Subtidal Natural Hard Substrate Quantitative Habitat Mapping: Interlinking Underwater Acoustics and Optical Imagery with Machine Learning

Subtidal natural hard substrates (SNHS) promote occupancy by rich benthic communities that provide irreplaceable and fundamental ecosystem functions, representing a global priority target for nature conservation and recognised in most European environmental legislation. However, scientifically validated methodologies for their quantitative spatial demarcation, including information on species occupancy and fine-scale environmental drivers (e.g., the effect of stone size on colonisation) are rare. This is, however, crucial information for sound ecological management. In this investigation, high-resolution (1 m) multibeam echosounder (MBES) depth and backscatter data and derivates, underwater imagery (UI) by video drop-frame, and grab sediment samples, all acquired within 32 km2 of seafloor in offshore Belgian waters, were integrated to produce a random forest (RF) spatial model, predicting the continuous distribution of the seafloor areal cover/m2 of the stones’ grain sizes promoting colonisation by sessile epilithic organisms. A semi-automated UI acquisition, processing, and analytical workflow was set up to quantitatively study the colonisation proportion of different grain sizes, identifying the colonisation potential to begin at stones with grain sizes Ø ≥ 2 cm. This parameter (i.e., % areal cover of stones Ø ≥ 2 cm/m2) was selected as the response variable for spatial predictive modelling. The model output is presented along with a protocol of error and uncertainty estimation. RF is confirmed as an accurate, versatile, and transferable mapping methodology, applicable to area-wide mapping of SNHS. UI is confirmed as an essential aid to acoustic seafloor classification, providing spatially representative numerical observations needed to carry out quantitative seafloor modelling of ecologically relevant parameters. This contribution sheds innovative insights into the ecologically relevant delineation of subtidal natural reef habitat, exploiting state-of-the-art underwater remote sensing and acoustic seafloor classification approaches.

Download Full-text

Arctic benthic diversity research with PANABIO: scale, sharing, and modelling

10.7287/peerj.preprints.26681 ◽

2018 ◽

Author(s):

Dieter Piepenburg ◽

Jan Holstein ◽

Paul Kloss ◽

Thomas Brey ◽

Casper Kraan

Keyword(s):

Sea Ice ◽

Benthic Communities ◽

Public Information ◽

Environmental Drivers ◽

Ecosystem Functions ◽

Marine Biota ◽

Distribution Models ◽

Ice Dynamics ◽

Environmental Forcing ◽

Sea Ice Dynamics

Arctic marine biota areaffected profoundly and at large scales by accelerating environmental change, such as ocean warming and sea-ice decline. Moreover, increasing human activities add further cumulative pressures. Substantial shifts in ecosystem functions and services,including biodiversity, are expected. Tounderstand, predict, and mitigate the profound ecologicalconsequences of such shifts, it is critical to identify and analyzetherelationships between environmental drivers and ecosystem functionsata range of scales (local, regional, and pan-Arctic). We address this challenge by means of apan-Arctic knowledge system on benthicbiota(PANABIO). Underpinned by international efforts to combine data and expertise, PANABIO integrates quality-controlled and geo-referenceddata on benthiccommunities in a public information system. The system allows for (a) providingecological baseline-data to gauge ecosystem changes, (b) analysingcoupling mechanisms between environmental drivers and ecosystemfunctions/services on regional and pan-Arctic scales, (c) developing futureecosystem scenarios in response to external forcing, and (d) creating onlinestakeholder-orientedvisualization and analysis tools. The talk will demonstrate the huge up-scaling of benthic data with PANABIO, our achievements to support data-sharing, as well as first results of community-level distribution models to discern benthic communities in relation to multiple-factor environmental forcing, including sea-ice dynamics.

Download Full-text

Arctic benthic diversity research with PANABIO: scale, sharing, and modelling

10.7287/peerj.preprints.26681v1 ◽

2018 ◽

Author(s):

Dieter Piepenburg ◽

Jan Holstein ◽

Paul Kloss ◽

Thomas Brey ◽

Casper Kraan

Keyword(s):

Sea Ice ◽

Benthic Communities ◽

Public Information ◽

Environmental Drivers ◽

Ecosystem Functions ◽

Marine Biota ◽

Distribution Models ◽

Ice Dynamics ◽

Environmental Forcing ◽

Sea Ice Dynamics

Download Full-text

Predictive modelling of hospital readmission: Evaluation of different preprocessing techniques on machine learning classifiers

Intelligent Data Analysis ◽

10.3233/ida-205468 ◽

2021 ◽

Vol 25 (5) ◽

pp. 1073-1098

Author(s):

