Comparative study of mathematical models for the relationship between water temperature and brood development time of Gammarus fossarum and G. roeseli (Crustacea: Amphipoda)

SUMMARY. 1. The relationship between water temperature ( T°C ) and brood development time (d days for embryonic development time plus 'post-hatch time') was investigated experimentally using animals from four populations of Gammarus fossarum (populations 1–4) and two populations of G. roeseli (populations 5 and 6) in Austrian streams.
2. Eleven mathematical models were examined as suitable functions to describe the relationship between d and T . The models were compared with respect to their goodness-of-fit and suitability for further quantitative, statistical analyses of intra- and interspecific differences between natural populations of G. fossarum and G. roeseli .
3. Statistical analyses were performed with logarithmically transformed values of d , in order to provide homogeneous variances between the different constant temperatures used. The following three-parameter model was found to be the most suitable for the data:
This equation was fitted in its linear form:      

Life history and reproductive capacity of Gammarus fossarum and G. roeseli (Crustacea: Amphipoda) under naturally fluctuating water temperatures: a simulation study

SUMMARY 1. Mathematical functions developed in long‐term laboratory experiments at different constant temperatures were combined with daily water temperatures for 1991–93 in eight Austrian streams and rivers to simulate the complex life histories and reproductive capacities of two freshwater amphipods: Gammarus fossarum and G. roeseli. The functions describe brood development times, hatching success, times taken to reach sexual maturity, growth, and fecundity. The sex ratio was assumed to be 0.5 and an autumn–winter reproductive resting period was based on observations of six river populations. Simulations included summer‐cold mountain streams, summer‐warm lowland rivers, watercourses fed by groundwater or influenced by heated effluents, and varying amplitudes of change within each year. 2. A fortran 77 computer program calculated growth from birth to sexual maturity of first‐generation females born on the first day of each calendar month in 1991, and the numbers of offspring successfully released from the maternal broodpouch in successive broods. At the 1991–93 regimes of temperature, individual G. fossarum released 127–208 offspring and G. roeseli released 120–169 in seven or eight successive broods during life spans of less than 2 years in six rivers. Life spans extended into a third year in the relatively cool River Salzach (mean temperature 7.5 °C). They were not completed in the very cold River Steyr (mean 5.6, range 2.5–7.9 °C), where G. fossarum produced five broods (totalling 120 offspring) and G. roeseli only two broods (totalling 28 offspring) in the 3‐year period. Except in the Steyr, some offspring grew rapidly to maturity and produced several second‐generation broods during the simulation period; in the warmest rivers some third‐generation broods were also produced. Birth dates, early or late in the year, influenced the subsequent production of broods and young, depending on temperature regimes in particular rivers. Total numbers of offspring produced by the second and third generations represent the theoretical reproductive capacities of G. fossarum and G. roeseli. Minimum and maximum estimates mostly ranged from 100 to 17 300, were larger for G. fossarum except in the warmest river (March), where temperatures rose above 20 °C for 56–78 days in summer, and largest (maximum 37 600) in the River Voeckla heated by discharge from a power‐station (mean 11.5 °C). Results from the simulations agree with preliminary assessments of relative abundances for G. fossarum and G. roeseli in several of the study rivers, but in some one or both species appear to be absent. On a wider scale, the present study confirms that G. fossarum is potentially more successful than G. roeseli in cool rivers but indicates that neither species is likely to maintain viable populations in cold rivers strongly influenced by snow and ice‐melt. 3. The potential impacts of future river warming by increases of 1, 2 and 3 °C, due to climate change, vary according to river site, date of fertilisation, the extent of temperature increase, and the species of Gammarus. For Austrian rivers with mean temperatures in the range c. 7–10 °C, future warming would result in modest changes in the life histories and reproductive capacities of both G. fossarum and G. roeseli; the former would find improved temperature conditions in watercourses that are currently very cold throughout the year, and both would find warm rivers less tolerable. 4. The high potential reproductive capacity of gammarids, with rapid production of numerous successive broods when sexual maturity is finally achieved, indicates adaptation to high mortality during the relatively long period of growth to sexual maturity, and provides scope for an opportunistic strategy of emigration from centres of population abundance to colonise new territory when conditions are favourable. Rapid expansion of populations is desirable to combat the effects of environmental catastrophes, both frequent and short‐term floods and droughts, and more long‐term climatic changes that have occurred several times in glacial–interglacial periods during the current Ice Age.     

