Prédation exercée par le chabot (Cottus Gobio L.) sur l'alevin de truite commune (Salmo trutta L.): taille maximale de capture des alevins par les chabots

In laboratory conditions, where predation by sculpins (Cottus gobio L.) upon brown trout fry (Salmo trutta L.) was strong, predation was shown to be closely linked to total length of prey and predator. The limit of this predation can be defined by the equation: Y = 0.484 X + 5.8 (Y = total length of the trout fry in mm and X = total length of the sculpin in mm).

     

Distribution,production, and age structure of slimy sculpin in an Arctic lake

Synopsis Slimy sculpins (Cottus cognatus) were caught in the littoral region of Toolik Lake, an arctic lake. These sculpins grew slower and lived longer (71 mm at 8⁺ years) than more southerly populations. Sculpin distribution along the slope of the rocky littoral zone was greatest at the 3.5 m depth and coincided with the rock-mud interface. The hypothesis that this preferred area provided both increased prey and reduced predation was advanced. Yearly production estimate of littoral sculpins was 0.40 g · m^–2 · yr^–1 with a P/B ratio of 0.36. This is considerably less than estimates for more southerly populations of the same species.     

Evolution of rough sculpin (<Emphasis Type="Italic">Cottus asperrimus</Emphasis>) genetic divergence and late Quaternary displacement on the Hat Creek fault,California, USA

The rough sculpin (Cottus asperrimus) is a threatened species whose geographic range in northwestern California, USA is disrupted by Hat Creek fault. We tested whether the fluvial barriers (rapids and waterfalls) produced by this fault have generated significant phylogeographic structure among rough sculpin populations by analyzing variation in microsatellites and mitochondrial DNA. Rough sculpin isolated on either side of Hat Creek fault exhibited significant genetic divergence (microsatellite F _ST = 0.36; mitochondrial uncorrected p distance = 1 %). Independently derived estimates for the date of divergence, based upon a molecular clock and upon the age of slip on the Hat Creek fault are concordant and indicate divergence was initiated about 0.5–1 million years ago. Based upon the findings of our genetic analysis and the Pleistocene geologic history of midsections of the Pit River, we present a model of evolution of rough sculpin genetic divergence and late Quaternary displacement on Hat Creek fault. Our findings reveal that rough sculpin exhibit significant population structure and that two management units should be recognized within the species for future conservation planning.     

Partitioning of food resources by sculpins in Lake Michigan

Synopsis Resource partitioning was studied in two benthic Lake Michigan fishes, the deepwater sculpin,Myoxocephalus thompsoni, and the slimy sculpin,Cottus cognatus, that exhibit nearly disjunct distributions along a hypolimnetic depth gradient. Fish were collected in an area of sympatry over two 24 h periods. These sculpins exhibited food segregation—slimy sculpins ate primarilyPontoporeia affinis, deepwater sculpins ate bothP. affinis andMysis relicta — and their depth segregation was associated with the reported abundance of these prey. Different feeding behaviors may be responsible for this association, since slimy sculpins consumed intermediate size ranges ofPontoporeia and deepwater sculpins consumedPontoporeia in a pattern similar to the ambient size distribution. Neither fish showed a daily activity cycle, and both appear food-limited.     

Feeding rate of slimy sculpin and burbot on young lake charr in laboratory reefs

Synopsis Predation and contaminants are two possible factors in the poor recruitment of young lake charr Salvelinus namaycush in the Great Lakes. We measured the feeding rate of slimy sculpins Cottus cognatus and burbot Lota lota on young lake charr (uncontaminated young from eggs of a hatchery brood stock and contaminated young from eggs of Lake Michigan lake charr) in laboratory test chambers with a cobble substrate. The median daily consumption rate of sculpins for all tests was 2 lake charr eggs (N = 22 tests; 95% confidence interval, 0–13) and 2 lake charr free embryos (N = 31 tests; 95% confidence interval, 0–10). Feeding rate did not differ between hatchery and contaminated prey. Slimy sculpins continued to feed on lake charr when another prey organism, the deepwater amphipod Pontoporeia hoyi, was present. Feeding by burbot on free embryos (4–36 d^–1) increased as the mobility of young increased, but burbot consumed about 10% of their body weight weekly in free-swimming young (140–380 d^–1). Predation on lake charr eggs by sculpins could be considerable over the 100 to 140 d incubation period, and burbot could eat large numbers of free-swimming lake charr as the young fish left the reef. Predation pressure on young lake charr may inhibit rehabilitation of self-sustaining populations of lake charr on some reefs unless a critical egg density has been reached.     

