BENTHIC DIATOM AUTECOLOGY AND INFERENCE MODEL DEVELOPMENT FROM THE CANADIAN HIGH ARCTIC ARCHIPELAGO1

Diatom assemblages were analyzed from 64 lakes and ponds from Alert, Ellesmere Island and Mould Bay, Prince Patrick Island in the Canadian High Arctic Archipelago. Diverse water chemistry conditions and diatom communities were present in these sites. Small benthic taxa typically dominated diatom communities; however, assemblages were markedly different between Alert and Mould Bay sites in response to disparate water chemistry characteristics in the two regions. The most abundant taxa belonged to the genera Navicula, Cymbella, Achnanthes, Nitzschia, and Pinnularia. Canonical correspondence analysis indicated that pH, specific conductivity, dissolved organic carbon, and total phosphorus were the most important limnological variables in determining species composition. Diatom inference models were developed for pH, specific conductivity, and dissolved organic carbon using weighted averaging and weighted averaging partial least squares techniques; these had root mean square error of prediction/r²_boot values of 0.40/0.77, 0.28/0.70, and 0.24/0.55, respectively. These models are applicable to sites with large ranges of taxonomic and limnological variation and will allow the reconstruction of past changes of climate‐related limnological parameters from biostratigraphic records in future paleolimnological studies.     

A landscape approach to examining spatial patterns of limnological variables and long-term environmental change in a southern Canadian lake district

1. We explored patterns of limnological variables (physical, chemical and biological) with relation to landscape position (expressed as lake order) in 86 study lakes located on shield bedrock in south‐central Ontario, Canada. 2. Using anova s with lake order as the categorical variable, landscape position explained significant amounts of variation in major ion chemistry, physical and catchment characteristics, hypolimnetic oxygen, and community composition in algal (diatom, chrysophyte) and invertebrate (chironomid) assemblages preserved in surficial sediments. Several nutrient variables (TP, total phosphorus and TN, total nitrogen) and dissolved organic carbon did not have significant relationships with lake order. 3. The strongest relationships with lake order (as a fraction of variation explained in anova s) included silica concentrations (r² = 0.40) and SO₄ (r² = 0.29) concentrations, surface area (r² = 0.50) and hypolimnetic oxygen (r² = 0.29). 4. Bedrock geology (carbonate metasedimentary versus non‐carbonate bedrock) had strong influences on spatial gradients of pH and major ion chemistry. It was difficult to separate geological influences from spatial influences on limnological variables in this study, as drainage patterns in the region are highly influenced by surface features of underlying geological formations because of the very thin glacial till or exposed bedrock that exists in most catchments. 5. Patterns of limnological variables indicated that low‐order, headwater lakes had the lowest concentrations of major ions, and, from algal inferences of pH change, had been most susceptible to acidic deposition. High‐order, downstream lakes were larger and deeper, and had higher concentrations of hypolimnetic oxygen, indicating that optimal lake trout habitat was primarily located in high‐order lakes. 6. Variance partitioning analyses indicated that lake order as a metric of landscape position explained comparable portions of community variation in algal and invertebrate assemblages compared with geographic position (latitude, longitude) and Cartesian coordinate position (e.g. x, y, x², y², etc.) metrics. Lake order explained more community variation in chironomid assemblages compared with other landscape metrics, possibly because of the strong relationships between lake order and lake morphometry variables.     

Diatom–environmental relationships and nutrient transfer functions from contrasting shallow and deep limestone lakes in Ontario, Canada

