The distribution of Ephemerella ignita (Ephemeroptera) in streams: the role of pH and food resources

SUMMARY. 1. The mayfly Ephemerella ignita does not occur in upland streams of the River Duddon catchment which have waters with low pH values of 4.8-5.2, negative alkalinities and low ionic contents. However, it does occur in lowland streams of the catchment which have waters with higher pH values of 6.6 and above, large positive alkalinities and high ionic contents. This study addresses the question of whether the absence of the animal in the acid water streams is due directly to the water chemistry or to the absence of suitable food.
2. In comparative experiments using unfed animals of several species it was shown that E. ignita has a considerable tolerance of low pH, low ion water. In this it is similar to Amphinemura sulcicollis (Plecoptera), a stonefly which occurs in the acid water streams of the River Duddon catchment.
3. Growth rates in laboratory experiments with E. ignita were equally good whether the food supplied was that available in the low pH water streams (bryophyte Nardia compressa plus filamentous alga Hormidium subtile) or typical of that available in high pH water streams (moss Rhynchostegium riparioides with the epiphytic diatom Cocconeis placentula). However, a consideration of the natural habits of the animal suggests that it is the inaccessibility of the H. subtile algal food, rather than its unsuitability, which is relevant to the exclusion of E. ignita from the low pH streams. In addition, the very limited diversity of other algae in the low pH water streams, in the summer season, may well be influential in inhibiting the establishment of E. ignita.
4. The comparative experiments using unfed animals suggested that, contrary to the findings for E. ignita , the mayflies Baetis muticus and Beatis rhodani are probably excluded from the acid water streams of the Upper Duddon because of water chemistry.     

Intensive sampling and transplantation experiments reveal continued effects of episodic acidification on sensitive stream invertebrates

1. Although mean pH is increasing in acidified streams throughout Europe, benthic invertebrates are recovering only slowly. One developing hypothesis is that acid episodes continue to restrict recolonisation and recovery. Here, we used intra‐annual sampling in replicate circumneutral, acid and episodic streams at the Llyn Brianne experimental catchments to assess possible mayfly exposure to episodic effects quantified from a risk index based on long‐term discharge. Episodic effects were then tested using short‐term transplantations of animals in enclosures. 2. No mayflies occurred in acid streams while Baetis rhodani occurred in episodic streams but declined to zero density following low pH in autumn. Ephemerella ignita, Baetis vernus, B. muticus, Rhithrogena semicolorata, Ecdyonurus spp. and Heptagenia lateralis occurred only in the circumneutral streams. The first two species were present as nymphs only during July and August so that episodic exposure risk was minimal, but the remaining species occurred as nymphs in all months. 3. Baetis rhodani transplanted during base‐flow (September 2003) and high‐flow (April 2004) were exposed to either; (i) continual (chronic) exposure in the acid streams over 16 days or (ii) repeated short‐term (episodic) exposure to acid conditions for 2 × 4 day periods interspersed with 4‐day recovery periods in the circumneutral streams. Baetis survival in the circumneutral streams (always pH > 5.7) remained high during both low‐ and high‐flow. By comparison, mortality increased (P < 0.01) during chronic and episodic exposure, but only during high‐flow (mean pH 3.8–3.9, cf. 5.5–5.8 at low flow) when mortality varied significantly in the order chronic (>80%) > episodic (>40%) >circumneutral (<10%). 4. We conclude that, despite Europe‐wide trends towards chemical recovery from acidification, even short exposures to high‐flow events at Llyn Brianne are still sufficiently acid to reduce the survival of B. rhodani. Most mayfly species absent from acid and episodic streams have life cycles that would increase acid exposure risks during autumn and winter, and this may be sufficient to explain their current distribution.     

