies
All caddisflies 7   9·54< 0·001
All females 7   9·85< 0·001
All males 7   7·78< 0·001
Females ofPlectrocnemia conspersa5   4·60   0·010
Wormaldia occipitalis3   4·96   0·038
Males ofPlectrocnemia conspersa4   3·38   0·043
Wormaldia occipitalis3   4·18   0·053
Mayflies
All mayflies 4   5·26   0·013
All females 4   4·67   0·019
All males 4   5·34   0·013
Females ofBaetis sp.4   4·40   0·022
Males ofBaetis sp.4   5·45   0·012

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

abundance and species composition

Variation in species composition and abundance among catchments was associated with land use, with the highest species diversity at the moorland sites together with the highest abundance of caddisflies and mayflies. The lowest abundance of stoneflies was in the forested catchments. Studies of larval abundance and taxon richness from catchments in the same area revealed a similar distribution pattern and coincided with the physicochemical environment in the streams, with the highest pH and lowest aluminium concentration measured in the moorland streams (Weatherley, Rutt & Ormerod 1989; Weatherley & Ormerod 1990; Bradley & Ormerod 2002). Perhaps not surprisingly, this strongly suggests that the total number of adults, as well as the species composition, is affected by factors that determine the benthic (larval) community. Little is known about how much the size of larval population and species composition is determined by the factors influencing the adults (Zwick 1990; Enders & Wagner 1996). Peckarsky, Taylor & Caudill (2000) and Caudill (2003b) demonstrated that recruitment of eggs was limited by regional rather than local supply of adult mayflies, suggesting a movement of adult females. Further investigations are required to establish the links between larval and adult numbers, and models including both adult and larval stages will be useful tools for the analysis of population dynamics (Wilbur 1980; Speirs et al. 2000; Bohonak & Jenkins 2003; Caudill 2003b).

lateral distribution

No difference was found between catchments in the lateral distribution of caddisflies and within the sexes of mayflies, although there was a difference in the lateral distribution of stoneflies between catchments, with a steeper decline in the catches with distance in LI 6 than in LI 3 and LI 7. The distribution did not correlate with the slope of the catchment nor with land use. Similar abundances were found in LI 6 and LI 7 and therefore the difference in lateral distribution between these sites cannot be explained simply by abundance.

Overall, most adult insects stayed close to the stream channel and half of the stoneflies travelled less than 18 m, while 90% travelled less than 60 m. The local dispersal range of the caddisflies and female mayflies was even shorter, with half the individuals caught within 7–11 m of the stream channel. It is possible that the dispersal range of caddisflies and mayflies in this study is underestimated compared with that of the stoneflies as their abundance was much lower, and the chance of catching caddisflies and mayflies further from the stream channel was therefore lower. The general trend is that most adult aquatic insects stay close to the stream channel from which they emerged (Svensson 1974; Sode & Wiberg-Larsen 1993; Kuusela & Huusko 1996; Collier & Smith 1998). Hence, our estimates of the dispersal range of stoneflies, mayflies and caddisflies are similar to those derived previously (Jackson & Resh 1989b; Griffith, Barrows & Perry 1998; Petersen et al. 1999a; Delettre & Morvan 2000; Briers, Cariss & Gee 2002). Note, however, that we followed Bullock & Clarke (2000) by building GLM assuming a Poisson distribution likelihood but allowing for overdispersion of counts (McCullagh & Nelder 1989, chapter 6), which in effect is assuming that the variance was not constant across the sites. This makes comparison among our models potentially difficult.

effect of land use on dispersal

The fact that most adult aquatic insects stay close to the stream channel may also explain why there was little distinction between lateral dispersal in the different land uses. However, it is still too early to conclude that land use has no effect on lateral dispersal. For instance, adult abundance was much lower in the forested catchments than in the other catchments. Therefore, although there was no difference in the decline of the catch with distance (b in the models) between the land uses, the chances are that fewer individuals will reach substantial distances from the channel in the forested catchments than in the cleared or moorland ones (Fig. 6). Further, there could be an undetected effect of land use on long-distance dispersal, which may be important in the recovery of streams from acidification, pollution, physical changes and disturbance (Sode & Wiberg-Larsen 1993; Wilcock, Hildrew & Nichols 2001). Such rare dispersal events, although difficult to detect using conventional trapping techniques, could be sensitive to coniferous afforestation. Recent evidence based on genetic markers, for instance, suggests that populations of Plectrocnemia conspersa (the most abundant caddisflies at Llyn Brianne) are genetically homogeneous over distances of up to 20 km or more, inferring frequent dispersal over such distances (Wilcock, Nichols & Hildrew 2003).