Diatom assemblages were analysed in the surface sediments of 44 alkaline lakes in south-western Ontario, Canada, and combined with a pre-existing 58 south-eastern Ontario lake set: (1) to determine if shallow, polymictic Ontario lakes contain different diatom assemblages from deeper, dimictic lakes, and if so, which environmental variables most influence assemblages; (2) to improve the existing transfer functions; (3) to construct and compare transfer functions separately for dimictic, deep lakes and for polymictic, shallow lakes. Polymictic and dimictic lakes covered a similar nutrient range (spring total phosphorus (TP)=4–54 g/l, spring total nitrogen (TN)=200–927 g/l; n=101) and spring pH levels (7.6–9; n=101). However, polymictic lakes were shallower (median mean depth = 2.9 m vs. 7.3 m in dimictic lakes). Benthic diatoms (average 60% relative abundance) dominated the polymictic lakes, whereas planktonic diatoms (average 60%) dominated dimictic lakes. A Canonical Correspondence Analyses with forward selection (p < 0.05, 999 Monte Carlo permutations) identified TP, alkalinity, watershed to volume ratios and lake depth as the most important measured environmental variables influencing diatom distribution in both polymictic and dimictic lakes. Additionally, pH was identified as an important variable in polymictic lakes, whereas TN was also forward selected in the dimictic lakes. Adding more lakes to the original southern Ontario calibration set improved the TN transfer function (r² _jack=0.42, root mean squared error of prediction (RMSEP)_jack=0.11 [log g TN/l]), although there was a high systematic error in the revised model (r² _residual = 0.48). However, the strongest TP model was derived from the polymictic lakes (r² _boot =0.44, RMSEP_boot=0.20 [log g TP/l]), which was the smallest lake set (n=30) with the lowest number of diatom species. The stronger TP model from the polymictic lakes may be partly due to the relatively low macrophyte cover in our polymictic lakes, which may lead to stronger benthic–pelagic coupling than in lakes with large macrophyte populations. Additionally, our study suggests that the Chrysophyceae cyst:diatom frustule ratio may be useful for indicating trends in TP levels of 35 g/l in alkaline lakes that are dimictic, but is not necessarily indicative of trophic state changes in shallow, polymictic lakes. Our study demonstrates that it may be important to construct separate diatom-based nutrient transfer functions for polymictic and dimictic lakes.     

FRESHWATER DIATOMS FROM THE CANADIAN ARCTIC TREELINE AND DEVELOPMENT OF PALEOLIMNOLOGICAL INFERENCE MODELS1

Relationships between surface sediment diatom assemblages and measured environmental variables from 77 lakes in the central Canadian arctic treeline region were examined using multivariate statistical methods. Lakes were distributed across the arctic treeline from boreal forest to arctic tundra ecozones, along steep climatic and environmental gradients. Forward selection in canonical correspondence analysis determined that dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), total nitrogen (TN), lake surface area, silica, lake‐water depth, and iron explained significant portions of diatom species variation. Weighted‐averaging (WA) regression and calibration techniques were used to develop inference models for DIC, DOC, and TN from the estimated optima of the diatom taxa to these environmental variables. Simple WA models with classical deshrinking produced models with the strongest predictive abilities for all three variables based on the bootstrapped root mean squared errors of prediction (RMSEP). WA partial least squares showed little improvement over the simpler WA models as judged by the jackknifed RMSEP. These models suggest that it is possible to infer trends in DIC, DOC, and TN from fossil diatom assemblages from suitably chosen lakes in the central Canadian arctic treeline region.     

ENVIRONMENTAL FACTORS CORRELATED WITH CHRYSOPHYTE CYST ASSEMBLAGES IN LOW ARCTIC LAKES OF SOUTHWEST GREENLAND1

We analyzed the relationship between chrysophyte cyst assemblages in surface sediment samples and limnological and geographical variables for 70 lakes located along Søndre Strømfjord in southwest Greenland. Over 247 stomatocysts were identified and of these, 153 were sufficiently abundant for use in statistical analyses. Eight stomatocysts were considered to be new and are described formally. Canonical correspondence analysis indicated that conductivity was the dominant variable explaining cyst distribution, reflecting the large conductivity gradient in lake water chemistry in this area. High conductivity lakes had distinctive cyst assemblages with lower diversity than low alkalinity lakes, where assemblages were similar to alpine soft‐water lakes elsewhere. The high conductivity lakes, however, had similar cysts to other saline lakes elsewhere in the arctic. Additionally, pH, calcium, maximum depth, longitude, sulfate, total phosphorus, and altitude all explained significant amounts of variability of cyst assemblages. Longitude was the only geographical variable that explained cyst variability independently of other variables (i.e. had a unique effect), which suggests that the climatic gradient from the coast to the head of the fjord has a structuring effect on cyst assemblages. Conductivity (weighted‐averaging partial least squares, r²=0.917; root mean square error=0.142; r²_jack=0.861, root mean square error of prediction=0.191) and pH inference models (weighted averaging, r²=0.924; root mean square error=0.158; r²_jack=0.826, root mean square error of prediction=0.240) were developed. For the pH model, high conductivity lakes (>800 μ S₂₀·cm^?1) were removed. Both models are statistically robust and could be applied to lakes in west Greenland to reconstruct conductivity and/or pH. Such paleolimnological reconstructions provide the means of acquiring long‐term data for use in the evaluation of, for example, regional paleoclimatic models.     