Mortality of the terrestrial adult and aquatic nymphal life stages of Baetis vernus and Baetis rhodani in the Breitenbach, Germany (Insecta: Ephemeroptera)

1. The mortality of Baetis vernus Curtis and Baetis rhodani Pictet during the terrestrial-aerial and aquatic life stages was studied at the Breitenbach near Schlitz, Hesse, Germany. The number of females emerging from the stream was recorded with emergence traps. To estimate mortality of females of both species during terrestrial life, numbers of emerging females were compared with numbers of females returning to the stream for oviposition, as shown by numbers of egg masses found in the stream. 2. Mortality of female B. vernus during their terrestrial life stage was 98.8%. It was 91.2 and 96.6%, respectively, during the first and second generations of B. rhodani. 3. To estimate the mortality of both sexes of B. rhodani during the aquatic life stage, the number of eggs laid by the first generation was compared with the number of adults emerging during the second generation. 4. Mortality during the aquatic phase (both sexes combined) of B. rhodani was 91.1%. 5. Mortality during the terrestrial life stages was at least as high, if not higher than during the aquatic stages. Evidently, there is a considerable export of organic material into the terrestrial environment around the stream. Mortality during terrestrial life may be an important regulator of population size.     

Geographical and seasonal evidence of cryptic diversity in the <Emphasis Type="Italic">Baetis rhodani</Emphasis> complex (Ephemeroptera,Baetidae) revealed by means of DNA taxonomy

Previous phylogenetic investigations on the mayfly Baetis rhodani Pictet from several European countries, excluding Italy, strongly suggested the presence of cryptic species. Our paper reports a DNA-taxonomy phylogenetic analysis of B. rhodani with additional populations coming from Italian and UK sites, and aims to identify potential cryptic species with a coalescent-based method (GMYC model) and to understand the mechanisms of local coexistence of cryptic species. Twenty-five haplotypes of Italian samples and five haplotypes of UK samples were identified and added to a large European dataset. A total of 11 potential cryptic species have been recognised, and three of them co-occured in one Italian area. Such cryptic species seem to be phylogenetically over-dispersed on the tree and temporally segregated, and the seasonal substitution pattern of cryptic species could explain the apparently widespread distribution of the B. rhodani complex and its ability to adapt to different temperatures and food resources, justifying some of the differences observed in the relationship between water temperature, growth rates and phenology documented from field studies.     

Egg envelopes of Baetis rhodani and Cloeon dipterum (Ephemeroptera, Baetidae): a comparative analysis between an oviparous and an ovoviviparous species

A comparative ultrastructural investigation on the eggshell (vitelline and chorionic envelopes) has been carried out in the nymphs of two mayfly species encompassed into the Baetidae, namely Baetis rhodani and Cloeon dipterum. During oocyte differentiation in the meroistic telotrophic ovarioles, gametes are connected to discrete nurse cells by trophic cords. In B. rhodani, each ovariole contains several eggs arranged in sequence, whereas in C. dipterum each contains a single egg. Follicle cells are competent for vitelline and chorionic envelope synthesis. Baetis rhodani is an oviparous species and the chorion is fairly thick, formed by an alveolate endochorion and a fibrillar exochorion delimited by a honey‐comb roof. Cloeon dipterum stands out among Ephemeroptera for being ovoviviparous. In B. rhodani, ovulation starts in the older nymphs with dark wing‐pads, whereas in C. dipterum choriogenesis ends in the imaginal stage. Here the chorion is very thin and laid eggs hatch almost immediately, allowing the larvulae to move out. The maturation of a single egg per ovariole is synchronized with the achievement of the adult stage. The acquisition of ovoviviparity has led to remarkable changes in the ovariole organization along with a simplification of the eggshell structure.     

Microsatellite loci for the mayfly Baetis rhodani (Baetidae,Ephemeroptera)

The first microsatellite primers are described for the mayfly family Baetidae (order Ephemeroptera). Seven polymorphic loci were isolated and characterized for the species Baetis rhodani from two enriched genomic libraries. A total of 183 individuals were genotyped using these loci. Allelic diversity was high for all loci, and observed heterozygosities ranged between 0.382 and 0.772. A heterozygote deficiency was detected in some loci, suggesting the presence of ‘null’ alleles.     