Environmental calcium and mitochondrial DNA polymorphism among local populations of Gammarus fossarum (Crustacea, Amphipoda)

Mitochondrial DNA sequence analysis was used to investigate genetic diversity and phylogeographic population structure among geographically isolated alpine populations of the amphipod Gammarus fossarum. This study was performed across a region with an important glacial history and substantial geographical variations in calcium levels, in order to evaluate the relative impact on the genetic diversity of the population history and this environmental factor known to be associated with local adaptation and population growth. Analysis of the nucleotide sequences for a 376-bp segment of the cytochrome c oxidase I gene in 84 specimens from seven different populations revealed 20 different haplotypes distributed into five major lineages that are not geographically structured. Analysis of molecular variance indicates that the populations from the crystalline massif are less diverse than those living on limestone, suggesting a global correlation between genetic diversity and calcium concentration of the water. However, the role of this environmental factor appears to be indirect, with smaller population size leading to lower genetic diversity.     

Molecular evidence for further overlooked species within the Gammarus fossarum complex (Crustacea: Amphipoda)

The freshwater amphipod Gammarus fossarum Koch, in Panzer, 1836 is a locally abundant keystone species mainly occuring in European headwaters but also in larger rivers. Genetic studies in the past 25 years have revealed three cryptic species within nominal G. fossarum (types A, B and C). Assignments of specimens to these types were based on allozyme and 16S markers. Today, a fragment of the cytochrome c oxidase subunit 1 (CO1) is primarily used as a genetic marker for species assignments (‘DNA Barcoding’), yet not a single CO1 sequence of G. fossarum is available in the Barcode of Life Database. We analysed new CO1 and 16S data for German, Hungarian and Croatian G. fossarum specimens and compared them with 16S and CO1 sequences of G. fossarum from GenBank. Thereby, we close the gap between traditional allozyme- and 16S-based species assignments and modern CO1 barcoding. Studying genetic variation in 55 specimens from 29 populations, we identified between 11 and 23 novel and genetically distinct clades using distance- and tree-based methods. Our results suggest that G. fossarum comprises several additional, yet unrecognised, species in particular from the Balkan region. Therefore, a taxonomic revision and biogeographic reconsideration of the G. fossarum complex is urgently needed.     

Isolation and characterization of the first eight microsatellite loci in Gammarus fossarum (Crustacea, Amphipoda) and cross-amplification in Gammarus pulex and Gammarus orinos

We characterized the first microsatellite markers for Gammarus fossarum. Eight loci gave satisfactory amplification patterns in two stream populations (Southern France) with number of alleles ranging from 2 to 10 and expected heterozygosity from 0.076 to 0.857. We performed cross-amplification in two closely related gammarid species, Gammarus pulex and Gammarus orinos. Among the eight tested microsatellite loci, four correctly amplified in G. pulex and three in G. orinos.     

Oxygen dependent habitat selection in surface and hyporheic environments by Gammarus roeseli Gervais (Crustacea, Amphipoda): experimental evidence

Field distributions of benthic and hyporheic invertebrates are dynamic and are influenced by many physical and chemical factors. A laboratory flume containing natural gravel substrates was used to test the hypothesis that the amphipod Gammarus roeseli Gervais actively selects habitat based on two important environmental variables, dissolved oxygen concentration and direction of water flow. Under homogeneous oxygen concentrations throughout the flume, amphipods accumulated downstream. During trials with uniformly hypoxic conditions throughout the flume, G. roeseli moved to stones and screens at the water surface and above, into the zone normally saturated with humidity. This behavior, termed aquatic surface respiration, may enhance survival during periods of oxygen deficiency. Oxygen gradients were created in the flume by injecting water of differing oxygen concentrations into the head and the center of the channel. A statistically significant response to these gradients by G. roeseli demonstrates active selection of regions with more favorable oxygen concentrations. This revised version was published online in July 2006 with corrections to the Cover Date.     

