Predicting the Degree of Dissolved Oxygen Using Three Types of Multi-Layer Perceptron-Based Artificial Neural Networks

Predicting the level of dissolved oxygen (DO) is an important issue ensuring the sustainability of the inhabitants of a river. A prediction model can predict the DO level using a historical dataset with regard to water temperature, pH, and specific conductance for a given river. The model can be built using sophisticated computational procedures such as multi-layer perceptron-based artificial neural networks. Different types of networks can be constructed for this purpose. In this study, the authors constructed three networks, namely, multi-verse optimizer (MVO), black hole algorithm (BHA), and shuffled complex evolution (SCE). The networks were trained using the datasets collected from the Klamath River Station, Oregon, USA, for the period 2015–2018. We found that the trained networks could predict the DO level of 2019. We also found that both BHA- and SCE-based networks could predict the level of DO using a relatively simple configuration compared to that of MVO. From the viewpoints of absolute errors and Pearson’s correlation coefficient, MVO- and SCE-based networks performed better than BHA-based networks. In synopsis, the authors recommend MVO- and MLP-based artificial neural networks for predicting the DO level of a river.

Temperature dynamics in Klamath River off-channel floodplain restoration sites depend on location within the watershed

The mid-Klamath River is heavily impacted by altered streamflow and warm water temperatures, which contribute to the decline of native salmonids. In an effort to restore critical salmonid habitat, the Karuk Tribe, National Forest Service, and Mid-Klamath Watershed Council have created a variety of off-channel floodplain ponds that provide habitat for juvenile Coho Salmon (Oncorhynchus kisutch) and other juvenile salmonids such as Steelhead (O. mykiss). One purpose of these ponds is to provide cool water refuges for juvenile salmonids during high summer water temperatures. However, no studies have quantified how these ponds vary in temperature regimes across the river floodplain. In July 2020, we placed 30 temperature sensors in 9 off-channel ponds and 2 creeks (Seiad Creek and Horse Creek) in the mid-Klamath floodplain. We used a multivariate auto-regressive state space (MARSS) models to determine the number and spatial arrangement of distinct thermal regimes in floodplain ponds and tributaries. We found that pond temperatures have lower daily maximums and fluctuate less than tributary temperatures. We also found that Seiad Creek, Seiad Creek ponds, Horse Creek, and Horse Creek ponds all have different patterns of temperature change throughout the summer. Historical data (2010-2019) for Alexander and Stender Ponds showed that over time, daily fluctuations in pond water temperature became less drastic. This pattern was also observed by MKWC in their reports on Alexander and Stender Ponds (MKWC 2020; Wickman et al. 2020). More stable water temperatures in the ponds contrast to creek temperatures, which continue to fluctuate widely on a daily basis during summer. Fish monitoring data from MKWC show that coho growth rates are higher in these two ponds, which suggests that coho experience a metabolic benefit from more stable water temperatures (MKWC 2020; Wickman et al. 2020). Overall, our analysis provides deeper insight into the thermal benefits of floodplain habitats and off-channel ponds on the mid-Klamath River, and informs the future collection of fish data that will reveal more precise information about how floodplains benefit salmonids.

Chronic temperature stress effects on the liver proteome of two threespine stickleback (Gasterosteus aculeatus) populations using a novel DIA assay library

A data-independent acquisition (DIA) assay library was generated for the liver of threespine sticklebacks to evaluate alterations in protein abundance and functional enrichment of molecular pathways following either chronic warm (25°C) or cold (7°C) three-week temperature challenge in two estuarine populations. The DIA assay library was created from a data-dependent acquisition (DDA) based raw spectral library that was filtered to remove low quality or ambiguous peptides. Functional enrichment analyses using STRING identified larger networks that were significantly enriched by examining both the entire liver proteome and only significantly elevated or depleted proteins from the various comparisons. These systems level analyses revealed the unique liver proteomic signatures of two populations of threespine sticklebacks acclimated to chronic temperature stress. The Big lagoon population (BL) had a stronger response than the Klamath river population (KL). At 7°C, BL showed alterations in protein homeostasis that likely fueled a higher demand for energy, but both populations successfully acclimated to this temperature. The warm acclimation induced major increases in proteins involved in chromatin structure and transcription, while there were decreases in proteins related to translation and fatty acid metabolism. Functional enrichment analyses of the entire liver proteome uncovered differences in glycolysis and carbohydrate metabolism between the two populations and between the cold acclimated and control groups. We conclude that the synchronous regulatory patterns of many proteins observed in the liver of threespine sticklebacks provide more comprehensive insight into population-specific responses to thermal stress than the use of less specific pre-determined biomarkers.