Life cycles and growth rates of Baetis spp. (Ephemeroptera: Baetidae) in the laboratory and in two stony streams in Austria

SUMMARY. Larvae of Baetis alpinus, B. lutheri and B. rhodani were reared in a stream channel (water temperature range 4.2–11.4°C) in the laboratory. The larval growth was exponential and the mean specific growth rate varied from 1.93 to 2.24% day^?1 for B. alpinus, 1.49 to 3.41% day^?1 for B. lutheri and 0.79 to 3.11% day^?1 for B. rhodani. These variations in growth rate were related to variations in mean temperature and this was the major factor affecting growth in the laboratory. Non-quantitative samples of the benthos in the Seebach and Unterseebach, two stony streams near Lunz, Austria, were taken at approximately monthly intervals from November 1965 to May 1968. In each year, there were two winter and three summer cohorts for B. alpinus from Seebach and two or three winter and one to three summer cohorts for B. lutheri and B. rhodani from Unterseebach. Over the study period of 30 months, eleven cohorts were recorded for B. alpinus and B. lutheri, and ten cohorts for B. rhodani. The life cycle of a cohort varied from 3 to 8 months in B. alpinus, from 2.5 to 9 months in B. lutheri and from 2.5 to 8 months in B. rhodani. Mean specific growth rate in length varied from 0.82 to 2.97% day^?1 for B. alpinus, 0.96 to 3.33% day^?1 for B. lutheri and 0.65 to 3.01% day^?1 for B. rhodani. The percentage of the variability in growth rate accounted for by variations in mean temperature was 63% for B. alpinus, 91% for B. lutheri and 82% for B. rhodani. Therefore mean temperature was clearly the major factor affecting the growth rates in the field. An agreement was found between the growth rates of B. alpinus in the field and the laboratory. The growth rates of B. lutheri and B. rhodani were slower in the field than in the laboratory at higher temperatures. The possible reasons for this latter discrepancy are discussed, and the growth rates of the three Baetis spp. are compared with those of other species of Ephemeroptera.     

Diel vertical movements by lotic mayfly nymphs under variable predation risk

1. The diel foraging periodicities of two grazing mayfly (Ephemeroptera) nymphs, Heptagenia dalecarlica and Baetis rhodani, under variable fish (European minnow) predation risk were examined in a series of laboratory experiments. 2. Heptagenia dalecarlica were almost exclusively nocturnal in their use of feeding areas on stone tops. There was a sharp increase in the proportion of nymphs out of refuge at nightfall, both in the control and fish-odour treatments. In a treatment with freely moving fish, H. dalecarlica decreased their activity during both the day and night. In feeding trials with three freely foraging minnows, nymphs were completely safe when they had access to refuges beneath stones, whereas almost half the nymphs were consumed when no refuges were available. 3. Baetis rhodani nymphs reduced their use of stone tops when exposed to four caged minnows or a freely moving fish, but this occurred both day and night. In feeding trials, B. rhodani were captured only while in the water column, and their mortality risk was independent of refuge availability. 4. It is suggested that because H. dalecarlica lack efficient post-encounter defences, they must rely on pre-encounter mechanisms to reduce the threat of fish predation. It appears that in order to dwell sympatrically with fish, H. dalecarlica have evolved a coexistence by hiding strategy. In contrast, Baetis are vulnerable to fish attacks only if they enter drift in daylight, and are thus able to remain on stone tops both during the day and at night.     

Influence of riparian cover on stream temperatures and the growth of the mayfly <Emphasis Type="Italic">Baetis rhodani</Emphasis> in an upland stream

