Minnesota Lakes Research Articles

Mobilizing place: Examining mobility, identity, and boundary in the politics of Asian carp

The movement of Asian carp species through connected river systems is a transboundary challenge for actors seeking to reduce their impacts. In this paper, I use “place” as an analytical lens, focusing on how concepts of mobility, identity, and boundary emerged as critical elements in the public framing of Asian carp problems and solutions. Drawing upon in-depth interviews as well as newspaper, policy, and media documents, I discuss how Asian carp policy discourses have centered on protecting Minnesota’s inland lakes or “Sky blue waters” from invasive Asian carp moving up the Mississippi River. Advocacy groups in Minnesota view Asian carp as threatening the place-based Minnesota “lake culture” and have used that discourse to motivate and mobilize public support. Consequently, management activities have focused on creating boundaries in the Mississippi River to deflect Asian carp movement into Minnesota lakes, which emerge as the places of primary concern and value. Ultimately, exploring the mobility of Asian carp and the ways actors mobilize in response illuminates the critical role of place in understanding the politics of Asian carp in the Upper Mississippi River.

Patterns and drivers of change in organic carbon burial across a diverse landscape: Insights from 116 Minnesota lakes

AbstractLakes may store globally significant quantities of organic carbon (OC) in their sediments, but the extent to which burial rates vary across space and time is not well described. Using 210Pb‐dated sediment cores, we explored patterns of OC burial in 116 lakes spanning several ecoregions and land use regimes in Minnesota, USA during the past 150–200 years. Rates for individual lakes (across all time periods) range from 3 to 204 g C m−2 yr−1 (median 33 g C m−2 yr−1) and show strong geographic separation in accordance with the degree of catchment disturbance and nutrient enrichment. Climate and basin morphometry exercise subordinate control over OC burial patterns, and diagenetic gradients introduce little bias to estimated temporal trends. Median burial rates in agricultural lakes exceed urban lakes and have increased fourfold since Euro‐American settlement. The greatest increase in OC burial occurred prior to the widespread adoption of industrial fertilizers, during an era of land clearance and farmland expansion. Northern boreal lakes, impacted by historical logging and limited cottage development yet comparatively undisturbed by human activity, bury OC at rates 3X lower than agricultural lakes and exhibit much smaller increases in OC burial. Scaling up modern OC burial estimates to the entire state, we find that Minnesota lakes annually store 0.40 Tg C in their sediments, equal to 1.5% of annual statewide CO2 emissions from fossil fuel combustion. During the period of Euro‐American settlement (circa 1860–2000), cumulative OC burial amounted to 36 Tg C, 40% of which can be attributed to anthropogenic enhancement.

Is the island biogeography model a poor predictor of biodiversity patterns in shallow lakes?

Summary The classic island biogeography model (IBM) predicts highest species richness in large, connected habitats due to colonisation and reduced risk of extinction. Promoting large, connected habitats has subsequently become a common theme in conservation biology. However, the IBM does not account for direct and indirect interactions among species. For example, planktivorous and benthivorous fish may reduce biodiversity in shallow lakes by inducing shifts to a turbid‐water lake state with low habitat complexity. We assessed relationships between species richness, landscape features, fish biomass and lake state in 104 shallow lakes in Minnesota, U.S.A. First, we tested whether lake size and connectivity influenced species richness of fish and biomass of planktivores and benthivores (fish biomass). We subsequently tested whether fish biomass affected the probability that lakes were in turbid versus clear‐water states. Finally, we tested whether species richness of macrophytes and taxon richness of aquatic invertebrates showed stronger relationships with lake size and connectivity or with fish biomass and lake state. Fish richness and biomass both increased with lake size and were higher in connected basins. Fish biomass, in turn, increased the probability that lakes would be turbid. In contrast, macrophyte and invertebrate richness were unrelated to lake size or connectivity. Instead, macrophyte richness was best predicted by lake state, while invertebrate richness was predicted by lake state and fish biomass. Richness of both macrophytes and invertebrates was higher in clear lakes, and invertebrate richness was inversely related to fish biomass. Our results indicate the IBM poorly explains the diversity of macrophytes and invertebrates in shallow lakes, with diversity more strongly driven by biotic interactions and influences associated with fish. We suggest that ecological implications of increased connectivity and lake size should be considered in future conservation strategies for shallow lakes.

