Stable Oxygen and Carbon Isotopic Compositions of Lates stappersii Otoliths from Lake Tanganyika,East Africa

In this study, we measured growth trends in oxygen and carbon isotopic ratios in whole sagittal otoliths from three adult centropomid fish (Lates stappersii) from each of three sub-basins of Lake Tanganyika, East Africa. Sampling density was 20 to 50 samples per otolith. Both δ¹⁸O and δ¹³C values increase with age. The δ¹⁸O data suggest that otoliths were precipitated near the expected equilibrium with the ambient environment (ca. +3.5‰) and support a migration pattern from surface waters during larval stages to deeper waters (40 to 80 m) for mature fish. Relatively high δ¹⁸O values in the southern sub-basin are consistent with cooler temperatures in the region during seasonal upwelling. The δ13C increase from otolith core to edge is large (up to 4‰) and is interpreted as due to ontogenetic changes in diet and contributions from a decrease in the proportion of respired CO₂ incorporated into otolith carbonate as metabolic rates of the fish dropped with maturity. The data seem to successfully reveal life strategy and migration patterns of L. stappersii, document regional differences in lake conditions, and provide a record of temperature within the water column during which the fish lived. Higher resolution studies and analyses of historical samples could be used to constrain modern and past growth patterns, and to reconstruct past temperature gradients and productivity patterns in the lake.     

Machine-learning paradigms for selecting ecologically significant input variables

Harmful algal blooms, which are considered a serious environmental problem nowadays, occur in coastal waters in many parts of the world. They cause acute ecological damage and ensuing economic losses, due to fish kills and shellfish poisoning as well as public health threats posed by toxic blooms. Recently, data-driven models including machine-learning (ML) techniques have been employed to mimic dynamics of algal blooms. One of the most important steps in the application of a ML technique is the selection of significant model input variables. In the present paper, we use two extensively used ML techniques, artificial neural networks (ANN) and genetic programming (GP) for selecting the significant input variables. The efficacy of these techniques is first demonstrated on a test problem with known dependence and then they are applied to a real-world case study of water quality data from Tolo Harbour, Hong Kong. These ML techniques overcome some of the limitations of the currently used techniques for input variable selection, a review of which is also presented. The interpretation of the weights of the trained ANN and the GP evolved equations demonstrate their ability to identify the ecologically significant variables precisely. The significant variables suggested by the ML techniques also indicate chlorophyll-a (Chl-a) itself to be the most significant input in predicting the algal blooms, suggesting an auto-regressive nature or persistence in the algal bloom dynamics, which may be related to the long flushing time in the semi-enclosed coastal waters. The study also confirms the previous understanding that the algal blooms in coastal waters of Hong Kong often occur with a life cycle of the order of 1–2 weeks.     

Evaluation of landscape and instream modeling to predict watershed nutrient yields

The project goal was to loosely couple the SWAT model and the QUAL2E model and compare their combined ability to predict total phosphorus (TP) and NO₃-N plus NO₂-N yields to the ability of the SWAT model with its completely coupled water quality components to predict TP and NO₃-N plus NO₂-N yields from War Eagle Creek watershed in Northwest Arkansas. Model predictions were compared using a statistical approach to identify significant differences between the two modeling methods. Results from two variations of the Pearson product-moment correlation (p < 0.05) indicated that correlation coefficients and regression slopes for the two data sets were not significantly different. This implies that neither modeling method was significantly better in predicting monthly TP and NO₃-N plus NO₂-N yields from the watershed. Additionally, no significant differences were present between predicted outputs of the SWAT model with instream components active compared with when instream components were inactive, indicating a need for further testing and refinement of the SWAT algorithms simulating instream processes. We can further infer that the instream processes available in SWAT may not be enhancing its predictive abilities as far as simulating instream components.     