image

Figure 6. Comparison of absolute dispersal distances between populations with low (dashed lines) and high (solid lines) population densities. Given that two populations have the same dispersal potential, i.e. equal slope of the lines (b in the models in Table 6), individuals from populations with a low abundance will have less chance of reaching suitable habitats at further distances than individuals from larger populations.

Download figure to PowerPoint

male and female distribution

The overall sex ratio in the riparian zone did not differ from 1 : 1 for caddisflies and stoneflies. While only a few female mayflies were caught at any distance from the stream, a relatively high number of males was caught in the traps placed 15 m from the stream channel. The adult life span of mayflies is relatively short (Brittain 1987). It is possible that females are time limited in finding suitable oviposition sites and are, therefore, more concentrated over the stream channel and less likely than males to be caught in the riparian zone. It should be emphasized that most mayflies were caught over the stream channel, however, and this was the case for both males and females. Only for the stoneflies was there a significant deviation from a 1 : 1 sex ratio over the stream channel, with more females than males. The same pattern was observed by Petersen et al. (1999a), where the females of several species were in the majority over the stream channel towards the end of the flight period. The search for egg-laying sites, combined with a longer female life span, may provide a plausible explanation for the female-biased catches over the stream channel (Sheldon & Jewett 1967; Hynes 1974; Petersen 2002). Thus, differences between the distributions of the sexes and of the taxa may be explained by differences in mating and reproductive behaviour.

direction of movement close to the stream channel

More individual mayflies and caddisflies were caught in traps placed across the stream channels than in the trap on the banks, indicating that most flight activity of these taxa was concentrated over the stream channel itself. Of the stoneflies, some of the females were also more likely to be taken in flight along the channel than on the stream bank, whereas there was no difference for the males. The reason for this difference in distribution between the sexes in stoneflies may be found in their mating and oviposition behaviour. As mating for many species of stoneflies occurs on the ground shortly after emergence (Brinck 1949), it is possible that male stoneflies have an equal chance of encountering receptive females on the stream bank as on emergent substrata in the stream. Females, once they are mated and their eggs matured, may be governed by the search for oviposition sites, and hence may have a higher flight activity along the stream channel.

routes of dispersal

The vast majority of the insects caught in this study was taken either over the stream channel or in the first trap on the stream bank (Table 4). This suggests that most flight activity is in the ‘stream corridor’ itself rather than laterally away from the stream channel. Many empirical and a few theoretical studies have dealt with migration along the stream channel, inspired by Müller's (1954) postulated ‘colonization cycle’. Empirical studies have not been able to provide evidence for universal upstream flight (Allan 1995; Petersen et al. 1999b; Kopp, Jeschke & Gabriel 2001) and theoretical studies have raised questions about its ‘necessity’, in terms of population persistence (Anholt 1995; Kopp, Jeschke & Gabriel 2001; Speirs & Gurney 2001). Computer simulations suggest that the exact compensation of stream drift by upstream movement is an evolutionarily stable strategy (Kopp, Jeschke & Gabriel 2001). Exact compensation by upstream flight seems rather unlikely under natural conditions, however, and random dispersal by adults may be enough to maintain the population (Speirs & Gurney 2001). Some studies of genetic population structure suggest that movement of aquatic invertebrates within streams is limited, resulting in significant genetic differentiation at the reach scale, and that larvae at the reach scale are likely to be the offspring of only a few matings (Bunn & Hughes 1997). In contrast, Peckarsky, Taylor & Caudill (2000) explained local egg recruitment by inter-reach dispersal by females of Baetis, and Hershey et al. (1993) found that between one-third and one-half of an adult Baetis population flew 1·6–1·9 km upstream from where it emerged. Similarly, swarms of caddisflies marked with a radioisotope were found up to 16 km upstream of a reactor effluent (Coutant 1982) and dispersal among ponds by adult females of the mayfly Callibaetis ferrugineus hageni was demonstrated by means of stable isotope techniques (Caudill 2003b). Apart from the studies by Coutant (1982) and Hershey et al. (1993), little is known about actual adult dispersal distances of aquatic insects along the stream channel and more information is required (Bohonak & Jenkins 2003). Future research should therefore be aimed at quantifying the extent and distance of dispersal within the stream corridor in both upstream and downstream directions. The research should relate dispersal within the stream corridor to land use, to clarify whether the various types of riparian vegetation act as a barrier or corridor for dispersal (Delettre & Morvan 2000). Focus should be brought on landscape connectivity within the stream corridor in relation to population dynamics (Taylor et al. 1993).