Changes in Relative Abundance of Several Minnesota Fishes from 1970 to 2013

AbstractLarge‐scale ecological stressors such as climate change, habitat degradation, and invasive species increasingly affect fisheries across watershed or political boundaries within which they are typically managed. Though broad‐scale analyses of trends in fish populations and changes in aquatic communities can be useful for detecting the effects of these stressors, they are rarely possible due to a lack of long‐term data sets. The Minnesota Department of Natural Resources has used gill nets and trap nets to monitor fish communities in lakes since 1940, and the techniques have been standardized since 1970. We applied a random‐coefficient mixed‐effects model to mean CPUE data to describe statewide and local population trends of eight key fish species (Black Crappie Pomoxis nigromaculatus, Bluegill Lepomis macrochirus, Largemouth Bass Micropterus salmoides, Northern Pike Esox lucius, Smallmouth Bass M. dolomieu, Walleye Sander vitreus, White Sucker Catostomus commersonii, and Yellow Perch Perca flavescens) sampled from 1970 to 2013 in 119–1,499 lakes. We found increasing statewide trends in gill‐net CPUE of Black Crappie, Bluegill, Largemouth Bass, Smallmouth Bass, Northern Pike, and Walleye and decreasing trends for Yellow Perch and White Sucker. Statewide trends in average trap‐net CPUE were positive for Bluegill, negative for Black Crappie, Walleye, White Sucker, and Yellow Perch, and stable for Largemouth Bass, Northern Pike, and Smallmouth Bass. We compared the CPUE trends across lake types developed for Minnesota lakes and found some differences, especially in medium and large mesotrophic lakes, suggesting that some species display changes in abundance more in those lakes types than in others. This study quantifies temporal changes in Minnesota's fish community structure and can help managers focus future research and management activities in the state and in similar waters across central North America.Received May 14, 2014; accepted September 9, 2014

Analysis of macrophyte indicator variation as a function of sampling, temporal, and stressor effects

Biological indicators that signal changes in lake condition are essential tools for guiding resource management decisions. Macrophyte-based indicators have traditionally been selected and evaluated in the context of nutrient-based stressors, although the need to evaluate indicators that are sensitive to climate stressors has been increasingly relevant. Moreover, indicators should ideally exhibit minimal sampling variation and have low natural temporal variation so there is high power to detect changes in the mean value over time. Eight macrophyte indicators were estimated in 23 Minnesota (USA) lakes using four years of repeated surveys to estimate sampling and temporal variation, response to development (phosphorus concentration) and climate stress (annual growing degree days), and power to detect significant change at various annual sampling intervals. Indicators included a macrophyte index of biotic integrity, floristic quality index, maximum depth of growth, total species richness, common species richness, mean richness, and frequency occurrence of rooted species and Chara sp. Overall, regression and smoothed additive models indicated significant relationships of indicators to total lake phosphorus and mean annual growing degree days. The macrophyte index of biotic integrity, floristic quality index, and the frequency rooted species had minimal sampling variation in this study, were responsive to development or climate stress, and had low annual variation (coefficients of variation 0.08, 0.10, and 0.19, respectively) resulting in high to moderate power (>50%) for detecting significant change over a 20 year period. Results from these analyses will facilitate the use of precise and powerful indicators that respond to stressors that are of concern for the management of freshwater glacial lakes.

Summary of microcystin concentrations in Minnesota lakes

Minnesota has compiled a fairly extensive database on the cyanobacterial toxin microcystin (MC). These data have been acquired through targeted lake studies, incident-based investigations, and stratified random surveys as a part of the US EPA's 2007 and 2012 National Lakes Assessments. This paper updates previous efforts and provides a statistical summary of all Minnesota MC data collected by the Minnesota Pollution Control Agency from 2004 to 2012. These data provide a basis for (1) describing the influence sampling regimes have on reported MC values, (2) describing regional patterns in MC across 3 distinct Minnesota ecoregions, and (3) describing the relative risk of encountering elevated MC with respect to sampling regimes. This summary places MC concentrations in perspective for Minnesota and also provides a comparative database, which other entities that collect MC may use to place their results in perspective.