Nor Hamizah Miswan ◽

Chee Seng Chan ◽

Chong Guan Ng

Keyword(s):

Machine Learning ◽

Hospital Readmission ◽

Performance Metrics ◽

Predictive Performance ◽

Predictive Modelling ◽

Machine Learning Classifiers ◽

Learning Classifiers ◽

Set Up ◽

The Right ◽

The Impact

Hospital readmission is a major cost for healthcare systems worldwide. If patients with a higher potential of readmission could be identified at the start, existing resources could be used more efficiently, and appropriate plans could be implemented to reduce the risk of readmission. Therefore, it is important to predict the right target patients. Medical data is usually noisy, incomplete, and inconsistent. Hence, before developing a prediction model, it is crucial to efficiently set up the predictive model so that improved predictive performance is achieved. The current study aims to analyse the impact of different preprocessing methods on the performance of different machine learning classifiers. The preprocessing applied by previous hospital readmission studies were compared, and the most common approaches highlighted such as missing value imputation, feature selection, data balancing, and feature scaling. The hyperparameters were selected using Bayesian optimisation. The different preprocessing pipelines were assessed using various performance metrics and computational costs. The results indicated that the preprocessing approaches helped improve the model’s prediction of hospital readmission.

Download Full-text

Posterior contraction in sparse generalized linear models

Biometrika ◽

10.1093/biomet/asaa074 ◽

2020 ◽

Author(s):

Seonghyun Jeong ◽

Subhashis Ghosal

Keyword(s):

Generalized Linear Models ◽

Bayesian Methods ◽

Generalized Linear Model ◽

Linear Models ◽

Continuous Distribution ◽

Fractional Power ◽

Regression Coefficients ◽

Link Function ◽

Convergence Properties ◽

Set Up

Summary We study posterior contraction rates in sparse high-dimensional generalized linear models using priors incorporating sparsity. A mixture of a point mass at zero and a continuous distribution is used as the prior distribution on regression coefficients. In addition to the usual posterior, the fractional posterior, which is obtained by applying Bayes theorem with a fractional power of the likelihood, is also considered. The latter allows uniformity in posterior contraction over a larger subset of the parameter space. In our set-up, the link function of the generalized linear model need not be canonical. We show that Bayesian methods achieve convergence properties analogous to lasso-type procedures. Our results can be used to derive posterior contraction rates in many generalized linear models including logistic, Poisson regression and others.

Download Full-text

Modelling temporal dynamics of Culicoides Latreille (Diptera: Ceratopogonidae) populations on Reunion Island (Indian Ocean), vectors of viruses of veterinary importance

Parasites & Vectors ◽

10.1186/s13071-019-3812-1 ◽

2019 ◽

Vol 12 (1) ◽

Cited By ~ 2

Author(s):

Yannick Grimaud ◽

Hélène Guis ◽

Frédéric Chiroleu ◽

Floriane Boucher ◽

Annelise Tran ◽

...

Keyword(s):

Vegetation Index ◽

Temporal Dynamics ◽

Control Strategies ◽

Host Density ◽

Environmental Drivers ◽

Hurdle Model ◽

Reunion Island ◽

Annual Variations ◽

Réunion Island ◽

Set Up

Abstract Background Reunion Island regularly faces outbreaks of epizootic haemorrhagic disease (EHD) and bluetongue (BT), two viral diseases transmitted by haematophagous midges of the genus Culicoides (Diptera: Ceratopogonidae) to ruminants. To date, five species of Culicoides are recorded in Reunion Island in which the first two are proven vector species: Culicoides bolitinos, C. imicola, C. enderleini, C. grahamii and C. kibatiensis. Meteorological and environmental factors can severely constrain Culicoides populations and activities and thereby affect dispersion and intensity of transmission of Culicoides-borne viruses. The aim of this study was to describe and predict the temporal dynamics of all Culicoides species present in Reunion Island. Methods Between 2016 and 2018, 55 biweekly Culicoides catches using Onderstepoort Veterinary Institute traps were set up in 11 sites. A hurdle model (i.e. a presence/absence model combined with an abundance model) was developed for each species in order to determine meteorological and environmental drivers of presence and abundance of Culicoides. Results Abundance displayed very strong heterogeneity between sites. Average Culicoides catch per site per night ranged from 4 to 45,875 individuals. Culicoides imicola was dominant at low altitude and C. kibatiensis at high altitude. A marked seasonality was observed for the three other species with annual variations. Twelve groups of variables were tested. It was found that presence and/or abundance of all five Culicoides species were driven by common parameters: rain, temperature, vegetation index, forested environment and host density. Other parameters such as wind speed and farm building opening size governed abundance level of some species. In addition, Culicoides populations were also affected by meteorological parameters and/or vegetation index with different lags of time, suggesting an impact on immature stages. Taking into account all the parameters for the final hurdle model, the error rate by Normalized Root mean Square Error ranged from 4.4 to 8.5%. Conclusions To our knowledge, this is the first study to model Culicoides population dynamics in Reunion Island. In the absence of vaccination and vector control strategies, determining periods of high abundance of Culicoides is a crucial first step towards identifying periods at high risk of transmission for the two economically important viruses they transmit.