implications for management

Throughout much of the UK and Europe, the interaction between forest management, riparian zones and river ecosystems has been increasingly emphasized in forest policy (Forestry Commission 2000). Indeed, in many locations, active and sympathetic forest management offers one of the best mechanisms for enhancing upland stream and river quality. With much forest in the UK reaching the end of its first major rotation after initial planting during the 1950s and 1960s, there are now opportunities to re-examine forest design with particular emphasis on forest–water linkages. As in other areas of stream management and ecology, however, the requirements of adult insects have so far received little emphasis in such design issues.

These results confirm the importance of the riparian zone, i.e. the area close to the stream channel (e.g. 20 m on either side), as a buffer zone. Here, care is required over aspects of forest management such as habitat structure and the application of pesticides. This study extends previous work, by emphasizing the value of buffer zones intrinsically in the protection of aquatic insects, the adult stages of which are concentrated in this zone, but also for the riparian insectivores for which aquatic insects provide an important food source (Jackson & Fisher 1986; Collier, Bury & Gibbs 2002). Further, we recommend that attention should be paid to the role of landscape structure, catchment vegetation and interbasin connectivity in the life of aquatic insects, the dispersal of which may determine how aquatic habitats are linked. As most flight activity appears to occur along the stream corridors, the dispersal of adults could well be dependent on the vegetation structure of the stream and river network in facilitating linkage directly across headwaters.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

We thank Tilhill Forestry, Forest Enterprises and Mr Roger Davies for access and permission to use Malaise traps on their land, and Dr Fred Slater for the use of facilities at the Cardiff University Field Station. We are grateful to Professor Malcolm Elliott and Dr Roger Wotton for advice and to two anonymous referees for helpful comments. Finally we thank all the people who helped with the field work. The study was supported by a Draper's Scholarship, Queen Mary, University of London and the Danish Research Academy for I. Petersen, and by Ray Beverton and Freshwater Biological Association studentships for Z. Masters.

Supplementary material

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Appendix a. Lateral distribution of stoneflies in relation to catchments, examined at species level.

Appendix b. Lateral distribution of stoneflies, caddisflies amd mayflies in relation to catchments, examined at order level.

Appendix c. Lateral distribution of stoneflies in relation to land use, examined at species level.