Basin-scale Patterns in the Drift of Embryonic and Larval Fishes and Lamprey Ammocoetes in two Coastal Rivers

We studied the distribution and abundance of drifting embryonic and larval fishes and lampreys in the Smith and Van Duzen rivers of northern California, U.S.A. We collected seven fish species in four families and at least one lamprey species in the drift. All taxa drifted almost exclusively at night. Sculpins, Cottus aleuticus and C. asper, outnumbered all other taxa, comprising 63% of the catch in the Van Duzen River and 90% of the catch in the Smith River. We estimated that sculpin drift reached 3×10⁷ individualsh^–1 during the relatively high flow period from late winter through early summer. Most sculpin in these two rivers appeared to drift to the estuaries; we estimated 2.5×10⁹ sculpin embryos and larvae reached the Smith River estuary in 1995. In contrast to the sculpins, the patterns in the drift of other taxa suggest limited transport to the estuaries. Suckers, Catostomus occidentalis in the Van Duzen River, C. rimiculus in the Smith River, threespine stickleback, Gasterosteus aculeatus, and lamprey, Lampetra tridentata and possibly L. richardsoni, drifted at much lower rates and later in the year than sculpins. In the Van Duzen River, drift appeared to serve as a dispersal mechanism for only one of three introduced cyprinids. California roach, Lavinia symmetricus, drifted at low rates throughout the summer while we captured only seven Sacramento pikeminnow, Ptychocheilus grandis, and no speckled dace, Rhynichthys osculus. The information gathered on the drift of early life history phases is germane to both the conservation of native fishes and management of non-indigenous species in coastal rivers.     

Temperature preferences and oxygen consumption of three species of sculpin (Cottus) from the Pit River drainage,California

Synopsis I examined the temperature preferences and routine metabolic rates of Pit sculpin, Cottius pitensis, marbled sculpin, C. klamathensis macrops, and rough sculpin, C. asperrimus, of the Pit River drainage of California to determine if the distributional patterns of these species can be explained on the basis of physiological or behavioral responses to temperature. The routine metabolic rates of these species did not increase significantly between 10 and 15°C, indicating an area of thermal compensation. Metabolic rates then rapidly increased between 15 and 20°C (Q₁₀ values>4.0) followed by little increase between 20 and 25°C (Q₁₀ values >2.0), indicating another area of thermal compensation. The final temperature preferenda of Pit, marbled and rough sculpin were 11.2, 12.1 and 13.5°C, respectively. Marbled and rough sculpin appear to be more stenothermal than Pit sculpin. At acclimation temperatures of 10, 15 and 20°C the acute preferred temperatures of marbled and rough sculpin ranged from 11.1 to 14.7° C and 13.3 to 14.4°C, respectively. Values for Pit sculpin ranged from 9.9 to 16.4°C at acclimation temperatures of 10, 15 and 20°C. The distributions of marbled and rough sculpin are consistent with their behavioral and metabolic responses to temperature. The widespread distribution of Pit sculpin is consistent with its greater tolerance of high temperatures and eurythermal behavior, but the absence of Pit sculpin from habitats dominated by marbled and rough sculpin is not consistent with a temperature related explanation.     

The reproductive seasonality and gametogenic cycle of Acropora cervicornis off Broward County, Florida, USA

Reproductive characters of the Caribbean reef-building coral Acropora cervicornis were investigated based on histological samples collected from April 2001 through October 2002. Oogenesis commenced in early to mid-October through November and spermatogenesis was initiated from January to March. The onset of gametogenesis was staggered, exhibiting up to approximately a 1-month delay within colonies. In the hermaphroditic polyps, the observed male-to-female gonad ratio was nearly 1:1 and ripe oocytes represented over 70% of the total gonadal volume. Fecundity estimates based on Stage IV ova ranged between 10.4 and 21.8 mm³ per square centimeter per year, comparable to A. cervicornis in Puerto Rico and other broadcasting Indo-Pacific Acropora. Fecundity estimates based on Stage III vitellogenic oocytes indicated statistically significant differences among study sites. Spawning in field conditions was observed in 2001, 2003, and 2004 from 2300 to 2330 h. Gamete release generally occurred synchronously between nights two and seven after the full moon of July or August. However in 2003, multiple, small-scale gamete release episodes occurred over more than one lunar cycle. This coincided with the full moon occurring early in the month of July. While prolific gamete production is reported in this study, low levels of recruitment have been reported for this species. Thus, the highly fragmenting A. cervicornis may rely heavily on asexual reproduction for population maintenance and expansion, and recovery after disturbance may be greatly protracted.     