This research aimed to assess whether stream temperature and the size of Baetis rhodani larvae (Insecta: Ephemeroptera) differed between sites with contrasting riparian vegetation cover. Stream temperatures and Baetis sizes were monitored at sites along a 5-km section of a small (30 km²) upland Scottish stream. Some sites had riparian tree cover dominated by mature broadleaved woodland, while others were located in areas of open Calluna moorland. The 13 month data period included two contrasting summers, one typical for the region and the other unusually warm. Daytime water temperatures were generally highest at the open moorland sites. The difference between moorland and broadleaved woodland sites was most apparent in the summer months, with the woodland shading apparently acting to reduce maximum temperatures and mediate temporal fluctuations. Size data suggest that Baetis has a bivoltine life-cycle in the stream. Significant differences in the mean size of Baetis were found between moorland and woodland sites, with differences occurring primarily during the summer months. Mean size of Baetis was significantly greater in the moorland during the time of summer adult emergence. During the months that followed (growth of the next generation), Baetis larvae in the woodland sites were significantly larger. Size distribution data for the population suggest earlier adult emergence in the moorland section and different egg-to-larvae hatch times. Overall, the results indicate that subtle differences in growth and life cycles, linked to differences in riparian cover and streamwater temperature, can occur over short distances in upland streams.     

Low‐level effects of inert sediments on temperate stream invertebrates

1. The delivery, entrainment and deposition of inert fine sediments are among the most significant contributors to stream and river impairment worldwide. Associated ecological effects have been observed frequently, but specific experiments to identify sensitivity and avoidance behaviour in stream organisms are few, particularly in headwaters. 2. In a field‐experiment, we added fine sand at low levels (c. 4–5 kg m^?2) to 10 m reaches of two replicate headwater streams in the Usk catchment (Wales, U.K.) over two periods (autumn and summer). Upstream reaches were used as control in a classic before‐after‐control‐impact design. Invertebrate drift and benthic composition were measured for 2 days before and 1 day after sediment impact. 3. Sediment addition significantly increased overall drift density (by 45%) and propensity (by 200%), with effects largest on the night following addition rather than immediately (i.e. within 9 h). The mayflies Baetis rhodani, B. muticus and Ecdyonurus spp., simuliid and chironomid dipterans, and helodid beetles were the strongest contributors. 4. There were no marked effects on benthic composition, but density declined in treated reaches by 30–60%, particularly in B. rhodani, Ecdyonurus spp. and Leuctra hippopus + L. moselyi. 5. All effects were consistent between both seasons and streams. 6. These data show how even low‐level, short‐term, increases in fine sediment loading to upland, stony streams can reduce overall benthic density through increased drift. We suggest that the likely cause of the delayed drift response was a change in habitat quality which prompted avoidance behaviour. Longer‐term experiments are required to assess whether these effects reduce fitness or explain the losses of some types of organisms observed recently in sediment‐impaired reaches of this and other catchments.     

Stream drift as a consequence of disturbance by invertebrate predators

Summary We carried out an experimental field study in a Swedish stream in order to determine whether mobile predators enhance the drift of stream insects. We increased the density of nymphs of the predaceous perlid stonefly, Dinocras cephalotes, in an experimental section of a stream up to densities in another more densely populated part of the same stream. The drift of several benthic species increased significantly compared to a control section where D. cephalotes were rare. Experiments carried out in September showed a strongly elevated drift response in nymphs of the mayfly Baetis rhodani only, whereas May experiments resulted in increased drift in B. rhodani as well as the amphipod Gammarus pulex, the stonefly Leuctra fusca, chironomids, and the total number of drifting animals. In September, we found that the drift response of Baetis rhodani to predator disturbance was dependent on the size of mayfly nymphs; small nymphs appeared in greater numbers in the drift nets than did large nymphs. A subsequent laboratory analysis of drift lengths of B. rhodani nymphs supported the hypothesis that small nymphs travel in the drift for longer than do large nymphs, particularly in darkness. We suggest that morphological constraints in vision or swimming performance, or both, cause small nymphs to drift longer. In May, size-dependent drift was less obvious, probably because the size of the nymphs was considerably greater than in September.     