Application of neural networks to quantify the utility of indices of biotic integrity for biological monitoring

Indices of Biotic Integrity (IBIs) or multimetric indices have been developed as an approach for monitoring and evaluating biological condition of aquatic organisms. Quantitative evaluations of IBIs to determine whether they can explicitly link environmental condition with anthropogenic activities are needed to effectively use them in management. Analytical approaches using supervised neural networks are potentially powerful techniques to evaluate IBIs. The goal of this study was to evaluate the use of neural networks to identify ecosystem characteristics related to IBI response and to explicitly quantify relationships between variables using sensitivity analyses. An aquatic macrophyte-based IBI developed for Minnesota lakes was used as an example. The study was particularly interested in the usefulness of neural networks to highlight key predictors of IBI performance and to be used as a technique to evaluate multimetric index performance in other systems or regions. Neural networks made accurate predictions of overall IBI scores using an independent dataset, whereas predictive performance of the models varied for individual metrics. Bootstrap analyses to evaluate the effects of different training data on model performance indicated that predictions were highly sensitive to the training data. More conventional modeling techniques, such as multiple regression, performed similarly in predicting IBI scores, although diagnostic tools developed for neural networks provided novel insight into variables influencing IBI response. We suggest that neural networks have the ability to quantify ecological relationships that affect biotic integrity, but the statistical uncertainty associated with multimetric indices may limit the use of predictive models to infer causation. Accordingly, the statistical properties of multimetric indices should be carefully evaluated during index development, with specific attention given to the diagnostic capabilities of individual metrics.

Common Loon (Gavia immer) Nesting Habitat Models for North-Central Minnesota Lakes

Abstract. Shoreline attributes and extensive field surveys of aquatic vegetation and animal presence were used to determine probabilities of Common Loon (Gavia immer) nesting for segments of lakeshore on 35 lakes in north-central Minnesota. Model development used both a general linear mixed model and random forest classifier approach. The resulting nesting habitat models were used to predict nesting sites for a small set of independent lakes. Shoreline segments with low mean fetch and littoral slope, fewer developed shoreline parcels, and higher aquatic plant richness had higher probabilities of nesting. In addition, significantly more nesting sites were on islands than on mainland shoreline segments. The locations of predicted nesting sites on the independent lake set compared favorably to the locations of observed nests. The ability to predict suitable Common Loon nesting sites should lead to the greater protection or restoration of these valuable areas and enhance conservation efforts across the state.

Vulnerability to harvest by anglers differs across climate, productivity, and diversity clines

We contrast catchability of walleye (Sander vitreus) and northern pike (Esox lucius) populations with angling fisheries across regions that differ twofold in growing-degree-days and productivity and sixfold in fish diversity. Populations of both species in Alberta, Wisconsin, Minnesota, and Oneida Lake, New York, had density-dependent catchability with approximately tenfold higher catchability in Alberta than in the other regions when density was controlled for. There is no evidence that the higher catchability estimates for Alberta walleye and northern pike are due to differential spatial distributions, enhanced hook avoidance due to catch and release or to differential size structure of the populations, or to differences in harvest regulations. We argue that the most likely explanation for the tenfold higher catchability is increased hunger resulting in enhanced foraging activity in the region with a substantially shorter growing season, lower prey productivity, and lower prey community diversity. Regardless of the proximate causes, higher catchability of fish harvested in recreational fisheries in Alberta substantially increases their vulnerability to overharvest and collapse if angling effort is unabated.

Simulation and Validation of Cisco Habitat in Minnesota Lakes Using the Lethal-Niche-Boundary Curve

Fish survival in lakes is strongly influenced by water temperature and dissolved oxygen (DO) concentration. A one-dimensional (vertical) lake water quality model MINLAKE 2012 was calibrated in 23 Minnesota lakes and used to simulate daily water temperature and DO concentrations in 36 representative lake types under past (1992–2008) climate conditions and a future climate scenario (MIROC 3.2). The 36 representative Minnesota lake types were developed based on three maximum depths (H max = 4, 13, and 24 m), three surface areas (As = 0.2, 1.7, 10 km 2 ), and four Secchi depths (SD = 1.2, 2.5, 4.5, and 7 m, from eutrophic to oligotrophic lake). A fish habitat model using the lethal-niche

Persistence of clear-water, shallow-lake ecosystems: the role of protected areas and stable aquatic food webs