Groundwater-surface water interactions and agricultural nutrient transport in a Great Lakes clay plain system

Nutrient export from agricultural land to surface waters is a significant environmental concern within the Great Lakes Basin (GLB). A field-based watershed-scale study was completed to investigate spatial and temporal variations of phosphorus and nitrate to assess nutrient transport pathways and groundwater-surface water interactions in an agriculturally dominated clay plain system. This was conducted in the 127 km² Upper Parkhill Watershed, near Lake Huron in southwestern Ontario, Canada. Data collection occurred from June 2018 to May 2019 via continuous sensor deployment and discrete sampling of stream water, groundwater, hyporheic zone, and tile drainage water. Samples were analyzed for various nutrient species (total, total dissolved, soluble reactive, and particulate phosphorus, and nitrate-N) to examine the hydrological dynamics of principal transport pathways of agriculturally-derived nutrients. Total phosphorus and nitrate concentrations in stream water ranged from 0.007 to 0.324 mg/L and 0.32 to 13.13 mg NO₃^?-N/L, respectively. Tile drainage water total phosphorous concentrations varied from 0.006 to 0.066 mg/L. Groundwater total dissolved phosphorus concentrations ranged from <0.003 to 0.085 mg/L. Transport of phosphorus through tile drainage was observed to be greater than through groundwater over the study period. No distinct relationship was observed between nutrient concentrations in the hyporheic zone and the vertical hydraulic gradient within this zone in the studied stream reach. Preliminary correlations were discerned between water quality observations and recognized land management practices. Given the elevated stream nutrient concentrations, these results are consequential for the continual improvement of strategies and programs devised to conserve water resources within the GLB.     

Towards improving an Area of Concern: Main-channel habitat rehabilitation priorities for the Maumee River

The Maumee River watershed in the Laurentian Great Lakes Basin has been impacted by decades of pollution and habitat modification due to human settlement and development. As such, the lower 35 km of the Maumee River and several smaller adjacent watersheds comprising over 2000 km² were designated the Maumee Area of Concern (AOC) under the revised Great Lakes Water Quality Agreement in 1987. As part of pre-rehabilitation assessments in the Maumee AOC, we assessed fish and invertebrate communities in river km 24–11 of the Maumee River to identify: 1) areas that exhibit the highest biodiversity, 2) habitat characteristics associated with high biodiversity areas, 3) areas in need of protection from further degradation, and 4) areas that could feasibly be rehabilitated to increase biodiversity. Based on benthic trawl data, shallow water habitats surrounding large island complexes had the highest fish diversity and catch per unit effort (CPUE). Electrofishing displayed similar fish diversity and CPUE patterns across habitat types early in the study but yielded no discernable fish diversity or CPUE patterns towards the end of our study. Although highly variable among study sites, macroinvertebrate density was greatest in shallow water habitats <2.5 m and around large island complexes. Our results provide valuable baseline data that could act as a foundation for developing rehabilitation strategies in the lower Maumee River and for assessing the effectiveness of future aquatic habitat rehabilitation projects. In addition to increasing in-channel habitat, watershed-scale improvements of water quality might be necessary to ensure rehabilitation strategies are successful.     

Foredune dynamics at a Lake Michigan site during rising and high lake levels

On Great Lakes dunes, the link between foredune dynamics and coastal processes is seen in dune responses to changing lake levels. This paper investigates foredune dynamics during a recent period of rising and high lake levels. The study location was an active foredune in P.J. Hoffmaster State Park on the east coast of Lake Michigan, where field data were collected from 2000 through the final destruction of the foredune by wave removal in November 2019. Foredune dynamics were studied with erosion pins, direct observations, photographs, mapping, and on-site wind measurements. Regional climate and lake-level data were obtained from established data collection programs. The response of the foredune to rising lake levels was compared to several models of foredune behavior. During the study, the Lake Michigan-Huron level rose 1.89 m from January 2013 to July 2020. After an early transitional period, foredune activity was characterized by scarp retreat (4–19 m per year) and dune narrowing from 2014 to 2019. When the foredune completely disappeared in November 2019, erosion/scarping began on the next landward dune. The foredune activity fits Olson’s (1958) model for foredune growth and erosion through lake-level cycles. The foredune migration predicted by the revised Davidson-Arnott (2021) model of foredune response to relative water level rise did not occur, most likely because the rate of lake-level rise was too high. The six years of foredune narrowing before wave erosion started affecting the next landward dune represent a time-lag in Lake Michigan dune history models of increased dune activity during high lake-level stands.     