Hybridizing Neural Network with Multi-Verse, Black Hole, and Shuffled Complex Evolution Optimizer Algorithms Predicting the Dissolved Oxygen

The great importance of estimating dissolved oxygen (DO) dictates utilizing proper evaluative models. In this work, a multi-layer perceptron (MLP) network is trained by three capable metaheuristic algorithms, namely multi-verse optimizer (MVO), black hole algorithm (BHA), and shuffled complex evolution (SCE) for predicting the DO using the data of the Klamath River Station, Oregon, US. The records (DO, water temperature, pH, and specific conductance) belonging to the water years 2015 - 2018 (USGS) are used for pattern analysis. The results of this process showed that all three hybrid models could properly infer the DO behavior. However, the BHA and SCE accomplished this task by simpler configurations. Next, the generalization ability of the developed patterns is tested using the data of the 2019 water year. Referring to the calculated mean absolute errors of 1.0161, 1.1997, and 1.0122, as well as Pearson correlation coefficients of 0.8741, 0.8453, and 0.8775, the MLPs trained by the MVO and SCE perform better than the BHA. Therefore, these two hybrids (i.e., the MLP-MVO and MLP-SCE) can be satisfactorily used for future applications.

Successful collection and captive rearing of wild-spawned larval Klamath suckers

Shortnose Chasmistes brevirostris and Lost River Suckers Deltistes luxatus endemic to the Klamath River Basin on the California-Oregon border have experienced dramatic population declines in parallel with many other Catostomid species. Captive propagation has become a key element of many endangered fish recovery programs, although there is little evidence of their success in restoring or recovering fish populations. A novel rearing program for Klamath suckers was initiated in 2016 with the goal of developing a husbandry strategy that better balances the ecological, genetic, and demographic risks associated with captive propagation. We collected 4,306 wild-spawned Klamath sucker larvae from a major spawning tributary May-June 2016 and reared them at a geothermal facility established through a partnership with a local landowner and aquaculture expert. Mortality during collection was less than 1%. Larvae were reared in glass aquaria for 17-78 days until reaching approximately 30 mm total length, upon which they were moved to round fiberglass tanks for 14-46 days or until reaching approximately 60 mm total length. Overall survival of larvae to ponding for final growout was 71%. Larval tank rearing survival was 98% for 37 days until an isolated fish health incident affected three aquarium populations, reducing survival to transfer to 75%. Survival after transfer to round fiberglass tanks for 14-46 days was 94%. This study outlines the first successful collection and early life history husbandry of wild-spawned endangered Klamath suckers that we are aware of.

Disease in Central Valley Salmon: Status and Lessons from Other Systems

hinook Salmon (Oncorhynchus tshawytscha) are increasingly vulnerable to anthropogenic activities and climate change, especially at their most southern range in California’s Central Valley. There is considerable interest in understanding stressors that contribute to population decline and in identifying management actions that reduce the effects of those stressors. Along the west coast of North America, disease has been linked to declining numbers of salmonids, and identified as a key stressor that results in mortality. In the Central Valley, targeted studies have revealed extremely high prevalence of infectious agents and disease. However, there has been insufficient monitoring to understand the effect that disease may have on salmon populations. To inform future research, monitoring, and management efforts, a two-day workshop on salmon disease was held at the University of California, Davis (UC Davis) on March 14-15, 2018. This paper summarizes the science presented at this workshop, including the current state of knowledge of salmonid disease in the Central Valley, and current and emerging tools to better understand its effects on salmon. We highlight case studies from other systems where successful monitoring programs have been implemented. First, in the Klamath River where the integration of several data-collection and modeling approaches led to the development of successful management actions, and second in British Columbia where investment in researching novel technologies led to breakthroughs in the understanding of salmon disease dynamics. Finally, we identify key information and knowledge gaps necessary to guide research and management of disease in Central Valley salmon populations.