Tracing lake trophic history with a chironomid–total phosphorus inference model

1. In the absence of historical water chemistry data, predictive biological indicator groups preserved in lake sediments can be employed to reconstruct the history of lake eutrophication. Diatoms are well established in this role, but to augment diatom‐based inferences of nutrient status we investigate the potential use of chironomid midges (Insecta: Chironomidae). 2. Canonical correspondence analysis (CCA) of modern chironomid assemblages in surface sediments from 44 lakes in the English Midlands and Wales, U.K., shows that five environmental variables (total phosphorus (TP), bottom dissolved oxygen, maximum lake depth, Secchi depth and surface water temperature) make a statistically significant (P < 0.05) contribution to explaining the variance in the chironomid data, of which TP makes the largest contribution (29%). 3. The relationship is used to develop a series of weighted averaging (WA) and partial least squares (PLS), (WA‐PLS) models to infer log₁₀TP. The models are evaluated by leave‐one‐out (jack‐knifing) cross‐validation. The simplest minimal adequate model is provided by WA with unweighted inverse deshrinking of root mean square error of prediction (RMSEP_jack=0.34 and r²_jack=0.60). 4. Using this model, the trophic history of Betton Pool, Shropshire, U.K., is reconstructed from the mid‐19th century to the present day and the results from the chironomid‐TP model are compared with inferences from a diatom‐TP model ( 13 ). Both reconstructions suggest that there was a gradual rise in TP since 1850 AD until about 1974, followed by a more pronounced and rapid increase that has continued until the present. Inferred TP values from the WA chironomid inference model agree with diatom‐inferred values. 5. The study demonstrates that fossil chironomid assemblages can be used to investigate quantitatively the trophic history of lakes.     

Chironomid-based inference models for estimating end-of-summer hypolimnetic oxygen from south-central Ontario shield lakes

1. Subfossil chironomid head capsules were sampled from surficial sediments from 86 boreal shield lakes in south‐central Ontario, Canada. Lake characteristics ranged from shallow to very deep (> 80 m), ultraoligotrophic to mesotrophic, and with end‐of‐summer hypolimnetic oxygen conditions ranging from near‐saturation to anoxic. 2. Subfossil chironomid assemblages, comprising 44 taxa from 59 lakes, were analysed using multivariate ordination techniques such as redundancy analysis (RDA) and canonical correspondence analysis (CCA). Forward selection in RDA and CCA both showed that measures of oxygen, such as end‐of‐summer volume‐weighted hypolimnetic oxygen concentration (VWHO) and bottom oxygen concentration (botO₂), were the strongest explanatory variables for the chironomid data. Maximum depth and major ion chemistry were also important explanatory variables. 3. Oxygen inference models were developed using partial‐least‐squares regression (PLS), weighted‐averaging partial‐least‐squares regression (WA‐PLS), and weighted averaging regression (WA). Models were developed using both the full 44 taxa assemblage (which included littoral taxa) and using only 15 profundal‐type taxa. 4. Cross‐validated models (jackknifing) using full‐assemblage or profundal‐only taxa had similar statistical power (similar root mean squared error of prediction, RMSEP). The best models had moderate predictive power, with an r²_jack as high as 0.56, and an RMSEP as low as 2.15 mg L^–1 for [VWHO], and an r²_jack of 0.49 and an RMSEP of 0.24 for log([botO₂] + 1). 5. Reconstruction of [VWHO] and [botO₂] using a previously published chironomid profile that showed strong lake response to land‐clearance and logging suggests that oxygen inference models are reliable and accurate, reflecting the qualitative changes occurring in subfossil assemblages. However, the profundal‐only models may be misleading in situations where the ratio of littoral‐to‐profundal subfossils changes drastically in response to lake disturbance.     