Wild juvenile salmonid abundance in Wisconsin tributaries indicates limited contributions to Lake Michigan fisheries

Natural reproduction of salmonids occurs in many Lake Michigan tributaries, yet little is known about abundance and the potential contribution of wild fish hatching in Wisconsin tributaries. The objectives of our study were to determine if: 1) abundance of wild juvenile salmonids (primarily adfluvial rainbow trout, Oncorhynchus mykiss, referred to as steelhead) varied among selected Wisconsin streams based on available spawning and age-0 habitat; 2) stream temperature regimes could limit survival of juvenile salmonids, and 3) wild juvenile salmonids outmigrate from Wisconsin tributaries into Lake Michigan or larger tributaries. In 2016 and 2017, juvenile salmonid abundance was estimated in six Wisconsin tributaries to Lake Michigan by multiple-pass depletion sampling using backpack electrofishing. Habitat assessments included steelhead redd surveys, age-0 habitat surveys, and stream temperatures were monitored using in-stream loggers. Passive integrated transponder (PIT) tagging and PIT antennas were used to detect outmigration from three streams (Willow, Stony and Hibbard creeks). Population estimates for individual streams ranged from 75 to 2276 for juvenile steelhead and from 0 to 243 for juvenile coho salmon, Oncorhynchus kisutch. No correlation was detected between juvenile steelhead abundance and quality age-0 habitat. Stream temperatures rarely exceeded the thermal limit for steelhead (27 °C). Outmigration rates for three streams ranged from 0.6% to 3.1%, but these estimates were considered minimum values. Low abundance of wild juvenile steelhead and coho salmon alone suggest that the contributions of these tributaries to Lake Michigan fisheries are likely small. Furthermore, relying on returns of wild steelhead produced in these streams is probably insufficient to maintain stream fisheries.     

Building a research network to better understand climate governance in the Great Lakes

Climate-driven disturbances threaten the sustainability of coastal communities in the Great Lakes Basin. Because such disturbances are unpredictable, their magnitude, number and intensity are changing, and they occur at varying temporal and spatial scales. Consequently, communities struggle to respond in effective ways. The expected intensification of climate-driven disturbances will require that community capacity and governance structures match the spatial and temporal scales of these disturbances, as the most sustainable social and economic systems will be those that can respond at similar frequencies to key natural system drivers. The Climate Governance Variability in the Great Lakes Research Coordination Network (CGVG-RCN) was recently established to address questions about the relationship between climate-driven disturbances and community response. The objective of this short communication is to introduce the ideas behind the CGVG-RCN, outline its goals, and facilitate engagements and collaboration with social and natural scientists interested in social-ecological systems in the Great Lakes Basin.     

Interannual and intra-annual cycles of satellite-derived chlorophyll-a concentrations in the Caspian Sea

Variability of phytoplankton in the Caspian Sea (CS) is related to the complex orography, the variety of physical and chemical processes, and climatic indices. Here, interannual and intra-annual variability of satellite-derived chlorophyll-a concentrations (Chl-a) were studied using wavelet analysis during 2002–2019 in different regions of the CS. Self-Organizing Maps (SOM) analysis performed to classify the CS into the areas of similar variability of satellite-derived Chl-a. Wavelet spectral analysis showed that the Chl-a variability regulated by four primary periodic cycles: 0.5-year, 1-year, 2 to 2.5-year, and 3 to 5-year. The 0.5 and 1-year wavelet cycles mostly depicted the intensity of seasonality patterns. The 2–2.5-year and 3–5-year cycles of Chl-a showed non-stationary coherence with corresponding low-frequency cycles of NAO and ENSO. The 3–5-year wavelet amplitudes of Chl-a strongly correlated with NAO and ENSO in the southern CS. Weak correlations of 2–2.5-year cycle wavelet amplitudes of Chl-a with NAO and ENSO suggested that variations do not always directly translate to a biological response. A negative anomaly in the Chl-a autumn peaks observed during 2011–2016 in the middle and southern CS, when NAO phases were persistently positive. The interannual variations of summer peaks in the northern CS, and autumn peaks in the middle and southern CS were broadly related to the precipitation. SST and wind stress. Moreover, it was shown that the Volga discharge has a significant influence on Chl-a variability in the northern CS.