Classification and ordination of macroinvertebrate assemblages to predict stream acidity in upland Wales

Macroinvertebrates were collected from riffles at 104 sites in upland Wales during April and July 1984. Species assemblages were ordinated by DECORANA, classified by TWINSPAN and related to stream chemistry and other environmental factors using correlation and multiple discriminant analysis. DECORANA axis 1 was most strongly correlated with pH and aluminium concentration whilst axis 2 correlated with stream gradient and flow. Four TWINSPAN site groups established in each season were also principally related to pH and aluminium concentration, and reflected overall taxon-richness; differences between groups were most apparent during spring, when catchment forest cover and taxon-richness were also related. A dichotomous key based on indicator species was established for each season with the coleopteran Hydraena gracilis Germar and the Ephemeroptera, including Baetis rhodani Pictet, important indicators at Level 1. We propose that these indicator systems may be used for the rapid detection and assessment of acid waters throughout Wales, and that the methodology is applicable generally.     

Short-term experimental acidification of a Welsh stream: comparing the biological effects of hydrogen ions and aluminium

SUMMARY. 1. A soft-water stream iti upland Wales was dosed with sulphuric acid and aluminiutn sulphate at two successive points to create sitnultaneous episodes of low pH, and low pH with increased aluminiutn. Chemical atid biological responses were measured before, during and after the episode and were compared with a reference zone. 2. The pH fell frotn ～7.0 to 4.28 (±0.18 SD) and 5.02 (±0.10) respectively in the acid and aluminium zones. Corresponding aluminium concentrations during the episode were 0.052 g Al m^?3 (±0.008) and 0.347 g Al nr³ (±0.047), the former not differing significantly from the reference zone. The concentration of cadmium rose to 0.002- 0.011 g Cd m^?3in both treated areas, but the concentrations of other metals were unchanged. 3. In situ toxicity tests were performed with macroinvertebrates and fish. Chironomus riparius. Hydropsyche angustipennis and Dinocras cephalotes suffered no mortality. Ecdyonurus venosus, Baetis rhodani and Gammarus pulex showed up to 25% mortality in both treatment zones and further mortalities occurred after the episode. Brown trout Salmo trutta and salmon Valmo salar s howed 7–10% mortality in the acid zone, but 50–87% in the aluminium zone, where salmon had a significantly shorter LT₅₀than trout. 4. The drift of Simuliidae increased during treatment in both acid and aluminium zones. Drift densities of Dixa puherula, Protonemura meyeri, Ephemeralla ignita and Dicranota sp. increased in the aluminium zone. The most pronounced response was by Baetis rhodani in the aluminium zone where drift density increased by ×8.4 during the episode. 5. Baetis rhodani was the only taxon to show a significant decline in benthic density during the treatment, and then only in the aluminium zone. Drift could account for most of the losses. 6. The depth distribution of invertebrates in the substratum differed between zones following treatment. More individuals were present at the surface of the reference zone (1287 m^?2±747) than at the surface of the other zones (<400 m^?2); however, densities at greater depths were similar. These patterns probably reflected differences prior to the treatments.     

Longitudinal extent of acidification effects of plantation forest on benthic macroinvertebrate communities in soft water streams: evidence for localised impact and temporal ecological recovery

Plantation conifer forest can increase the risk of acidification in acid-sensitive catchments with consequences for macroinvertebrates and ecosystem functioning. This study compared headwater streams in forested and non-forested catchments to appraise the distance required for the acid effect on macroinvertebrates to diminish downstream. Strict criteria were followed in the selection of paired streams, including similarities in elevation, aspect, stream order, geographical proximity, geology and soil type, with no inflowing tributaries in the first 2.5 km and no major land-use other than plantation forest and moorland. Consequently two headwater streams (one forested, one non-forested) drained Ordovician sedimentary geology and two headwater streams (one forested, one non-forested) drained Old Red Sandstone (ORS) were selected. All streams drained peaty soils. Up to six sites at 500 m intervals were sampled in triplicate by multihabitat kick sampling. Ecological impact and recovery from acid effects involved the entire macroinvertebrate community, but varied between seasons. Acid-sensitive Ephemeroptera revealed marked effects of episodic acidification, with Baetis rhodani and Rhithrogena semicolorata found to be transient between seasons. The increase in acid-sensitive ephemeropteran species with increasing distance downstream indicates the finite effects of forest mediated acidification on soft water streams in Ireland. Ecological impact appears, therefore, to be localised within the catchment, with ecological recovery occurring a short distance downstream. This is the first study to have reported such results.     