Spawning ecology of flannelmouth sucker, Catostomus lattipinnis (Catostomidae), in two small tributaries of the lower Colorado River

We report the first published accounts of spawning behavior and spawning site selection of the flannelmouth sucker in two small tributaries of the lower Colorado River in the Grand Canyon, Arizona. Spawning was observed on 20 March 1992 and from 28 March to 10 April 1993 in the Paria River, and from 16 to 19 March 1993 in Bright Angel Creek. Flannelmouth suckers exhibited promiscuous spawning behavior–individual females were typically paired with two or more males for a given event and sometimes changed partners between events. Multiple egg deposits by different females sometimes occurred at one spawning site. Flannelmouth sucker selected substrates from 16 to 32 mm diameter in both streams. Spawning occurred at depths of 10 to 25 cm in the Paria River and 19 to 41 cm in Bright Angel Creek. Mean column water velocities at spawning locations ranged from 0.15 to 1.0 m sec^-1 in the Paria River and from 0.23 to 0.89 m sec^-1 in Bright Angel Creek. Water temperatures recorded during spawning ranged from 9 to 18° C in the Paria River and 13 to 15° C in Bright Angel Creek. Spawning flannelmouth sucker ascended 9.8 km upstream in the Paria River and 1.25 km in Bright Angel Creek. Spawning females (410–580 mm) were significantly larger than spawning males (385–530 mm) in the Paria River. The mean size of spawning fish in the Paria River was significantly smaller than the entire stock, averaged throughout the study period (380–620 mm). However, fish spawning in 1992–1993 averaged 53 mm larger than fish spawning in the same reach of the Paria River in 1981, indicating a shift in the size structure of this stock.     

Balancing risks? Responses and non-responses of mayfly larvae to fish and stonefly predators

Summary In a series of laboratory experiments we examined the hypothesis that larvae of stream mayflies would respond to the presence of two different types of predators in such a way as to minimize their risk of being consumed by each. Positioning of larvae (whether they frequent the top, sides, or bottom of stones) of Baetis tricaudatus and Ephemerella subvaria was altered by the presence of predaceous stoneflies (Agnetina capitata) with a larger proportion of the population occurring on the upper surfaces, where the probability of encountering the predator was lowest. The presence of a benthivorous fish (Cottus bairdi) had no significant effects on positioning of the mayfly larvae. Lack of fish effects may reflect an inability of the mayflies to detect or respond to sculpins, or alternately may indicate that sculpins do not normally present a important predation risk for these mayflies. Failure of mayfly prey to account for fish predators when responding to the presence of stoneflies appcars to explain facilitation previously observed between stoneflies and sculpins.     

Biology of Volk’s Sculpin <Emphasis Type="Italic">Cottus volki</Emphasis> (Cottidae) from the Serebryanka River (Central Primor’e)

Biology of Volk’s Sculpin Cottus volki from the Sakhalin Channel is investigated (the lower reaches of the Serebryanka River). This fish species is slow and secretive, attains the length of 150 mm, weight 40 g, and age 6+. Males are larger and older than females. The highest rate of linear growth is observed in the first and second year of life, and that of weight growth is observed at the age of 4+ and 5+. They spawn from the end of April until the first ten-day period of June and are lithophilous. Females lay eggs on the lower surface of stones. Fecundity is low, up to 500 eggs. One male may spawn with several females. The feeding type is zoobenthophagous. Principal foods objects are chironomids, amphipods, and Trichoptera. It is an abundant species.     

Contemporary hybrid speciation in sculpins (Cottus spp.)