Appendix d. Lateral distribution of stoneflies and caddisflies in relation to land use, examined at order level.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information
  • Allan, J.D. (1995) Stream Ecology: Structure and Function of Running Waters. Chapman & Hall, London, UK.
  • Anholt, B.R. (1995) Density dependence resolves the stream drift paradox. Ecology, 76, 22352239.
  • Begon, M., Harper, J.H. & Townsend, C. (1990) Ecology: Individuals, Populations and Communities. Blackwell Science, Cambridge, UK.
  • Bohonak, A.J. & Jenkins, D.G. (2003) Ecological and evolutionary significance of dispersal by freshwater invertebrates. Ecology Letters, 6, 783796.
  • Bradley, D.C. & Ormerod, S.J. (2002) Long-term effects of catchment liming on invertebrates in upland streams. Freshwater Biology, 47, 161171.
  • Briers, R.A., Cariss, H.M. & Gee, J.H.R. (2002) Dispersal of adult stoneflies (Plecoptera) from upland streams draining catchments with contrasting land-use. Archiv für Hydrobiologie, 155, 627644.
  • Briers, R.A., Cariss, H.M. & Gee, J.H.R. (2003) Flight activity of adult stoneflies in relation to weather. Ecological Entomology, 28, 3140.
  • Brinck, P. (1949) Studies on Swedish stoneflies (Plecoptera). Opuscula Entomologica, Supplement XI, 1250.
  • Brittain, J.E. (1987) Life history strategies in Ephemeroptera and Plecoptera. Mayflies and Stoneflies: Life Histories and Biology. Proceedings of the 5th International Ephemeroptera Conference and the 9th International Plecoptera Conference (ed. I.C.Campbell), pp. 112. Kluwer, Dordrecht, the Netherlands.
  • Brittain, J.E. & Eikeland, T.J. (1988) Invertebrate drift: a review. Hydrobiologia, 166, 7793.
  • Bullock, J.M. & Clarke, R.T. (2000) Long distance seed dispersal by wind: measuring and modelling the tail of the curve. Oecologia, 124, 506521.
  • Bunn, S.E. & Hughes, J.M. (1997) Dispersal and recruitment in streams: evidence from genetic studies. Journal of the North American Benthological Society, 16, 338346.
  • Caudill, C.C. (2003a) Empirical evidence for nonselective recruitment and a source–sink dynamic in a mayfly metapopulation. Ecology, 84, 21192132.
  • Caudill, C.C. (2003b) Measuring dispersal in a metapopulation using stable isotope enrichment: high rates of sex-biased dispersal between patches in a mayfly metapopulation. Oikos, 101, 624630.
  • Collier, K.J. & Smith, B.J. (1995) Sticky trapping of adult mayflies, stoneflies and caddisflies along three contrasting streams near Hamilton, New Zealand. New Zealand Natural Sciences, 22, 19.
  • Collier, K.J. & Smith, B.J. (1998) Dispersal of adult caddisflies (Trichoptera) into forests alongside three New Zealand streams. Hydrobiologia, 361, 5365.
  • Collier, K.J., Bury, S. & Gibbs, M. (2002) A stable isotope study of linkages between stream and terrestrial food webs through spider predation. Freshwater Biology, 47, 16511659.
  • Collier, K.J., Smith, B.J. & Baillie, B.R. (1997) Summer light-trap catches of adult Trichoptera in hill-country catchments of contrasting land use, Waikato, New Zealand. New Zealand Journal of Marine and Freshwater Research, 31, 623634.
  • Coutant, C.C. (1982) Evidence for upstream dispersion of adult caddisflies (Trichoptera: Hydropsychidae) in the Columbia River. Aquatic Insects, 4, 6166.
  • Delettre, Y.R. & Morvan, N. (2000) Dispersal of adult aquatic Chironomidae (Diptera) in agricultural landscapes. Freshwater Biology, 44, 399411.
  • Elliott, J.M. (2002a) A continuous study of the total drift of freshwater shrimps, Gammarus pulex, in a small stony stream in the English Lake District. Freshwater Biology, 47, 7586.
  • Elliott, J.M. (2002b) Time spent in the drift by downstream-dispersing invertebrates in a Lake District stream. Freshwater Biology, 47, 97106.
  • Elliott, J.M. & Humpesch, U.H. (1983) A Key to the Adults of the British Ephemeroptera with Notes on their Ecology. Scientific Publication No. 47. Freshwater Biological Association, Ambleside, UK.
  • Enders, G. & Wagner, R. (1996) Mortality of Apatania fimbriata (Insecta: Trichoptera) during embryonic, larval and adult life stages. Freshwater Biology, 36, 93104.
  • Forestry Commission (2000) Forest and Water Guidelines. Forestry Commission, Wetherby, UK.
  • NAG (1997) GENSTAT for Windows, Version 5, Release. 4·1. Command Language Manual. Numerical Algorithms Group, Oxford, UK.
  • Griffith, M.B., Barrows, E.M. & Perry, S.A. (1998) Lateral dispersal of adult aquatic insects (Plecoptera, Trichoptera) following emergence from headwater streams in forested Appalachian catchments. Annals of the Entomological Society of America, 91, 195201.