Mitochondrial DNA polymorphism among alpine populations of Gammarus lacustris (Crustacea, Amphipoda)

1. Mitochondrial DNA sequence analysis was used to investigate the phylogeographic structure among populations of the amphipod Gammarus lacustris isolated in high altitude lakes from the Northern to the Southern Alps, and to investigate the historical and demographic events that may have accompanied this differentiation.
2. The comparison of a 376‐base segment of the mitochondrial cytochrome c oxidase I gene in fifty‐four specimens from nine different populations reveals no variation within populations. However, northern populations appear divergent from those from the Southern Alps. A particular population, located at the limit between the Northern and Southern Alps shows an intermediate haplotype related to the southern lineage.
3. Consideration of the different mitochondrial lineages, together with geographical differences among populations, suggests a scenario for the post‐glacial colonization of G. lacustris in the alpine lakes.     

Parasite altered micro-distribution of Gammarus pulex (Crustacea: Amphipoda)

In a river survey, Gammarus pulex amphipods both unparasitised and parasitised with the acanthocephalan Echinorhynchus truttae were distributed similarly with respect to flow regimen, tending to be more abundant in faster, shallower, riffle patches. However, there was a higher prevalence of parasitism in faster, shallower areas than in slower, deeper areas and abundance correlated with macrophyte coverage for unparasitised but not parasitised amphipods, indicating subtle differences in habitat usage. A laboratory 'patch' simulation indicated that parasitism influenced micro-distribution. There were higher proportions of unparasitised amphipods in/under stone substrates and within weed. In contrast, there were higher proportions of parasitised amphipods in the water column and at the water surface. As the experiment progressed, unparasitised but not parasitised amphipod habitat usage shifted from those micro-habitats above the substrate and in the water column to those in/under the substrates. Experiments also demonstrated that parasitised amphipods were more active and had a greater preference for illumination. Previous studies of the effects of acanthocephalan parasitism of amphipod hosts have focussed on how drift behaviour is altered, now we show that subtle differences in micro-habitat usage could translate to greatly increased vulnerability to fish predation. We discuss how aggregation of parasitised individuals within specific habitats could promote parasite transmission.     

The dynamics of predation on Gammarus spp. (Crustacea: Amphipoda)

Gammarus spp. (Crustacea: Amphipoda) are widespread throughout a diverse range of marine, freshwater and estuarine/brackish habitats, often dominating benthic macroinvertebrate communities in terms of both numbers and/or biomass. Gammarus spp. are the dominant macroinvertebrate prey items of many fish, whether as a seasonal food source or a year-round staple. Selective predation by fish on Gammarus spp. is often linked to parasitism and the body size of the prey. Gammarus spp. populations are under increasing threat from both pollution and replacement/displacement by introduced species. Loss of populations and species invasions/replacements could have significant impacts on native predator species if the predator(s) cannot successfully adapt their feeding patterns to cope with non-indigenous Gammarus prey species. Despite this, many fish predation studies do not identify Gammarus prey to species level. This lack of precision could be important, as Gammarus spp. exhibit wide variations in physiochemical tolerances, habitat requirements, abilities to invade and susceptibility to replacement. Although rarely acknowledged, the impacts of nonpiscean predators (particularly macroinvertebrates) on Gammarus prey species may frequently be stronger than those exerted by fish. A major aim of this review is to ascertain the current importance of Gammarus as a prey species, such that the implications of changes in Gammarus spp. populations can be more accurately assessed by interested groups such as ecologists and fisheries managers. We also review the dynamics of Gammarus spp. as prey to a diverse array of mammals, birds, amphibians, insects, flatworms, other crustaceans such as crabs and crayfish and, perhaps most importantly, other Gammarus spp.     

The nature and implications of differential predation between Gammarus pulex and G. duebeni celticus (Crustacea: Amphipoda)

The mutual predatory behaviour between Gammarus pulex (Linnaeus, 1758) and Gammarus duebeni celticus (Stock & Pinkster, 1970) is described. The implications of the predatory nature of these species are discussed in terms of the feeding ecology of Gammarus and other amphipods. Predation rates by males on moulted congeneric females are assessed and compared. When males are present at the moult of a congeneric female, 100%of females are eaten in both species/sex combinations. When females are allowed to recover from moulting prior to presentation to a congeneric male, predation rates are significantly reduced in both species. However, significantly more G. d. celticus females are eaten by G. pulex males compared with the reciprocal situation. The predatory 'clumping' behaviour of both species is described and shown to be a form of feeding frenzy upon congenerics. Clumping behaviour also results in significantly higher rates of congeneric predation on G. d. celticus females than on G. pulex females. Support is thus given to the hypothesis that differential predation by males on moulted females may be the primary mechanism by which the introduced G. pulex has displaced the native Irish species G. d. celticus . The implications of amphipod congeneric predation are discussed in terms of observed patterns of interspecific coexistence and exclusion.     