Natural hybridization between closely related taxa is frequent in many organismal groups, yet it has long been perceived as a force preventing diversification and speciation, especially so in animals. In recent years, growing evidence in favour of hybridization facilitating adaptive divergence has accumulated ( Mallet 2007 ; Mavárez & Linares 2008 ; Nolte & Tautz 2010 ). Homoploid hybrid speciation (the formation of hybrid lineages without changes in chromosome number) occurs when distinct species come into contact, hybridize, and at least in part of their range, produce hybrid swarms. If the hybrid genotypes can then colonize areas of the adaptive landscape inaccessible to ancestral species, they may eventually form new distinct lineages, reproductively isolated from their ancestors. Invasive sculpins (Cottus sp.) are one of a few good examples of homoploid hybrid speciation in animals. In this issue, Stemshorn et al. (2011) identified three distinct hybrid lineages, which have emerged out of a secondary contact situation of Cottus rhenanus and Cottus perifretum. Hybrids have recently invaded large river habitats unsuitable to ancestral species. Through the use of genetic mapping, the authors established that contrary to expectations, chromosomal rearrangements were not apparent in the hybrid lineages. In addition, different population genetic models were tested and the results suggest that contemporary gene flow from ancestral species represents an important component of the system. As such, recent and ongoing hybridization appears to be promoting the appearance of phenotypes adapted to novel environments. The examination of partially isolated lineages such as invasive hybrid sculpins should permit to identify early adaptive genetic changes before they become confounded by differences arising once speciation is complete.     

Migration and reproductive biology of Mugil liza (Teleostei: Mugilidae) in south Brazil

The mullet Mugil liza occurs along the Atlantic coast of South America from Venezuela to Argentina, but 95% of the commercial catch is collected from south Brazil between São Paulo and Argentina. Mugil liza is a single spawner with oocyte development occurring synchronously in two groups. Spawning happens in marine areas and occurs after migration. The reproductive migration occurs from Argentina (38° S) to the southern Brazilian states (24–26° S) from April to July, with peak spawning in June between northern Santa Catarina and Paraná. The presence of hyaline oocytes was associated with high salinity and sea surface temperatures of 19–21° C, and followed the seasonal northward displacement of these oceanographic conditions. The average size at first maturity (L_m) for both sexes was 408·3 mm total length, L_T. Males (L_m = 400·1) matured earlier than females (L_m = 421·9 mm). Fecundity ranged from 818 992 to 2 869 767 oocytes (mean = 1 624 551) in fish that were between 426 and 660 mm L_T.     

Behavioral effects of invaders: alien crayfish and native sculpin in a California stream

Invasive crayfish have been shown to have negative impacts on a range of taxa, though the mechanisms for those effects have not always been evaluated. Signal crayfish (Pacifastacus leniusculus) in Sagehen Creek were associated with reduced growth rates and gut fullness of Paiute sculpin (Cottus beldingi) in earlier experiments. This paper assesses potential behavioral mechanisms of competition between the two species. I conducted experiments to determine crayfish effects on sculpin behavior and habitat use in a stream observation facility at the Sagehen Research Station, California, USA. Sculpin reduced their use of refuges and pools, shifted into higher-velocity microhabitats, and spent more time fleeing in the presence of crayfish. Crayfish used refuges, pools, and low-velocity habitats more than sculpin in either treatment. Both species were notably nocturnal, with most activity at dusk and night observations, although crayfish were more strongly so than sculpin. Detailed field surveys of lower Sagehen Creek found that potential refuges (unembedded rocks) were closely associated with total crayfish and sculpin numbers, suggesting that cover is at least sometimes limiting under natural conditions. By displacing sculpin from refuges and pools and increasing their activity rate, crayfish may increase the likelihood of predation on sculpin. Behavioral shifts in sculpin appear to be at least partly responsible for the reduced growth rates of sculpin in the presence of crayfish.     

