This paper describes the results of a multivariate and univariate morphometric analysis of three groups of Arctic charr Salvelinus alpinus from Loch Rannoch, Perthshire, Scotland, distinguished a priori on the basis of coloration (bright v. cryptic) and site of capture (from two locations 12 km apart). The analysis produced a clear and highly significant distinction, not only between brightly and cryptically coloured charr, but also between the two cryptically coloured groups, one of which was more robust in terms of several size-independent head measurements. Stomach analysis showed that the brightly coloured charr fed entirely on zooplankton, the less robust, cryptically coloured fish, fed on benthic macroinvertebrates, whereas a significant proportion of the diet of the more robust cryptically-coloured form consisted of other fish. Differences in length at age also distinguished the three forms, with the robust, piscivorous charr, which live longer and potentially reach a larger final size, attaining smaller sizes at a given age. These data clearly confirm the previous identification of a distinct planktivorous morph of Arctic charr in Loch Rannoch and extend the morph analysis by distinguishing two additional morphs, a benthivorous morph and a piscivorous morph that are morphologically and ecologically distinct. The results are discussed in the context of other systems in which sympatric morphs of Arctic charr have been described. 1998 The Fisheries Society of the British Isles

Gill-netted samples of Arctic cham from Loch Rannoch, Scotland were bimodal when tested by univariate and multivariate morphometric analyses. The separation into two morphs corresponded very closely (95-98%) with fish classified subjectively in the field as benthic or pelagic, based largely on colour differences and ecological observations. The benthic charr had relatively longer heads, larger eyes and more powerful jaws than the pelagic charr. Unlike sympatric niorphotypes described from Scandinavia and Greenland, neither Rannoch morph is dwarfed.

The full extent of sympatric intraspecific population diversity (SIPD) i.e. structuring into multiple genetically distinct Mendelian populations, remains uncertain for most lacustrine fish species, particularly in northern lakes. However, increasing application of molecular genetics is advancing understanding and has shown both that many known intraspecific sympatric lacustrine morphological and behavioural polymorphisms represent SIPD, and resolved the existence of phenotypically cryptic structuring. Uncertainty remains as only a few northern lakes have been comprehensively surveyed and existing studies focus on known phenotypic polymorphisms or exploit ad hoc field sampling, marker selection and data analysis methods. Such unfocused approaches constrain resolving power and increase the likelihood of failing to detect SIPD when present. Brown trout (Salmo trutta L.) in a previously unstudied Scottish lake were collected by 3D stratified random netting. They were screened with both an arbitrary set of molecular markers and markers pre‐selected for their ability to resolve regional allopatric brown trout population structuring and data analysed by both Bayesian and non‐Bayesian approaches. Depth clines for mitochondrial and microsatellite markers variation were observed and found to arise from different, but overlapping, depth distributions of four genetically distinct piscivorous, limnetic, shallow benthic, and profundal benthic populations. This is the highest number of ecologically and phenotypically distinct sympatric brown populations identified in any lake to date, and includes the first reported profundal benthic form. The detection of the SIPD in Loch Laidon depended critically on the random 3D sampling and using markers preselected for their power to differentiate regional allopatric populations of trout, as well as aided by using both Bayesian and non‐Bayesian analytical approaches. The findings support the view that the extent of ecologically and evolutionarily significant SIPD is probably underestimated in brown trout and other fish species, and probably is a significant component of the biodiversity in many northern hemisphere lakes.