The cannibalistic behaviour of two Gammarus species (Crustacea: Amphipoda)

This study examines the cannibalistic behaviour of the freshwater amphipods Gammarus duebeni celticus Stock & Pinkster, 1970 and G. pulex (L., 1758). In the first experiment, interactions were staged among all combinations of single adult males, single adult females, adults in the precopulatory mate-guarding phase and juveniles. Cannibalism by inter-moult individuals on newly moulted conspecifics occurred in all interaction categories in both species. Gammarus d. celticus , however, were significantly more cannibalistic than G. pulex. Cannibalism between and within sex and size categories (males > females > juveniles) was facilitated by the vulnerability of individuals at moult. Individuals of smaller size categories, however, did not cannibalize newly moulted conspecifics of larger size categories. Males were less cannibalistic on newly moulted females than on newly moulted males and juveniles and, when in the precopulatory condition, appeared to defend females from cannibalistic attacks. In a second experiment, stream conditions were simulated in the laboratory and replicated populations monitored for nine weeks. High levels of cannibalism, and the species and sex differences in cannibalism identified in the first experiment, were confirmed under these heterogeneous conditions. Cannibalism by males on their newly moulted female mating partners, termed 'reversed' sexual cannibalism, was further investigated. When males were deprived of foraging opportunities, cannibalism of precopulatory partners was significantly more frequent. The occurrence of 'reversed' sexual cannibalism is thus interpreted as a conflict between motivation to feed and motivation to mate.     

The survival and growth of Gammarus tigrinus Sexton (Crustacea: Amphipoda) in relation to salinity and temperature

Field collections from an inland saline lake indicated that a rapid increase in numbers of a population of Gammarus tigrinus was positively correlated with high summer temperatures and salinities. The laboratory experiments on growth are an attempt to determine whether high summer temperatures or optimum salinity were primarily responsible for the rapid increase in numbers. Furthermore, G. tigrinus was not found in either of two adjacent lakes; one non-saline, the other at a higher salinity. The laboratory experiments on survival are an attempt to determine whether the restricted distribution of G. tigrinus was related to salinity. The implications of the field and laboratory studies are discussed and compared with other data from Europe and North America.     

Differential drift and parasitism in invading and native Gammarus spp. (Crustacea: Amphipoda)

Invading and native species often interact directly, such as by predation, producing patterns of exclusion and coexistence. Less direct factors, such as interactions with the broader abiotic and biotic environment, may also contribute to such patterns, but these have received less recognition. In Northern Ireland, the North American Gammarus tigrinus has invaded freshwaters populated with the native Gammarus duebeni celticus , with intraguild predation between the two implicated in their relative success. However, these species also engage in day and night "drifting", an activity that subjects amphipods to intense predation from fish and wildfowl. Sampling of two rivers where the invader and native co-occur showed that, compared with the benthos, G. tigrinus was underrepresented and G. d. celticus overrepresented in the drift. In addition, G. tigrinus were free from parasites, whereas some G. d. celticus harboured the acanthocephalans Polymorphus minutus and Echinorhynchus truttae and the muscle wasting microsporidian Pleistophora sp. (new species). Compared with the benthos, G. d. celticus parasitized with P. minutus were overrepresented and unparasitized individuals underrepresented in the drift. The opposite pattern was found with Pleistophora sp . In laboratory experiments, G. tigrinus were less positively phototropic and less "active" than G. d. celticus (unparasitized animals). Polymorphus minutus increased G. d. celticus positive phototropism and activity, while Pleistophora sp. increased positive phototropism but decreased activity. Previous studies show that the invader G. tigrinus is more disadvantaged by intraguild predation from the native G. d. celticus than vice versa. However, the native appears more disadvantaged with respect to drift, parasitism and the interaction of the two. These factors may mitigate direct interactions and help explain complex patterns of coexistence between these invader and native species.     