Life history and larval feeding of some species of Ephemeroptera and Plecoptera (Insecta) in the Sierra Nevada (Southern Iberian Peninsula)

We monitored two high mountain streams from the Sierra Nevada (Southern Spain) monthly throughout an annual cycle, measuring and analyzing physical and chemical parameters. Temperature was hourly registered. We studied life cycles and larval feeding of four mayfly species [Serratella ignita (Poda, 1761), Serratella spinosa nevadensis (Alba-Tercedor, 1981), Alainites muticus (Linnaeus, 1758), and Baetis alpinus Pictet, 1843] and two stonefly species [Amphinemura triangularis (Ris, 1902) and Capnioneura mitis Despax, 1932]. Some species changed their strategies (growth rate, life cycle duration, and/or voltinism) in accordance with different conditions (depending on the altitude and temperature). The gut content data showed that species fed mainly on detritus (as collector-gatherers or as shredders), except B. alpinus that behaved as a scraper in one of the streams. Differences in the larval diet of the same species in different streams, and even through its life, support the idea of not using general models to extrapolate the function of a species from a higher taxonomical level. Handling editor: D. Dudgeon     

Long-term effects of catchment liming on invertebrates in upland streams

1. Catchment liming to mitigate acidification causes major chemical change in freshwaters but longer‐term effects are poorly understood. Using a replicated basin‐scale experiment with a multiple BACI design (= before‐after‐control‐impact), we assessed chemical and biological effects for 10 years after the catchments of three acidified Welsh streams at Llyn Brianne were limed in 1987/88. 2. Stream chemistry was measured weekly to monthly, and macroinvertebrates monitored annually, between 1985 and 1998. Biological change through time was assessed from the abundance and taxon richness of invertebrates. We paid particular attention to 18 species known to be acid‐sensitive. The effects of liming were assessed by comparing chemical and biological trends among the three replicate limed streams, three acid reference streams and two naturally circumneutral streams. 3. Following single lime applications, acid‐base chemistry in treated streams changed significantly. High mean pH (> 6), increased calcium (> 2.5 mg L^?1) and low aluminium (< 0.1 mg L^?1) persisted throughout the 10 years following liming. 4. The effects of liming on invertebrates were modest. Acid sensitive taxa increased significantly in abundance in limed streams, but only during 2 years following treatment. Significant effects on richness were more sustained, but on average added only 2–3 acid‐sensitive species to the treated streams, roughly one‐third of their average richness in adjacent circumneutral streams. Only the mayfly Baetis rhodani and the stonefly Brachyptera risi occurred significantly more often in limed streams after treatment than before it. 5. Despite these modest long‐term effects on invertebrates, nearly 80% of the total pool of acid‐sensitive species has occurred at least once in the limed streams in the 10 years since treatment. This pattern of occurrence suggests that the colonization of limed streams by acid‐sensitive taxa reflects limited persistence rather than restricted dispersal. We present evidence to show that episodes of low pH continued to affect acid‐sensitive taxa even after liming. We highlight the importance of extending the time‐periods over which the effects of large‐scale ecological experiments are assessed.     

Cryptic diversity and multiple origins of the widespread mayfly species group Baetis rhodani (Ephemeroptera: Baetidae) on northwestern Mediterranean islands

How the often highly endemic biodiversity of islands originated has been debated for decades, and it remains a fervid research ground. Here, using mitochondrial and nuclear gene sequence analyses, we investigate the diversity, phylogenetic relationships, and evolutionary history of the mayfly Baetis gr. rhodani on the three largest northwestern Mediterranean islands (Sardinia, Corsica, Elba). We identify three distinct, largely co‐distributed, and deeply differentiated lineages, with divergences tentatively dated back to the Eocene–Oligocene transition. Bayesian population structure analyses reveal a lack of gene exchange between them, even at sites where they are syntopic, indicating that these lineages belong to three putative species. Their phylogenetic relationships with continental relatives, together with the dating estimates, support a role for three processes contributing to this diversity: (1) vicariance, primed by microplate disjunction and oceanic transgression; (2) dispersal from the continent; and (3) speciation within the island group. Thus, our results do not point toward a prevailing role for any of the previously invoked processes. Rather, they suggest that a variety of processes equally contributed to shape the diverse and endemic biota of this group of islands.     