A weighted—averaging regression and calibration model for inferring total phosphorus concentration from diatoms in British Columbia (Canada) lakes

1. The relationship between surficial sediment diatom taxa (Bacillariophyceae) and measured limnological variables in forty-six British Columbia lakes was explored using canonical correspondence analysis (CCA). Lake-water total phosphorus concentration (TP), maximum lake depth, conductivity, and calcium concentration each accounted for independent and statistically significant directions of variation in the distribution of diatom taxa. 2. Weighted-averaging (WA) models were developed to infer lake-water TP from the relative abundances of 131 diatom taxa in the surficial sediments of thirty-seven lakes. WA regression and calibration with classical deshrinking provided the best model for TP reconstructions. 3. Our quantitative inference model has two major advantages over existing multiple linear-regression models: (i) inferences are based on the responses of individual taxa to TP, and do not involve grouping the taxa into a small number of ecological categories; and (ii) the model assumes that diatoms respond to TP in a unimodal, rather than a linear, fashion. 4. The WA model can now be used to infer past lake-water TP, within the range 5–28νgr1^?1, from diatoms preserved in the sediments of British Columbia lakes. The model can provide quantitative estimates of the onset, rate, and magnitude of lake eutrophication in response to natural processes and human disturbances.     

Validation of a diatom–phosphorus calibration set for Sweden

1. A weighted averaging (WA) regression and calibration model for diatoms and total phosphorus (TP) was developed from a dataset of 45 surface‐sediment samples from Swedish lakes. Jack‐knifed error statistics were comparable with those for similar diatom–TP datasets: r²_jack=0.47, root mean squared error of prediction (RMSEP)=0.24 log₁₀μg TP L^–1 and mean bias=–0.002 log₁₀ μg TP L^–1 for the simple WA model; r²_jack=0.36, RMSEP=0.27 log₁₀ μg TP L^–1 and mean bias=0.017 log₁₀ μg TP L^–1 for WA with tolerance downweighting. 2. The model was used to estimate TP concentrations for the Ekoln basin of Lake Mälaren using a ²¹⁰Pb‐dated sediment core record. Highly eutrophic conditions developed in the basin in the 1960s as a result of nutrient inputs from cultivated land and the city of Uppsala. A reduction in the supply of phosphorus from sewage outlets in the late 1960s resulted in less eutrophic conditions. 3. The model results indicated levels of 50–60 μg TP L^–1 prior to 1900. The rapid eutrophication of the lake basin after the 1950s and the subsequent recovery were evident from the diatom data. 4. Diatom‐inferred TP (DI–TP) values were validated by comparison with monitored data for the period 1966–95. The diatom model tended to underestimate TP at high levels (> 80 μg L^–1) but overestimate at lower concentrations. 5. A good agreement was observed between the trends in TP concentration and the DI–TP concentration and the timing of the maximum was well reflected by the diatom‐based reconstruction. A significant correlation (r²=0.69, P < 0.01) was found between DI–TP and measured TP at this site.     

Surface sediment diatom assemblages and epilimnetic total phosphorus in large, shallow lakes of the Yangtze floodplain: their relationships and implications for assessing long-term eutrophication

1. The Yangtze floodplain (SE China) is characterized by a number of large shallow lakes, many of which have undergone eutrophication due to the intensification of agriculture and urban growth over recent decades. As monitoring data are limited and in order to determine lake baseline nutrient concentrations, 49 lakes were sampled, covering a total phosphorus (TP) gradient (c. 30–550 μg L⁻¹) to develop a diatom-based inference model. 2. There are three dominant diatom assemblages in these shallow lakes with a marked change in assemblage structure near the boundary between eutrophic and hypereutrophic nutrient levels (as indicated by their TP value). Canonical correspondence analysis indicated that TP was the most important and significant variable in explaining the diatom distributions, independently accounting for 9.5% variance of diatoms. 3. Forty-three lakes were used to generate a transfer function using weighted averaging (WA) with inverse deshrinking. This model had low predictive error (root mean squared error of prediction; RMSEP_jack = 0.12) and a high coefficient of prediction (R²_jack = 0.82), comparable with regional TP models elsewhere. The good performance of this TP model may reflect the low abundance of benthic diatom species which are commonly regarded as the main error source in European shallow lake WA models. 4. The WA model was used to reconstruct the past-TP concentrations for Taibai Lake, a shallow hypereutrophic lake in Hubei province. The results showed that TP concentration varied slightly (43–62 μg L⁻¹) prior to the 1920s, indicating an eutrophic state since the 1800s. A period of sustained eutrophication occurred after 1950, because of the development of agriculture, reflecting by maximum values of Aulacoseira alpigena and increased abundance of Cyclotella meneghiniana, C. atomus and Cyclostephanos dubius. The steep increase in nutrient concentration after 1970 was related to the overuse of chemical fertilizer and fish farming in the catchment. 5. The shift in fossil diatoms from epiphytic to planktonic forms in the lake sediment core during 1950–70 provides useful information on the transformation from macrophyte-dominated to alga-dominated states. It is plausible that the TP concentration of 80–110 μg L⁻¹ observed in this study is the critical range for switching between the two stable states in the lake. 6. The regional diatom-TP model developed in this study allows, therefore, the possibility of reconstructing historical background nutrient concentrations in lakes. It will provide an indication of the onset and development of eutrophication at any site. This is particularly important for the many lakes in the Yangtze floodplain where information about historical changes in water quality is lacking.     