Identifying Chinese species of Gammarus （Crustacea： Amphipoda） using DNA barcoding

Using a standard cytochrome c oxidase I sequence, DNA barcoding has been shown to be effective to distinguish known species and to discover cryptic species. Here we assessed the efficiency of DNA barcoding for the amphipod genus Gammarus from China. The maximum intraspecific divergence for widespread species, Gammarus lacustris, was 3.5%, and mean interspecific divergence reached 21.9%. We presented a conservative benchmark for determining provisional species using maximum intraspecific divergence of Gammarus lacustris. Thirty-one species possessed distinct barcode clusters. Two species were comprised of highly divergent clades with strong neighbor-joining bootstrap values, and likely indicated the presence of cryptic species. Although DNA barcoding is effective, future identification of species of Gammarus should incorporate DNA barcoding and morphological detection[Current Zoology 55(2):158-164,2009].     

Respiratory Levels and Adaptations in Four Freshwater Species of Gammarus (Crustacea: Amphipoda)

Respiration of four freshwater species of the amphipod crustacean Gammarus: G. fossarum, G. lacustris (river and lake), G. pulex and G. roeseli were measured in a closed, stirred respirometric chamber with a micro-electrode. Oxygen consumption, expressed as weight-specific oxygen uptake (R_s) in relation to decreasing oxygen concentration, varied at air saturation from 0.86 (G. lacustris, lake) to 2.06 μl O₂ mg⁻¹ AFDW h⁻¹ (G. pulex). R_s also differed intra-specifically among the two populations of G. lacustris. G. lacustris (river), G. pulex and G. roeseli expressed moderate ability to regulate their oxygen consumption at decreasing oxygen concentrations, whereas the regulation ability was higher in G. lacustris (lake) and in G. fossarum, which maintain high oxygen uptake at oxygen levels >2 mg O₂ l⁻¹. All four Gammarus species are partial regulators in response to variations in oxygen concentration. The differences between species are considered too small to account for their natural distributions. It appears that the tolerances of Gammarus species to organic pollution depend only in part on oxygen conditions.     

Relation between habitat and antennular segmentation in Gammarus and Echinogammarus (Crustacea: Amphipoda)

In 140 species of the amphipod genera Gammarus and Echinogammarus the ratio was calculated between the number of segments of the accessory and main antennular flagellum. Minor inter-and intrapopulational variation of this ratio was found in four species of the genus Gammarus. Seasonal variation was studied in one species of this genus. For both genera a clear correlation was found between antennular composition and the occurrence in marine or freshwater habitats.     

The geometry of natural selection in cave and spring populations of the amphipod Gammarus minus Say (Crustacea: Amphipoda)

The shape of the fitness function was investigated in a spring population (Davis Spring) and a cave population (Organ Cave) of the amphipod Gammarus minus in southern West Virginia. The first three principal components of a series of size, antennal and eye measurements and two fitness components (amplexus and number of eggs) were investigated. Smoothing splines using generalized cross-validation scores were found as were least-squares quadratic regression equations. Confidence intervals on shape, as determined by a smoothing parameter, were calculated using a bootstrap approach. Out of 18 cases, there were six cases of no selection, six cases of directional (monotonic) selection, one case of stabilizing selection, one case of disruptive selection, and four cases of selection with multiple internal maxima and minima. Selection on the cave population was frequently directional (5), resulting from strong directional selection for increased size, larger antennae, and smaller eyes. Selection on the spring population, the ancestral habitat for the species, was frequently non-monotonic (five cases) and rarely directional (one case). In both populations, stabilizing selection was rare and no more common than disruptive selection.     