The roles of both landscape alteration and in-lake processes need to be considered in conservation strategies for shallow lakes in the prairie regions of North America. Here we focus on shallow lakes in west-central Minnesota, USA, highlighting the long-term ecological history and response to known landscape changes of a clear-water, macrophyte-dominated, shallow lake. Contemporary limnological data suggest the aquatic ecosystem has been very stable and fishless for the last ~15 years. Sediment proxies for primary production and ecological change confirm that a stable ecosystem likely prevailed for the last ~200 years. However, sedimentary indicators of catchment erosion detail a distinct response to land-use change during the conversion of native grassland to agricultural land, and following establishment of a protected waterfowl production area (WPA) around the lake. Post-WPA, the rate of sediment accrual decreased dramatically within 5–10 years and sources of organic matter were similar to those of the pre-settlement period. The aquatic ecosystem has been able to withstand nutrient enrichment and allochthonous inputs because stable trophic interactions have likely been in place for more than 200 years. We conclude that lack of hydrologic connectivity and isolated, small catchments are important factors in the promotion of clear-water shallow lake ecosystems, mainly because they prevent colonization by fish and associated ecological consequences. This study highlights the importance of managing both the landscape and in-lake processes to maintain stable, clear-water, shallow lakes.

Geospatial and Temporal Analysis of a 20‐Year Record of Landsat‐Based Water Clarity in Minnesota's 10,000 Lakes

AbstractA large 20‐year database on water clarity for all Minnesota lakes ≥8 ha was analyzed statistically for spatial distributions, temporal trends, and relationships with in‐lake and watershed factors that potentially affect lake clarity. The database includes Landsat‐based water clarity estimates expressed in terms of Secchi depth (SDLandsat), an integrative measure of water quality, for more than 10,500 lakes for time periods centered around 1985, 1990, 1995, 2000, and 2005. Minnesota lake clarity is lower (more turbid) in the south and southwest and clearer in the north and northeast; this pattern is evident at the levels of individual lakes and ecoregions. Temporal trends in clarity were detected in ~11% of the lakes: 4.6% had improving clarity and 6.2% had decreasing clarity. Ecoregions in southern and western Minnesota, where agriculture is the predominant land use, had higher percentages of lakes with decreasing clarity than the rest of the state, and small and shallow lakes had higher percentages of decreasing clarity trends than large and deep lakes. The mean SDLandsat statewide remained stable from 1985 to 2005 but decreased in ecoregions dominated by agricultural land use. Deep lakes had higher clarity than shallow lakes statewide and for lakes grouped by land cover. SDLandsat decreased as the percentage of agriculture and/or urban area increased at county and catchment levels and it increased with increasing forested land.

Depths and Thermal Habitat Used by Large versus Small Northern Pike in Three Minnesota Lakes

AbstractWe monitored depths and temperatures used by large (>71‐cm) versus small Northern Pike Esox lucius in three north‐central Minnesota lakes with either acoustic telemetry or archival tags. Individual Northern Pike demonstrated flexibility in depths used within a season and between years. The fish had some tolerance for low levels of dissolved oxygen (<3 mg/L), but depth selection was generally constrained by low dissolved oxygen in summer and winter. The fish more fully exploited all available depths during winter and thermal turnover periods. During July and August, large Northern Pike tended to follow the thermocline into cooler water as upper water layers warmed. Selection ratios indicated that large Northern Pike preferred water temperatures of 16–21°C during August when temperatures up to 28°C were available. In two lakes providing dense overhead cover from water lilies in shallow water, small Northern Pike used warmer, shallower water compared with large fish during summer. In a third lake providing no such cover, small fish were more often in deeper, cooler water. For small Northern Pike, temperature seemed to be a secondary habitat consideration behind the presence of shallow vegetated cover. This study provided detailed temperature selection information that will be useful when considering temperature as an ecological resource for different sizes of Northern Pike.

Land-use change, not climate, controls organic carbon burial in lakes

Lakes are a central component of the carbon cycle, both mineralizing terrestrially derived organic matter and storing substantial amounts of organic carbon (OC) in their sediments. However, the rates and controls on OC burial by lakes remain uncertain, as do the possible effects of future global change processes. To address these issues, we derived OC burial rates in (210)Pb-dated sediment cores from 116 small Minnesota lakes that cover major climate and land-use gradients. Rates for individual lakes presently range from 7 to 127 g C m(-2) yr(-1) and have increased by up to a factor of 8 since Euro-American settlement (mean increase: 2.8×). Mean pre-disturbance OC burial rates were similar (14-22 g C m(-2) yr(-1)) across all land-cover categories (prairie, mixed deciduous and boreal forest), indicating minimal effect of the regional temperature gradient (approx. 4 °C) on background carbon burial. The relationship between modern OC burial rates and temperature was also not significant after removal of the effect of total phosphorus. Contemporary burial rates were strongly correlated with lake-water nutrients and the extent of agricultural land cover in the catchment. Increased OC burial, documented even in relatively undisturbed boreal lake ecosystems, indicates a possible role for atmospheric nitrogen deposition. Our results suggest that globally, future land-cover change, intensification of agriculture and associated nutrient loading together with atmospheric N-deposition will enhance OC sequestration by lakes.