  • Hanski, I. (2001) Population dynamic consequences of dispersal in local populations and in metapopulations. Dispersal (eds J.Clobert, E.Danchin, A.A.Dhondt & J.D.Nichols), pp. 283299. Oxford University Press, Oxford, UK.
  • Harker, J. (1989) Mayflies. Richmond Publishing Co., Slough, UK.
  • Harper, P.P. (1973) Life histories of Nemouridae and Leuctridae in Southern Ontario (Plecoptera). Hydrobiologia, 41, 309356.
  • Harrison, S.S.C. & Harris, I.T. (2002) The effects of bankside management on chalk stream invertebrate communities. Freshwater Biology, 47, 22332245.
  • Harrison, S.S.C. & Hildrew, A.G. (1998) Distribution dynamics of epilithic insects in a lake littoral. Archiv für Hydrobiologie, 143, 275293.
  • Harrison, S.S.C., Harris, I.T., Croeze, A. & Wiggers, R. (1998) The influence of bankside vegetation on the distribution of aquatic insects. Verhandlungen der Internationale Vereinigung für Theorestische und Angewandte Limnologie, 27, 14801484.
  • Hershey, A.E., Pastor, J., Peterson, B.J. & Kling, G.W. (1993) Stable isotopes resolve the drift paradox for Baetis mayflies in an Arctic River. Ecology, 74, 23152325.
  • Holm, S. (1979) A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6, 6570.
  • Hynes, H.B.N. (1974) Observations on adults and eggs of Australian Plecoptera. Australian Journal of Zoology, 29, 3752.
  • Hynes, H.B.N. (1977) A Key to the Adults and Nymphs of British Stoneflies (Plecoptera). Scientific Publication No. 17. Freshwater Biological Association, Ambleside, UK.
  • Jackson, J.K. & Fisher, S.G. (1986) Secondary production, emergence, and export of aquatic insects of a Sonoran Desert stream. Ecology, 67, 629638.
  • Jackson, J.K. & Resh, V.H. (1989a) Distribution and abundance of adult aquatic insects in the forest adjacent to a northern California stream. Environmental Entomology, 18, 278281.
  • Jackson, J.K. & Resh, V.H. (1989b) Activities and Ecological Role of Adult Aquatic Insects in the Riparian Zone of Streams. General Technical Report PSW 110. USDA Forest Service, Berkeley, CA, USA.
  • Jackson, J.K., McElravy, E.P. & Resh, V.H. (1999) Long-term movements of self-marked caddisfly larvae (Trichoptera: Sericostomatidae) in a California coastal mountain stream. Freshwater Biology, 42, 525536.
  • Kopp, M., Jeschke, J.M. & Gabriel, W. (2001) Exact compensation of stream drift as an evolutionarily stable strategy. Oikos, 92, 522530.
  • Kovats, Z.E., Ciborowski, J.H. & Corkum, L.D. (1996) Inland dispersal of adult aquatic insects. Freshwater Biology, 36, 265276.
  • Kuusela, K. & Huusko, A. (1996) Post-emergence migration of stoneflies (Plecoptera) into the nearby forest. Ecological Entomology, 21, 171177.
  • Macan, T.T. (1973) A Key to the Adults of the British Trichoptera. Scientific Publication No. 28. Freshwater Biological Association, Ambleside, UK.
  • McConway, K.J., Jones, M.C. & Taylor, P.C. (1999) Statistical Modelling Using GENSTAT. Arnold, London, UK.
  • McCullagh, P. & Nelder, J.A. (1989) Generalized Linear Models, 2nd edn. Chapman & Hall, London, UK.
  • McCullough, D.R. (1996) Metapopulation management: what patch are we in and which corridor should we take? Metapopulations and Wildlife Conservation (ed. D.R.McCullough), pp. 405410. Island Press, Washington, DC.
  • Mackay, R.J. (1992) Colonization by lotic macroinvertebrates: a review of processes and patterns. Canadian Journal of Fisheries and Aquatic Sciences, 49, 617628.
  • Malaise, R. (1937) A new insect trap. Entomologisk Tidskrift, 38, 148160.
  • Malicky, H. (1983) Atlas of European Trichoptera. Dr W. Junk Publishers, The Hague, the Netherlands.
  • Malmqvist, B., Rundle, S., Bronmark, C. & Erlandsson, A. (1991) Invertebrate colonization of a new, man-made stream in southern Sweden. Freshwater Biology, 26, 307324.
  • Miller, M.P., Blinn, D.W. & Keim, P. (2002) Correlation between observed dispersal capabilities and patterns of genetic differentiation in populations of four aquatic insect species from the Arizona White Mountains, USA. Freshwater Biology, 47, 16601673.
  • Müller, K. (1954) Investigations on the organic drift in north Swedish streams. Report of the Institute of Freshwater Research Drottingholm, 35, 133148.
  • Ormerod, S.J., Donald, A.P. & Brown, S.J. (1989) The influence of plantation forestry on the pH and aluminum concentration of upland welsh streams: a reexamination. Environmental Pollution, 62, 4762.
  • Ormerod, S.J., Rundle, S.D., Lloyd, E.C. & Douglas, A.A. (1993) The influence of riparian management on the habitat structure and macroinvertebrate communities of upland streams draining plantation forests. Journal of Applied Ecology, 30, 1324.
