Using explainable machine learning methods to evaluate vulnerability and restoration potential of ecosystem state transitions.

Ecosystem state transitions can be ecologically devastating or be a restoration success. State transitions are common within aquatic systems worldwide, especially considering human-mediated changes to land use and water use. We created a transferable conceptual framework to enable multiscale assessments of state resilience and early warnings of state transitions that can inform strategic restorations and avoid ecosystem collapse. The conceptual framework integrated machine learning predictions with ecosystem state concepts (e.g., state classification, gradients of vulnerability, and recovery potential leading to state transitions) and was devised to investigate possible environmental drivers. As an application of the framework, we generated prediction probabilities of submersed aquatic vegetation (SAV) presence at nearly 10,000 sites in the Upper Mississippi River (United States). Then, we used an interpretability method to explain model predictions to gain insights into possible environmental drivers and thresholds or linear responses of SAV presence and absence. Model accuracy was 89% without spatial bias. Average water depth, suspended solids, substrate, and distance to nearest SAV were the best predictors and likely environmental drivers of SAV habitat suitability. These environmental drivers exhibited nonlinear, threshold-type responses for SAV. All the results are also presented in an online dashboard to explore results at many spatial scales. The habitat suitability model outputs and prediction explanations from many spatial scales (4 m to 400 km of river reach) can inform research and restoration planning.

Métodos de aprendizaje automático para evaluar la vulnerabilidad y el potencial de restauración de las transiciones de estados de los ecosistemas Resumen Las transiciones de estado de los ecosistemas pueden tener consecuencias ecológicas graves o ser un éxito de restauración. Estas transiciones son comunes en los sistemas acuáticos, especialmente si consideramos los cambios mediados por humanos en el uso del suelo y del agua. Generamos un marco conceptual transferible para permitir las evaluaciones multiescalares de la resiliencia del estado y las alertas tempranas de la transición de estado que sirvan para guiar las restauraciones estratégicas y evitar el colapso del ecosistema. Este marco integró predicciones de aprendizaje automático a los conceptos de estado del ecosistema (p. ej.: clasificación del estado, gradientes de vulnerabilidad y potencial de restauración que lleve a transiciones de estado) y fue diseñado para investigar los posibles factores ambientales. Para aplicar el marco, generamos probabilidades de predicción de la presencia de vegetación acuática sumergida (VAS) en casi 10,000 sitios del Alto Río Mississippi en los Estados Unidos. Después usamos un método de interpretabilidad para explicar las predicciones del modelo y obtener información de los posibles factores ambientales y umbrales o respuestas lineales de la presencia y ausencia de la VAS. La precisión del modelo fue de 89% sin sesgo espacial. Los mejores pronosticadores y probables factores ambientales de la idoneidad de hábitat de la VAS fueron la profundidad promedio del agua, los sólidos suspendidos, el sustrato y la distancia a la VAS más cercana. Estos factores ambientales exhibieron respuestas no lineales de tipo umbral para la VAS. Todos los resultados también están presentados en un tablero virtual para explorar los resultados a varias escalas espaciales. Los resultados del modelo de idoneidad de hábitat y las explicaciones de las predicciones de varias de las escalas espaciales (4 m-400 km de alcance del río) pueden guiar la investigación y la planeación de la restauración.

Land conservation can mitigate freshwater ecosystem services degradation due to climate change in a semiarid catchment: The case of the Portneuf River catchment, Idaho, USA.

There is increasing evidence of environmental change impacts on freshwater ecosystem services especially through land use and climate change. However, little is known about how land conservation could help mitigate adverse water-sustainability impacts. In this paper, we utilized the InVEST tool and the Residual Trends method to assess the joint effects and relative contributions of climate change and land conservation on freshwater ecosystem services in the Portneuf River catchment in Idaho, USA. We developed five hypothesized scenarios regarding gain and loss in the enrollment of Conservation Reserve Program (CRP), the largest agricultural land-retirement program in the U.S., plus riparian buffer and assessed their interactions with climate change. Results suggest that the realized water yield in the Portneuf River catchment would possibly be 56% less due to climate change and 24% less due to the decline of CRP enrollment. On the contrary, if CRP enrollment is promoted by ~30% and riparian buffer protection is implemented, the water supply reduction in the year 2050 could be changed from 56% to 26%, the total phosphorus (TP) and total nitrogen (TN) export would be reduced by 10% and 11%, and the total suspended sediment (TSS) reduced by 17%. This study suggests that increasing implementation of the CRP would likely preserve key freshwater ecosystem services and assist proactive mitigation, especially for semiarid regions vulnerable to changing climate conditions.

Ecotoxicological assessments show sucralose and fluoxetine affect the aquatic plant, Lemna minor.