Between- and within-lake responses of macrophyte richness metrics to shoreline development

Aquatic habitat in littoral environments can be affected by residential development of shoreline areas. We evaluated the relationship between macrophyte richness metrics and shoreline development to quantify indicator response at 2 spatial scales for Minnesota lakes. First, the response of total, submersed, and sensitive species to shoreline development was evaluated within lakes to quantify macrophyte response as a function of distance to the nearest dock. Within-lake analyses using generalized linear mixed models focused on 3 lakes of comparable size with a minimal influence of watershed land use. Survey points farther from docks had higher total species richness and presence of species sensitive to disturbance. Second, between-lake effects of shoreline development on total, submersed, emergent-floating, and sensitive species were evaluated for 1444 lakes. Generalized linear models were developed for all lakes and stratified subsets to control for lake depth and watershed land use. Between-lake analyses indicated a clear response of macrophyte richness metrics to increasing shoreline development, such that fewer emergent-floating and sensitive species were correlated with increasing density of docks. These trends were particularly evident for deeper lakes with lower watershed development. Our results provide further evidence that shoreline development is associated with degraded aquatic habitat, particularly by illustrating the response of macrophyte richness metrics across multiple lake types and different spatial scales.

A comparison of survey methods to evaluate macrophyte index of biotic integrity performance in Minnesota lakes

Aquatic macrophytes shape trophic web dynamics, provide food and refuge for macroinvertebrates and fish, and increase nutrient retention, sediment stabilization, and water clarity. Macrophytes are well-suited as indicators of ecological health because they are immobile, relatively easy to sample and identify, and respond to anthropogenic disturbance on an ecological time scale. Aquatic plant monitoring programs can provide valuable information to water resource managers, especially in conjunction with macrophyte-based indices of biotic integrity (IBI). However, there are several current sampling designs and the precision of IBI scores has not been evaluated across different surveys. We evaluated the performance of the Minnesota macrophyte-based IBI for two survey designs; a point intercept (PI) survey and a belt transect (BT) survey. PI surveys are time intensive, especially on large lakes, whereas BT are less time intensive and have been used historically in Minnesota. Our objectives were to compare the PI surveys with BT surveys on the same lakes, and to modify the BT survey (MT survey) to improve information obtained from BT surveys. BT surveys consistently overestimated IBI scores compared to the PI method (t=6.268, df=60, p<0.001). Overall IBI scores calculated from MT surveys differed significantly from PI scores, but on average, MT surveys predicted scores only 3% lower than PI scores. Implementation of the Minnesota macrophyte-based IBI through the adoption of the MT survey approach would improve sampling efficiency and enable widespread documentation of the effects of landscape change, shifts in hydrologic regimes, and other anthropogenic activities on the integrity of lacustrine systems.

Governing the ice. Ice fishing villages on Lake Mille Lacs and the creation of environmental governance institutions

We identify four choice dimensions that determine the configuration and evolution of governance: formal-informal institutions, network-central steering, local-scientific knowledge and representation-participation. Choices on one dimension affect choices on the other dimensions, which naturally leads to historical dependency. We integrate these insights in a model of governance evolution that revolves around actor/institution configurations and power/knowledge configurations. In a case study of ice fishing villages on Minnesota's Lake Mille Lacs, we investigate one specific set of couplings between the choice dimensions. As we can study the local ice fishing tradition from its very beginning, the evolutionary paths of technology and institutions provide insights into how choices were made along the different dimensions and how they interacted. The case study illustrates how to apply the model, but also contributes to its further development as it draws attention to possible extensions: concepts of scale and identity.