  • Palmer, M.A., Allan, J.D. & Butman, C.A. (1996) Dispersal as a regional process affecting the local dynamics of marine and stream benthic invertebrates. Trends in Ecology and Evolution, 11, 322326.
  • Peckarsky, B.L., Taylor, B.W. & Caudill, C.C. (2000) Hydrologic and behavioural constraints on oviposition of stream insects: implications for adult dispersal. Oecologia, 125, 186200.
  • Petersen, I. (2002) Spatial and temporal distribution of adult aquatic insects. PhD Thesis. University of London, London, UK.
  • Petersen, I., Winterbottom, J.H., Orton, S., Friberg, N., Hildrew, A.G., Speirs, D.C. & Gurney, W.S.C. (1999a) Emergence and lateral dispersal of adult Plecoptera and Trichoptera from Broadstone Stream, UK. Freshwater Biology, 42, 401416.
  • Petersen, I., Winterbottom, J.H., Orton, S. & Hildrew, A.G. (1999b) Does the colonization cycle exist? Nordic Benthological Meeting (eds N.Friberg & J.D.Carl), p. 142. Technical Report No. 266. NERI, Silkeborg, Denmark.
  • Quinn, G.P. & Keough, M.J. (2002) Experimental Design and Data Analysis for Biologists. Cambridge University Press, Cambridge, UK.
  • Rundle, S.D., Foggo, A., Choiseul, V. & Bilton, D.T. (2002) Are distribution patterns linked to dispersal mechanism? An investigation using pond invertebrate assemblages. Freshwater Biology, 47, 15711581.
  • Sheldon, A.L. & Jewett, S.G. (1967) Stonefly emergence in a Sierra Nevada Stream (Plecoptera). Pan-Pacific Entomologist, 43, 18.
  • Sode, A. & Wiberg-Larsen, P. (1993) Dispersal of adult Trichoptera at a Danish forest brook. Freshwater Biology, 30, 439446.
  • Söderström, O. (1987) Upstream movements of invertebrates in running waters: a review. Archiv für Hydrobiologie, 111, 197208.
  • Sokal, R.R. & Rohlf, F.J. (1995) Biometry. W.H. Freeman, New York, NY.
  • Speirs, D.C. & Gurney, W.S.C. (2001) Population persistence in rivers and estuaries. Ecology, 82, 12191237.
  • Speirs, D.C., Winterbottom, J.H., Gurney, W.S.C. & Hildrew, A.G. (2000) Long-term demographic balance in the Broadstone Stream insect community. Journal of Animal Ecology, 69, 4558.
  • Statzner, B. (1978) The effects of flight behaviour on the larval abundance of Trichoptera in the Schierenseebrooks (North Germany). Proceedings of the 2nd International Symposium on Trichoptera, 1977 (ed. M.Crichton), pp. 121134. Dr W. Junk Publishers, The Hague, the Netherlands.
  • Stewart, K.W. (1994) Theoretical considerations of mate finding and other adult behaviors of Plecoptera. Aquatic Insects, 16, 95104.
  • Svensson, B.W. (1974) Population movement of adult Trichoptera at a south Swedish stream. Oikos, 25, 157175.
  • Sweeney, B.W. (1993) Effects of streamside vegetation on macroinvertebrate communities of White Clay Creek in Eastern North America. Proceedings of the Academy of Natural Sciences of Philadelphia, 144, 291340.
  • Taylor, P.D., Fahrig, L., Henein, K. & Merriam, G. (1993) Connectivity is a vital element of landscape structure. Oikos, 68, 571573.
  • Wallace, I.D., Wallace, B. & Philipson, G.N. (1990) A Key to the Case-Bearing Caddis Larvae of Britain and Ireland. Scientific Publication No. 51. Freshwater Biological Association, Ambleside, UK.
  • Waters, T.F. (1972) The drift of stream insects. Annual Review of Entomology, 17, 253272.
  • Weatherley, N.S. & Ormerod, S.J. (1987) The impact of acidification on macroinvertebrate assemblages in Welsh streams: towards an empirical model. Environmental Pollution, 46, 223240.
  • Weatherley, N.S. & Ormerod, S.J. (1990) The constancy of invertebrate assemblages in soft-water streams: implications for the prediction and detection of environmental change. Journal of Applied Ecology, 27, 952964.
  • Weatherley, N.S., Rutt, G.P. & Ormerod, S.J. (1989) Densities of benthic macroinvertebrates in upland Welsh streams of different acidity and land use. Archiv für Hydrobiologie, 115, 417431.
  • Wilbur, H.M. (1980) Complex life cycles. Annual Review of Ecology and Systematics, 11, 6793.
  • Wilcock, H.R., Hildrew, A.G. & Nichols, R.A. (2001) Genetic differentiation of a European caddisfly: past and present gene flow among fragmented larval habitats. Molecular Ecology, 10, 18211834.
  • Wilcock, H.R., Nichols, R.A. & Hildrew, A.G. (2003) Genetic population structure and neighbourhood population size estimates of the caddisfly Plectrocnemia conspersa. Freshwater Biology, 48, 18131824.
  • Zwick, P. (1990) Emergence, maturation and upstream oviposition flights of Plecoptera from the Breitenbach, with notes on the adult phase as a possible control of stream insect populations. Hydrobiologia, 194, 207223.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Supplementary material
  9. References
  10. Supporting Information