Pharmaceuticals and personal care products (PPCP) are prevalent in aquatic systems, yet the fate and impacts on aquatic plants needs quantification for many compounds. We measured and detected sucralose (an artificial sweetener), fluoxetine (an antidepressant), and other PPCP in the Portneuf River in Idaho, USA, where Lemna minor (an aquatic plant in the environment and used in ecotoxicology studies) naturally occurs. Sucralose was hypothesized to negatively affect photosynthesis and growth of L. minor because sucralose is a chlorinated molecule that may be toxic or unusable for plant metabolism. A priori hypotheses were not created for fluoxetine due to lack of previous studies examining its impacts on plants. We conducted laboratory ecotoxicological assessments for a large range of concentrations of sucralose and fluoxetine on L. minor physiology and photosynthetic function. Frond green leaf area, root length, growth rate, photosynthetic capacity, and plant carbon isotopic composition (discrimination relative to a standard; δ13C) were measured among treatments ranging from 0 to 15000nmol/L-sucralose and 0-323nmol/L-fluoxetine. Contrary to our predictions, sucralose significantly increased green leaf area, photosynthetic capacity, and δ 13C of L. minor at environmentally relevant concentrations. The increase of δ 13C from sucralose amendments and an isotope-mixing model indicated substantial sucralose uptake and assimilation within the plant. Unlike humans who cannot break down and utilize sucralose, we documented that L. minor-a mixotrophic plant-can use sucralose as a sugar substitute to increase its green leaf area and photosynthetic capacity. Fluoxetine significantly decreased L. minor root growth, daily growth rate, and asexual reproduction at 323nmol/L-fluoxetine; however, ambiguity remains regarding the mechanisms responsible and the applicability of these extreme concentrations unprecedented in the natural environment. To our knowledge, this was the first study to show aquatic plants can uptake and metabolize sucralose as a carbon source. This study further supports the common notion that L. minor can be useful in bioremediation of PPCP from wastewaters.

Grassland fire and cattle grazing regulate reptile and amphibian assembly among patches.

Fire and grazing are common management schemes of grasslands globally and are potential drivers of reptilian and amphibian (herpetofauna) metacommunity dynamics. Few studies have assessed the impacts of fire and cattle grazing on herpetofauna assemblages in grasslands. A patch-burn grazing study at Osage Prairie, MO, USA in 2011-2012 created landscape patches with treatments of grazing, fire, and such legacies. Response variables were measured before and after the application of treatments, and I used robust-design occupancy modeling to estimate patch occupancy and detection rate within patches, and recolonization and extinction (i.e., dispersal) across patches. I conducted redundancy analysis and a permuted multivariate analysis of variance to determine if patch type and the associated environmental factors explained herpetofauna assemblage. Estimates for reptiles indicate that occupancy was seasonally constant in Control patches (ψ ~ 0.5), but declined to ψ ~ 0.15 in patches following the applications of fire and grazing. Local extinctions for reptiles were higher in patches with fire or light grazing (ε ~ 0.7) compared to the controls. For the riparian herpetofaunal community, patch type and grass height were important predictors of abundance; further, the turtles, lizards, snakes, and adult amphibians used different patch types. The aquatic amphibian community was predicted by watershed and in-stream characteristics, irrespective of fire or grazing. The varying responses from taxonomic groups demonstrate habitat partitioning across multiple patch types undergoing fire, cattle grazing, and legacy effects. Prairies will need an array of patch types to accommodate multiple herpetofauna species.

Ecosystem characteristics of remnant, headwater tallgrass prairie streams.

North America has lost >95% of its native tallgrass prairie due to land conversion, making prairie streams one of the most endangered ecosystems. Research on the basic ecosystem characteristics of the remaining natural prairie streams will inform conservation and management. We examined the structure and function of headwater streams draining tallgrass prairie tracts at Osage Prairie in Missouri and the Konza Prairie Biological Station in Kansas and compared those values with literature values for streams draining agricultural watersheds in the region. We quantified physicochemical and biological characteristics for 2 yr. Streams at Osage and Konza were characterized by low nutrients and low suspended sediments (substantially lower than impacted sites in the region), slight heterotrophic status, and high temporal variability. Suspended sediments and nutrient concentrations were generally low in all prairie streams, but storms increased concentrations of both by 3- to 12-fold. Spring prescribed burns were followed by a slight increase in chlorophyll and decreased nutrients, potentially due to greater light availability. Benthic macroinvertebrate communities at Osage showed seasonal patterns that were probably linked to variable hydrology. We found nine amphibian species using the Osage streams as habitat or breeding sites, but little usage at Konza was probably due to dry conditions and low discharge. Our study indicates that two remnant tallgrass prairie streams along a longitudinal gradient are fairly similar in terms of physicochemical features and have good water quality relative to agricultural watersheds but can differ considerably in macroinvertebrate and amphibian abundance.