Biogeography and lake morphometry drive diatom and chironomid assemblages’ composition in lacustrine surface sediments of oceanic islands

Subfossil biotic assemblages in lakes’ surface sediments have been used to infer ecological conditions across environmental gradients. Local variables are usually the major drivers of assemblage composition, but in remote oceanic islands biogeographic filters may play a significant role. To assess the contribution of local and regional filters in the composition of subfossil diatom and chironomid assemblages in surface sediments, 41 lakes in Azores archipelago were studied and related to environmental variables. Ordination techniques were used to identify the forcing factors that best explain the composition of these assemblages. Both assemblages are influenced by multiple limnological variables (conductivity, pH and nutrients). However, diatom assemblages differed mainly in the proportion of planktonic versus benthic species along lakes’ depth gradient while chironomids differed significantly among islands but not among lake depths. Thus, biogeographic filters play an important role in shaping islands’ freshwater communities, particularly insect ones, more influenced by geographic variables. Results demonstrate the accuracy and potential of biotic remains in sediments for applied studies of lake ecology, trophic status, climatic trends and ecological reconstruction and evolution of lakes. In the Azores, the application of this information for the development of inference models is envisaged as a further step to accomplish these goals.     

DIATOM ASSEMBLAGES AS INDICATORS OF LAKE TROPHIC STATUS IN SOUTHEASTERN ONTARIO LAKES1

Canonical correspondence analysis (CCA) was used to explore the relationship between measured environmental variables and surficial diatom (Bacillariophyceae) assemblages in alkaline lakes from southeastern Ontario. Total nitrogen (TN), watershed area, alkalinity, and maximum depth each explain significant (P ≤ 0.05) directions of variance in the distribution of diatom taxa. TN was highly correlated to total phosphorus (TP) (r = 0.92), chlorophyll a (r = 0.86), and Secchi depth (r =0.77). When a series of CCAs were run with the first axis constrained to each of these variables in turn, the ratio of the eigenvalue of the first axis to that of the second axis (λ1/λ2) was highest for TN, indicating that TN best explained the distribution of the diatom assemblages in this set of lakes. Furthermore, results of Monte Carlo permutation tests indicated that these four variables did not act independently on the diatom assemblages. Therefore, TN was selected to represent these four closely related variables to infer lake trophic status. Weighted-averaging regression and calibration (with classical deshrinking) were used to develop transfer functions to infer TN from the relative abundances of 83 diatom taxa recovered from the surficial sediments of 51 lakes. There was a good correlation between diatom-inferred TN concentrations and measured TN concentrations (r²= 0.75, n = 51). The weighted-averaging regression and calibration model was used to infer lake trophic status (represented by TN) from diatom assemblages presented in the sediments from Little Round Lake, Ontario. These data were used in conjunction with historical land-use data in order to quantify the sequence and extent of nutrient enrichment related to human activity in the watershed area.     