The ultrastructure of the sinus gland of Gammarus oceanicus (Crustacea: Amphipoda)

Summary The sinus gland of Gammarus oceanicus, like that of other crustaceans, is composed of three elements: neurosecretory axons, glial cells and stromal sheath. Five neurosecretory axon types are identified on the basis of granule diameter, shape, and electron density, and axon matrix density. Exocytosis appears to be the major release mechanism of neurosecretory material. The preterminal regions of neurosecretory axons contain axoplasmic reticulum and neurotubules. Their arrangement in the axon and relationship with one another suggest a transport function. Multilamellar bodies are found in the terminal regions of neurosecretory axons. They arise from mitochondria and may be involved in granulolysis.The technical assistance of G.A. Bance, statistical assistance of D. MacCharles and D.W. Hagen, and financial support provided by the University of New Brunswick Research Fund to K.H. are gratefully acknowledged     

Effects of temperature, age and body size on moulting and growth in the freshwater amphipods Gammarus fossarum and G. roeseli

1. The total body length, cephalic length, wet weight and dry weight was measured in juveniles, males and females of Gammarus fossarum and G. roeseli kept in the laboratory. Numbers of flagellar segments on the first and second pairs of antennae of G. fossarum and G. roeseli were quantitatively related to body size and in stars. The addition of segments cannot be used to identify particular instars of individuals or to determine their ages in natural populations. 2. At experimental temperatures ranging from 3.8 to 20.2°C, the number of moults, duration of intermoults, maturation times and specific growth rates were studied from birth in isolated specimens. Sexual maturity was reached after 9 or 10 moults, at a mean wet weight of c. 5mg for females and c. 7mg for males of G. fossarum, and at c. 10 mg for females and c. 13 mg for males of G. roeseli. At 3.8°C neither species reached sexual maturity within 550 days. 3. The mean interval between moults was observed from birth to sexual maturity and was linearly related to moult number and exponentially related to age. The relationship between each intermoult interval, or the maturation time, and the experimental water temperature was described by a power function. Maturation times increased from 96 days at 20.2°C to 355 days at 7.9°C in G. fossarum, and 85 days to 403 days in G. roeseli. 4. Over the range 3.8–20.2°C there was a log—log relationship between temperature and specific growth rates. Growth was maximal at 20.2°C in newborn animals as well as in small sexually mature animals; interspecific differences were highly significant. 5. Increase in body wet weight of G. fossarum and G. roeseli fed ad libitum on a constant mixture of autumn-shed, naturally decaying, tree leaves and aquatic macrophytes was followed to senescence and death. The instantaneous or specific growth rate was maximal near birth, at c. 7.98% wet weight day^?1 in G. fossarum and 9.03% in G. roeseli. At ≥12°C, growth conformed to a logistic curve; maximum absolute increments in weight occurred about half-way through a life span of 280–300 days at 20°C, 380–420 days at 16°C and 550–600 days at 12°C. Some individuals lived longer than 850 days at ≤12°C. The wet weight at birth was 0.112mg for G. fossarum and 0.123mg for G. roeseli. Asymptotic mean body weights of males and females were, respectively, 61 and 41 mg for G. fossarum and 87 and 58 mg for G. roeseli. However, G. roeseli reached the inflection point of the logistic curve significantly faster than G. fossarum. In the latter species, growth and maturation were relatively faster at temperatures below 12°C, whereas they were faster in G. roeseli at 16–20°C. Thus G. fossarum is adapted to summer-cool streams and G. roeseli is adapted to summer-warm streams.     

Anatomy of the extrinsic gut musculature of Gammarus minus (Crustacea: Amphipoda)

Inserting on the buccal and esophageal foregut of Gammarus minus are numerous pairs of serially arranged dorsal dilator muscles, a single pair of lateral muscles, and two pairs of posterior muscles. Muscles of the cardiac stomach include three dorsal sets, a single pair associated with the pterocardiac ossicles, and two pairs inserting on the ventral aspect. A single pair of muscles inserts on the lateral aspect of the pyloric stomach. The extrinsic muscles of the foregut originate from exoskeletal apodemes of the cephalothoracic cuticle, sockets of the mandible, and a maxillary bridge that lies just ventral to the cardiac stomach. The extrinsic musculature of the hindgut is restricted to the rectal region and consists of paired dorsal and ventral series in an X-configuration. A single unpaired muscle inserts on the ventral midline. Extrinsic muscles of the hindgut originate from the integument of the last pleonic segment. The general arrangement of extrinsic gut muscles in G. minus is similar but not identical to that of other amphipods studied. However, the pattern is quite different from that of other malacostracans.