Influence of lake water pH and alkalinity on the distribution of core and intact polar branched glycerol dialkyl glycerol tetraethers (GDGTs) in lakes

Branched glycerol dialkyl glycerol tetraethers (GDGTs) are bacterial membrane lipids, ubiquitously present in soils and peat bogs, as well as in rivers, lakes and lake sediments. Their distribution in soil is controlled mainly by pH and mean annual air temperature, but the controls on their distribution in lake sediments are less well understood. Several studies have found a relationship between the distribution of branched GDGTs in lake sediments and average lake water pH, suggesting an aquatic source for them, besides that for soil transported to the lake via erosion. We sampled the surface water suspended particulate matter (SPM) from 23 lakes in Minnesota and Iowa (USA), that vary widely in pH, alkalinity and trophic state. The SPM was analyzed for the concentration and distributions of core lipid (presumed fossil origin) and intact polar lipid (IPL, presumed to derive from living cells) branched GDGTs. The presence of substantial amounts (18–48%) of IPL-derived branched GDGTs suggests that branched GDGTs are likely of autochthonous origin. Temperature estimates based on their distribution using lake-specific calibrations agree reasonably with water temperature at time of sampling and average air temperature of the season of sampling. Importantly, a strong correlation between the distribution of branched GDGTs and lake water pH was found (r2 0.72), in agreement with a predominant in situ production. An stronger correlation was found with lake water alkalinity (r2 0.83), although the underlying mechanism that controls the relationship is not understood. Our results raise the potential for reconstructing pH/alkalinity of past lake environments, which could provide important knowledge on past developments in lake water chemistry.

A new dimension to sediment magnetism: Charting the spatial variability of magnetic properties across lake basins

We have investigated the variability of the magnetic properties of surface sediments across eight Minnesota lake basins. The measured magnetic properties are controlled by the competing fluxes of allochthonous and autochthonous magnetic particles, and differ according to location in the basin. Shoreline sediments are dominated by detrital magnetic particles, whereas littoral and profundal sediments are characterized by a combination of bacterial magnetosomes and detrital particles. The position of the oxic–anoxic interface, which may occur in the water or within the sediment column, controls the depth at which living magnetotactic bacteria occur, and determines the degree of preservation of their magnetosome chains in the surface sediment. The preservation potential of undisturbed chains is higher for bacterial magnetite formed at the top of the sediment column in the littoral area than for magnetosomes originating in the water column in the profundal area. Bacterial magnetite in the profundal facies will contain a higher proportion of chains collapsed during settlement through the water column to the lake bottom. This process increases the fraction of interacting magnetosomes, which in turn artificially lowers the ARM ratio (χARM/IRM), which ceases to be a reliable grain size indicator in the profundal environment. Our results indicate that a holistic approach to interpreting limnologically-derived paleoecological data should be employed. Specifically, a thorough understanding of evolving and interrelated factors such as basin morphology and limnologic conditions is crucial for a more confident interpretation of the sedimentary record in terms of environmental conditions at the time of sediment deposition.

Assessing the performance of a diatom transfer function on four Minnesota lake sediment cores: effects of training set size and sample age

Paleolimnological information is often extracted from diatom records using weighted averaging calibration and regression techniques. Larger calibration sample sets yield better inferences because they better characterize the environmental characteristics and species assemblages of the sample region. To optimize inferred information from fossil assemblages, however, it is worth knowing if fewer calibration samples can be used. Furthermore, confidence in environmental reconstructions is greater if we consider the relative importance of (A) similarity between fossil and calibration assemblages and (B) how well fossil taxa respond to the environmental variable of interest. We examine these issues using ~200-year sediment profiles from four Minnesota lakes and a 145-lake surface sediment training set calibrated for total phosphorus (TP). Training set sample sizes ranging from 10 to 145 were created through random sample selection, and models based on these training sets were used to calculate diatom-inferred (DI) TP data from fossil samples. Relationships between DI-TP variability and sample size were used to determine the minimum sample size needed to optimize the model for paleo-reconstruction. Similarly, similarities between fossil and modern assemblages were calculated for each size training set. Finally, fossil and modern assemblages were compared to determine whether older fossil samples had poorer similarity with modern analogs. More than 50–80 samples, depending on lake, were needed to stabilize variability in DI-TP results, and >110 training set samples were needed to minimize modern-fossil assemblage dissimilarities. Dissimilarities appeared to increase with sample age, but only one of the four studied cores displayed a significant trend. We have two recommendations for future studies: (1) be cautious when dealing with smaller training sets, especially if they are used to interpret older fossil assemblages and (2) understand how well fossil taxa are attuned to the variable of interest, as it is critical to evaluating the quality of the diatom-inferred data.