Appendix S1. Lateral distribution of stoneflies in relation to catchments, examined at species level. Accumulated analysis of deviance: the numbers in the deviance column give the incremental variability (as ?deviance?) attributable to the factors in the model added sequentially. Approximate P indicates whether the model represents a significant improvement on the previous model (McConway, Jones & Taylor 1999). Appendix S2. Lateral distribution of stoneflies, caddisflies and mayflies in relation to catchments, examined at order level. Accumulated analysis of deviance: the numbers in the deviance column give the incremental variability (as ?deviance?) attributable to the factors in the model added sequentially. Approximate P indicates whether the model represents a significant improvement on the previous model (McConway, Jones & Taylor 1999). Appendix S3. Lateral distribution of stoneflies in relation to land use, examined at species level. Accumulated analysis of deviance: the numbers in the deviance column give the incremental variability (as ?deviance?) attributable to the factors in the model added sequentially. Approximate P indicates whether the model represents a significant improvement on the previous model (McConway, Jones & Taylor 1999). Appendix S4. Lateral distribution of stoneflies and caddisflies in relation to land use, examined at order level. Accumulated analysis of deviance: the numbers in the deviance column give the incremental variability (as ?deviance?) attributable to the factors in the model added sequentially. Approximate P indicates whether the model represents a significant improvement on the previous model (McConway, Jones & Taylor 1999).

FilenameFormatSizeDescription
JPE_942_sm_App1.doc14KSupporting info item
JPE_942_sm_App2.doc13KSupporting info item
JPE_942_sm_App3.doc13KSupporting info item
JPE_942_sm_App4.doc12KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.