A benthic predatory fish does not cause selection on armour traits in three‐spined stickleback Gasterosteus aculeatus (Gasterosteiformes: Gasterosteidae)

Predation can promote divergence between prey populations and contribute to ecological speciation. In theory, predators can also constrain prey population divergence. In coastal British Columbia, Canada, Gasterosteus aculeatus (three‐spined stickleback) species pairs only occur in lakes with a single species of predatory fish: Oncorhynchus clarkii (the cutthroat trout). Similar lakes containing additional predatory fish species (Cottus asper, prickly sculpins; Oncorhynchus mykiss, rainbow trout) contain only single species of morphologically intermediate stickleback, suggesting that these predators prevent the coexistence of stickleback species pairs. We conducted a mesocosm experiment to investigate how prickly sculpins might constrain divergence, by quantifying their impact on survival and natural selection on antipredator (armour) traits in F₂ stickleback from a cross between ecologically divergent populations. We tested three hypotheses: (1) sculpin predation on sticklebacks reduces survival in a way that could result in their exclusion from certain niches; (2) sculpins compete with stickleback; (3) sculpins respond to prey vulnerabilities in similar ways to cutthroat trout, tending to constrain rather than to enhance divergence. We found that sculpins significantly reduce stickleback survival, that their presence per se does not reduce growth in stickleback, and that predation did not result in selection on any of the armour traits measured, or on gill raker length, which is an important trophic trait. These results tend to refute hypotheses (2) and (3), while supporting hypothesis (1). © 2011 The Linnean Society of London, Biological Journal of the Linnean Society, 2011, 104 , 877–885.     

Evolutionary history of freshwater sculpins, genus Cottus (Teleostei; Cottidae) and related taxa, as inferred from mitochondrial DNA phylogeny

The freshwater sculpins, genus Cottus (Teleostei; Cottidae), comprise bottom-dwelling fishes that exhibit various life-history styles, having radiated throughout Northern Hemisphere freshwater habitats. The phylogenetic relationships among Cottus and related taxa were estimated from mitochondrial DNA 12S rRNA and control region (CR) sequences, the freshwater sculpins examined falling into five lineages (A-E). Lineage A consisted of Trachidermus fasciatus and C. kazika, both having a catadromous life-history. The remaining species (lineages B-E) spawn in freshwater habitats regardless of life-history (amphidromous, lacustrine or fluvial), suggesting that the various life-history types post-dated a common ancestor of lineages B-E. Molecular clock estimates suggested a Pliocene-Pleistocene radiation (or Miocene-Pliocene from the alternative clock) of lineages B-E. In eastern Eurasia, speciation with life-history changes to amphidromous or fluvial styles has apparently occurred independently in some lineages, as a general pattern. Mitochondrial DNA CR phylogeny showed the monophyletic Baikalian cottoids (Cottoidei) to be nested within Cottus and Trachidermus, suggesting that the former ecologically and morphologically divergent cottoids may have originated from a single lineage which invaded the ancient lake.     

Reproductive biology ofAwaous guamensis, an amphidromous Hawaiian goby

Synopsis Spawning season, size at first reproduction, oocyte maturation, and fecundity ofAwaous guamensis, an amphidromous Hawaiian goby, were studied from June 1989 through May 1991 in the Wainiha River, Kau'ai, Hawai'i. Female fish larger than 73 mm standard length (SL) had mature gonads from August through December in 1989 and 1990. Gonadosomatic index (GSI) values for mature females ranged from 0.2 to 14.5 during the spawning season. Male fish larger than 64 mm SL had elevated GSI values from June 1989 through December 1989 and from August 1990 through December 1990. Mature sperm were found in two male fish collected in January and February. GSI values for mature males ranged from less than 0.01 to 4.0 in the spawning season. Size-frequency distributions of measurements of vitellogenic oocyte diameters and microscopic observations of oocytes indicated this species has group-synchronous oocyte development. Ovarian maturation stages examined over a 29-month period suggest that members of the stock spawned at different times within the spawning season, although mass spawning events have been documented for this species. Estimates of clutch sizes from nests measured in situ were comparable to estimates of potential fecundity from in vitro examination of ovaries, and indicated that female fish deposited an entire clutch during a spawning event. No evidence for multiple spawning by an individual fish in a single season was found. However, microscopic observations of brown bodies in some ovaries suggested that individual fish probably spawn more than once in a lifetime.     