Development and evaluation of a diatom-conductivity model from lakes in West Greenland

1. The area around Kangerlussuaq (Søndre Strømfjord; West Greenland, 67°N 51°W) contains thousands of lakes ranging from coastal, dilute (conductivity < 30 μS cm^–1) oligotrophic systems to subsaline (～4000 μS cm^–1), closed basin lakes close to the ice sheet margin. In closed basins, salinity (or conductivity) is often a proxy for effective moisture, and thus palaeorecords of lake conductivity can provide valuable palaeoclimatic data. Little or nothing is known about the recent history of these lakes and hence it is difficult to evaluate how they will respond to effects of future changes. 2. Over 100 lakes have been sampled (1996–2000) between the ice sheet and the outer coast for a variety of water chemical and limnological variables. Surface sediments were taken from a subset of 40 lakes and analysed for diatoms. Diatom responses to 28 environmental variables were analysed by multivariate ordination techniques and indicate that the main gradient is highly correlated to conductivity (explaining ～12% of species variance). Despite the relatively short gradient (24–4072 μS cm^–1), diatom assemblages exhibit a clear response to conductivity. The most saline lakes do not contain a true saline flora. 3. We developed a range of weighted averaging (WA) and weighted‐averaging partial least squares (WA‐PLS) models from this training data set and found two component WA‐PLS models performed best. The effects of data transformation and omission of dissolution susceptible species (Diatoma spp.) on model performance were also examined. The error statistics for the preferred WA‐PLS (2) model (r²_jack=0.88, root mean square error of prediction (RMSEP)=0.217 log μS cm^–1) compare well with other published models. 4. A ²¹⁰Pb‐dated short core from a meromictic, subsaline lake (Braya Sø; location 67°N, 51°W, max. depth 23 m; conductivity 2600 μS cm^–1) was analysed for diatoms. Diatom preservation is poor and some taxa (e.g. Diatoma spp.) are badly corroded. Lake water conductivity was reconstructed using WA‐PLS models. Diatom‐inferred conductivity ranges from 1800 to 4400 μS cm^–1 over the last 600 years (extrapolated ²¹⁰Pb chronology). 5. The Kangerlussuaq area of West Greenland is an important area for palaeoclimatic research, located as it is between the Greenland ice sheet (ice core records) and the Davis Straits to the west. The development of a statistically robust transfer function for diatoms and conductivity will enable the reconstruction of conductivity from the many subsaline lakes around the head of the fjord and, hence, regional estimates of changing palaeoprecipitation.     

The fauna in the upper stony littoral of Danish lakes: macroinvertebrates as trophic indicators

1. The macroinvertebrate fauna living on stones in the exposed stony littorals of thirty-nine Danish lakes were examined by multivariate numerical methods. The data were derived from 125 semi-quantitative samples and a species list of 126 taxa. The mean number of individuals per sample was 960, and among the most common taxa were Asellus aquaticus, Gammarus, Oulimnius, Tinodes, Cricotopus and Dicrotendipes. 2. The total number of species and fourteen individual taxa were positively correlated to mean depth of the lakes and eleven taxa were correlated to the total phosphorus concentration. The Shannon diversity was negatively correlated to the chlorophyll a concentration ([Chl a]). 3. Community patterns were examined by detrended correspondence analysis (DCA), and the relationship between species data and selected environmental variables was analysed by canonical correspondence analysis (CCA). Mean lake depth was found to be the strongest environmental variable in explaining the species data. The [Chl a] and Secchi depth also explained significant variation in the distribution of the stony littoral invertebrates. Wind fetch and relative exposure did not explain any variation in the faunal composition among sites. 4. The abilities of the macroinvertebrates to predict the lake trophic state, expressed as log ([Chl a]), were explored by means of weighted averaging (WA) regression and calibration. Two tolerance-weighted WA models using inverse and classical regression for deshrinking are presented. The models were assessed by the root mean square error (RMSE) of prediction, using bootstrapping as cross validation, and by the correlation between observed and inferred log ([Chl a]). The model using inverse deshrinking had a RMSE_boot = 0.41 and r² = 0.63. By using classical regression, the predictability in the ends of the gradient was improved but the RMSE increased: RMSE_boot = 0.46. 5. Although the factors determining faunal distribution patterns in the Danish lowland lakes were highly multivariate and difficult to disentangle, it seems reasonable to use the WA estimated species optima and tolerances to [Chl a] in a bio-assessment model.     