Effects of Rotenone Treatment on Lotic Invertebrates

Two rivers in Norway were treated with rotenone in an effort to eradicate the salmon parasite, Gyrodactylus salaris. Rotenone induced an immediate catastrophic invertebrate drift, mainly of dead individuals. The response to rotenone with respect to degree and timing varied widely. Baetis rhodani, Isoperla sp. and Rhyacophila nubila displayed a strong increase in drift densities early in the treatment. Baetis niger, Ameletus inopinatus, Heptagenia sp. and Agapetus ochripes showed a delayed response. Ephemerella aurivillii, E. mucronata, Taeniopteryx nebulosa and Elmis aenea drifted at low densities throughout the treatment. Early larval instar tended to be more sensitive to rotenone than later instar, and the toxic effect of rotenone tended to increase with increasing water temperature. (© 2011 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Selective oviposition of the mayfly Baetis bicaudatus

Selective oviposition can have important consequences for recruitment limitation and population dynamics of organisms with complex life cycles. Temporal and spatial variation in oviposition may be driven by environmental or behavioral constraints. The goals of this study were to: (1) develop an empirical model of the substrate characteristics that best explain observed patterns of oviposition by Baetis bicaudatus (Ephemeroptera), whose females lay eggs under rocks protruding from high-elevation streams in western Colorado; and (2) test experimentally selective oviposition of mayfly females. We surveyed the number and physical characteristics of potential oviposition sites, and counted the number and density of egg masses in different streams of one watershed throughout two consecutive flight seasons. Results of surveys showed that variability in the proportion of protruding rocks with egg masses and the density of egg masses per rock were explained primarily by seasonal and annual variation in hydrology, and variation in geomorphology among streams. Moreover, surveys and experiments showed that females preferred to oviposit under relatively large rocks located in places with high splash associated with fast current, which may provide visual, mechanical or both cues to females. Experiments also showed that high densities of egg masses under certain rocks were caused by rock characteristics rather than behavioral aggregation of ovipositing females. While aggregations of egg masses provided no survival advantage, rocks selected by females had lower probabilities of desiccating during egg incubation. Our data suggest that even when protruding rocks are abundant, not all rocks are used as oviposition sites by females, due to female selectivity and to differences in rock availability within seasons, years, or streams depending on variation in climate and hydrogeomorphology. Therefore, specialized oviposition behavior combined with variation in availability of quality oviposition substrata has the potential to limit recruitment of this species.Electronic supplementary material Supplementary material is available in the online version of this article at and is accessible for authorized users.     

Flight Behaviour of Merolimnic Insects from the Leutra River (Thuringia,Germany)

1. The flight behaviour of adult merolimnic insects was studied on the first order limestone stream Leutra (Thuringia, Germany) using sticky traps. The focus of the present study was on testing the colonisation cycle hypothesis of Müller (1954) and on the small-scale dispersal of adult merolimnic insects. 2. A high number of sticky traps was used to guarantee a proper statistical analysis of the data and exclude effects of the heterogeneous environment on flight behaviour. 3. The flight behaviour of Leuctra hippopus, Nemoura cambrica, Nemoura flexuosa, Protonemura nitida, Protonemura praecox (Plecoptera), and Baetis rhodani (Ephemeroptera) was studied. 4. Specific flight corridors were observed for P. praecox, Nemoura spp., L. hippopus and B. rhodani. 5. Comparison of the catches on sticky traps set perpendicular to the water showed no significant flight upstream in any of the taxa studied. Thus, the results do not support the hypothesis of the colonisation cycle.