Comparative ecology of prickly sculpin,Cottus asper,and coastrange sculpin,C. aleuticus,in the Eel River,California

Synopsis We documented species' distributions, size structure of populations, abundance in mainstem and tributary streams, habitat use, and diets of prickly sculpin, Cottus asper, and coastrange sculpin, C. aleuticus, in the Eel River drainage of California, to determine the processes allowing coexistence of these very similar fishes. We observed prickly sculpins at 43 sites and coastrange sculpins at 34. The species co-occurred at 26 sites. Young-of-year coastrange sculpins were only observed within 42 km of the ocean, but young-of-year prickly sculpins were present throughout the species range. Mean, maximum, and minimum lengths of coastrange sculpins were positively correlated with distance from the ocean but no significant relationships were found for prickly sculpins. Absolute abundance of both species was highest in mainstem habitat (prickly sculpins = 0.6 sculpins m^–2 and coastrange sculpins = 0.4 sculpins m^–2) . Tributary densities of both species tended to be less than 0.1 sculpins m^–2. The species inhabited very similar habitats and had very similar diets. Coastrange sculpin populations in upstream areas were maintained by immigration from downstream areas in contrast with prickly sculpin populations that produced young-of-year fish throughout their range. Densities were probably not high enough for interspecific interactions to be important. The factors limiting the upstream distribution of the species may include high water temperatures, stability of the stream bed, and behavior of the fish. In the past, the range of sculpins within the Eel River drainage probably fluctuated with changing physical conditions. Recent introductions of exotic species that compete with and prey upon sculpins, and ongoing human activities in the drainage could result in major reductions in the distribution and abundance of one or both species.     

Genetic variations ofCottus nozawae populations from five tributaries of the Shiribetsu River of southern Hokkaido,Japan

To estimate genetic differentiation and heterogeneity in the landlocked river sculpin,Cottus nozawae, between tributary populations in the same river-system, 107 specimens were captured from 5 tributaries of the Shiribetsu river (course length 128 km), Hokkaido Island and surveyed for allozyme variations and restriction fragment length polymorphisms of mitochondrial DNA (mtDNA). Three and two alleles were seen at theIdh-2 andPgm loci, respectively, but only one locus,Idh-2, out of twenty loci examined was regarded as polymorphic, since the frequency of the most common allele did not exceed 0.95. Three different mtDNA haplotypes were detected, there being replacement of them between the tributary populations. Heterogeneities of allele and haplotype frequencies were significant between some tributary populations, suggesting that genetic differentiation has occurred between them.     

Coarse-grained population structure in Central European sculpin (Cottus gobio L.): secondary contact or ongoing genetic drift?

Population genetic affinities of 261 European sculpins Cottus gobio L. across the Rhenanian-Danubian and the Rhenanian-Rhónian watersheds were assessed by horizontal agarose-gel electrophoresis of up to 20 allozyme systems (encoded by 29 genetic loci). Polymorphism P_mean= 0.0689 (range: 0.000-0.1379), and heterozygosity H_e(mean)= 0.0167 (range: 0.000-0.0507) indicated low genetic variability within local stocks from single streams. Significant genetic distances D_mean= 0.1917 ± 0.0336 (D_max= 0.2407), based on differential fixation at 3–6 loci (Acp-1**, Pgdh**, Fh**, Est-1**, Gpi-2**, and Pgm-1**) distinguished populations from the Neckar catchment basin from those of the Hochhrein-Oberrhein and Danube basins. Differential fixation of alleles and pronounced genetic distances also separated sculpins of the Rh6nian tributary Doubs from Neckarian populations (D_men= 0.2131 ± 0.0033; Ah**, Acp-1**, Fh**), of the Doubs from the Danube (D_mean= 0.2177 ± 0.0028; Gpi-2**, Pgm-1**, Pgdh**, Ah**), and of the Doubs from the Hochrhein-Oberrhein (D_mean= 0.1780; Pgm-1**, Pgdh**, Ah**). Genetic distances between streams within these drainages proved low (Neckar: D_men= 0.0047 ± 0.0014, Danube: D_mean= 0.026 ± 0.0179, Rhine: D_mean= 0.0308). Screening of another 16 small-size samples consisting of 55 sculpins for nine diagnostic loci (Aat-2**, Acp-1**, Acp-2**, Ah**, Est-1**, Fh**, Gpi-2**, Pgdh** and Pgm-1**) confirmed the genetic homogeneity of sculpins within the Danubian and Neckarian drainage systems, but Neckarian sculpins were similar to those from the river Main. Populations from Hochrhein-Oberrhein resembled the Danubian stock but contained a decreasing frequency of ‘Neckarian’ markers when approaching the Danubian region. The genetic divergence between Neckarian, Danubian and Rhdnian sculpins suggests the existence of hitherto neglected taxa of anteglacial divergence.