Determining diatom ecotones and their relationship to terrestrial ecoregion designations in the central Canadian Arctic Islands

Ecotones are key areas for the detection of global change because many are predicted to move with shifts in climate. Prince of Wales Island, in the Canadian Arctic Archipelago, spans the transition between mid‐ to high‐Arctic ecoregions. We analyzed limnological variables and recent diatom assemblages from its lakes and ponds to determine if assemblages reflected this ecotone. Limnological gradients were short, and water chemistry explained 20.0% of diatom variance in a redundancy analysis (RDA), driven primarily by dissolved organic carbon, Ca and SO₄. Most taxa were small, benthic forms; key taxa such as planktonic Cyclotella species were restricted to the warmer, southern portion of the study area, while benthic Staurosirella were associated with larger, ice‐dominated lakes. Nonetheless, there were no significant changes in diatom assemblages across the mid‐ to high‐Arctic ecoregion boundary. We combined our data set with one from nearby Cornwallis Island to expand the study area and lengthen its environmental gradients. Within this expanded data set, 40.6% of the diatom variance was explained by a combination of water chemistry and geographic variables, and significant relationships were revealed between diatom distributions and key limnological variables, including pH, specific conductivity, and chl‐a. Using principal coordinates analysis, we estimated community turnover with latitude and applied piecewise linear regression to determine diatom ecotone positions. A pronounced transition was present between Prince of Wales Island and the colder, more northerly Cornwallis Island. These data will be important in detecting any future northward ecotone movement in response to predicted Arctic climate warming in this highly sensitive region.     

Diatom species–environment relationships and inference models from Isachsen, Ellef Ringnes Island, Canadian High Arctic

Surface sediment diatom assemblages were examined from 26 freshwater sites near Isachsen (78°47N, 103°32W), Ellef Ringnes Island, a region of diverse and atypical water chemistry for high arctic sites. One hundred and sixty eight diatom taxa were identified from these samples, over 50% of which had not previously been recorded in the Canadian High Arctic. Variations in diatom assemblages were related to changes in measured environmental variables using multivariate techniques. Canonical correspondence analysis (CCA) indicated that five variables contributed significantly to explaining patterns of diatom variation (i.e., COND, DIC, Mn, TPF, TPU). The first CCA axis (=0.44) was primarily controlled by conductivity-related variables, while CCA axis 2 (=0.21) was related to particulate concentrations. Diatom-based inference models were generated for the reconstruction of conductivity (RMSEP_jack=0.32, r²_jack=0.76) and pH (RMSEP_jack=0.40, r²_jack=0.69). The strengths of these models indicate that it will be possible to reliably infer past trends in conductivity and pH from diatom assemblages preserved in dated sediment cores from the Isachsen region.     

Evaluating diatom community composition in the absence of marked limnological gradients in the high Arctic: a surface sediment calibration set from Cornwallis Island (Nunavut,Canada)

Fossil diatom assemblages preserved within the sedimentary record in Arctic lakes provide the potential to reconstruct past changes in important limnological variables. During the summers of 1992 and 1993, we examined previously unstudied freshwater ecosystems on Cornwallis Island, Arctic Canada, with the specific objectives of (1) documenting the limnology and modern diatom assemblages from this region, and (2) determining which environmental variables most influence diatom species distributions. The Cornwallis Island study sites displayed the least amount of variance in measured water chemistry variables in comparison to nearly all of our labs’ previous freshwater surveys in the Arctic. The small limnological gradients precluded the development of a statistically robust diatom inference model, but perhaps more importantly, allowed us to explore variations in diatom composition in the absence of marked variations in water chemistry. Diatom species turnover was minimal, with the most common diatom taxa being Achnanthidium minutissima, Nitzschia perminuta, N. frustulum, with lesser percent abundances of Chaemaepinnularia soehrensis, Navicula chiarae, Psammothidium marginulata, and A. kryophila. A small number of study sites differed from the majority with respect to water chemistry (e.g., coastal sites with high specific conductivities) and habitat availability (e.g., ephemeral ponds with extensive moss habitats), and these sites had markedly different diatom assemblages. These data reinforce previous observations that water chemistry and other climate-related factors are the primary environmental controls influencing diatom distributions at high latitudes. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Diatoms and their relationship to environmental variables from lakes and ponds on Bathurst Island, Nunavut, Canadian High Arctic

Baseline and historical environmental data are sparse in the High Arctic, however diatom assemblages preserved in high arctic lake and pond sediment profiles can provide proxy data of past environmental changes. Ecological data, however, are still sparse. Diatom taxa preserved in the surficial sediment of lakes and ponds on Bathurst Island (75° 42 N, 97° 21 W), Nunavut, Canadian High Arctic, and their relationship to 34 environmental variables were examined using multivariate statistics. A total of 148 diatom taxa were identified from the surface sediments of 29 study sites. Five environmental variables, Fe³⁺, Total Phosphorus (Unfiltered) (TPU), Total Nitrogen (TN), Temperature (TEMP) and pH significantly (P0.05) accounted for most of the variation in the diatom assemblages. TN was also significantly correlated to other variables (e.g. TPF, DOC, POC). A CCA constrained to TN indicated that this variable best explained the species distributions, and a weighted-averaging (WA) model was developed to infer nutrient levels from the relative abundances of the 58 dominant taxa. Interestingly, previous limnological work indicated that nitrogen most likely limited algal production in some of these high arctic sites. This model may be used to quantitatively estimate past TN levels from diatom assemblages preserved in sediment cores from Bathurst Island, and may provide a means to track past environmental changes in the High Arctic.     

Development of diatom-based salinity models for paleoclimatic research from lakes in British Columbia (Canada)

Diatoms were identified and enumerated from the surface sediments of 65 lakes located on the Cariboo and Chilcotin Plateaux (British Columbia, Canada). These lakes span a large gradient in lakewater ionic concentration (fresh through hypersaline) and composition, as well as other physical/chemical variables. Almost all of the study lakes had higher salinities in the late-summer than in the spring. The lakes with spring salinities >8 g l^–1 showed the largest seasonal increases in salinity. Ionic composition was similar in the spring and late-summer for most lakes. Both ionic concentration (i.e. salinity) and composition were important environmental variables that could account for the different diatom floras in the lakes. Diatom assemblages characteristic of carbonate-dominated and sulfate-dominated waters were identified. Other variables such as water depth and phosphorus concentration were also important.The majority (87%) of diatom taxa had estimated salinity optima < 3 g l^–1 Halophilic diatom taxa had broader tolerances to salinity when compared to the fresh water taxa, however taxa with narrow and broad tolerances could be identified across the salinity gradient. Species diversity was weakly but significantly correlated to lakewater salinity (r ² = 0.18 to 0.3, P < 0.05).Salinity inference models were developed based on the relationship between the diatom assemblages and the spring, late-summer and average salinity. The correlations between the measured and diatominferred salinity, based on the spring (r = 0.95), late-summer (r = 0.94) and average (r = 0.95) salinity data, are high because there was an extremely strong correlation (r = 0.98) between the log transformed spring and late-summer measured salinities. These salinity reconstruction models provide a tool that can be used to infer past climatic changes as part of paleolimnological studies from appropriate closed-basin lakes in British Columbia.     

THE INFLUENCE OF SUBMERGED MACROPHYTES ON SEDIMENTARY DIATOM ASSEMBLAGES1

Submerged macrophytes are a central component of lake ecosystems; however, little is known regarding their long‐term response to environmental change. We have examined the potential of diatoms as indicators of past macrophyte biomass. We first sampled periphyton to determine whether habitat was a predictor of diatom assemblage. We then sampled 41 lakes in Quebec, Canada, to evaluate whether whole‐lake submerged macrophyte biomass (BiomEpiV) influenced surface sediment diatom assemblages. A multivariate regression tree (MRT) was used to construct a semiquantitative model to reconstruct past macrophyte biomass. We determined that periphytic diatom assemblages on macrophytes were significantly different from those on wood and rocks (ANOSIM R = 0.63, P < 0.01). A redundancy analysis (RDA) of the 41‐lake data set identified BiomEpiV as a significant (P < 0.05) variable in structuring sedimentary diatom assemblages. The MRT analysis classified the lakes into three groups. These groups were (A) high‐macrophyte, nutrient‐limited lakes (BiomEpiV ≥525 μg · L^?1; total phosphorus [TP] <35 μg · L^?1; 23 lakes); (B) low‐macrophyte, nutrient‐limited lakes (BiomEpiV <525 μg · L^?1; TP <35 μg · L^?1; 12 lakes); and (C) eutrophic lakes (TP ≥35 μg · L^?1; six lakes). A semiquantitative model correctly predicted the MRT group of the lake 71% of the time (P < 0.001). These results suggest that submerged macrophytes have a significant influence on diatom community structure and that sedimentary diatom assemblages can be used to infer past macrophyte abundance.