Download Full-text

Uncovering the environmental drivers of short-term temporal dynamics in an epibenthic community from the Western English Channel

Journal of the Marine Biological Association of the United Kingdom ◽

10.1017/s0025315419000663 ◽

2019 ◽

Vol 99 (7) ◽

pp. 1467-1479

Author(s):

Elizabeth Talbot ◽

Jorn Bruggeman ◽

Chris Hauton ◽

Stephen Widdicombe

Keyword(s):

Time Series Data ◽

Temporal Dynamics ◽

Anthropogenic Impacts ◽

Benthic Communities ◽

Food Limitation ◽

Environmental Drivers ◽

Sediment Surface ◽

Series Data ◽

English Channel ◽

Larval Recruitment

AbstractBenthic communities, critical to the health and function of marine ecosystems, are under increasing pressure from anthropogenic impacts such as pollution, eutrophication and climate change. In order to refine predictions of likely future changes in benthic communities resulting from these impacts, we must first better constrain their responses to natural seasonality in environmental conditions. Epibenthic time series data (July 2008–May 2014) have been collected from Station L4, situated 7.25 nautical miles south of Plymouth in the Western English Channel. These data were analysed to establish patterns in community abundance, wet biomass and composition, and to link any observed patterns to environmental variables. A clear response to the input of organic material from phytoplankton blooms was detected, with sediment surface living deposit feeders showing an immediate increase in abundance, while predators and scavengers responded later, with an increase in biomass. We suggest that this response is a result of two factors. The low organic content of the L4 sediment results in food limitation of the community, and the mild winter/early spring bottom water temperatures allow the benthos to take immediate advantage of bloom sedimentation. An inter-annual change in community composition was also detected, as the community shifted from one dominated by the anomuran Anapagurus laevis to one dominated by the gastropod Turitella communis. This appeared to be related to a period of high larval recruitment for T. communis in 2013/2014, suggesting that changes in the recruitment success of one species can affect the structure of an entire community.

Download Full-text

Impaired benthic macrofauna function 4 years after sediment capping with activated carbon in the Grenland fjords, Norway

Environmental Science and Pollution Research ◽

10.1007/s11356-020-11607-0 ◽

2020 ◽

Author(s):

Caroline Raymond ◽

Göran S Samuelsson ◽

Stefan Agrenius ◽

Morten T Schaanning ◽

Jonas S Gunnarsson

Keyword(s):

Activated Carbon ◽

Biological Effects ◽

Benthic Communities ◽

Ecosystem Functions ◽

Sediment Surface ◽

Sediment Remediation ◽

Vital Functions ◽

Southern Norway

AbstractThe sediments in the Grenland fjords in southern Norway are heavily contaminated by large emissions of dioxins and mercury from historic industrial activities. As a possible in situ remediation option, thin-layer sediment surface capping with powdered activated carbon (AC) mixed with clay was applied at two large test sites (10,000 and 40,000 m2) at 30-m and 95-m depths, respectively, in 2009. This paper describes the long-term biological effects of the AC treatment on marine benthic communities up to 4 years after treatment. Our results show that the capping with AC strongly reduced the benthic species diversity, abundance, and biomass by up to 90%. Vital functions in the benthic ecosystem such as particle reworking and bioirrigation of the sediment were also reduced, analyzed by using novel bioturbation and bioirrigation indices (BPc, BIPc, and IPc). Much of the initial effects observed after 1 and 14 months were still present after 49 months, indicating that the effects are long-lasting. These long-lasting negative ecological effects should be carefully considered before decisions are made on sediment remediation with powdered AC, especially in large areas, since important ecosystem functions can be impaired.

Download Full-text

Brief communication: Post-seismic landslides, the tough lesson of a catastrophe

10.5194/nhess-2017-363 ◽

2017 ◽

Cited By ~ 1

Author(s):

Qiang Xu ◽

Xuanmei Fan ◽

Gianvito Scaringi

Keyword(s):

Rock Avalanche ◽

Predictive Modelling ◽

Early Warning Systems ◽

Landslide Risk ◽

Seismic Behaviour ◽

Policy Makers ◽

Seismic Slope Stability ◽

Set Up ◽

The Village

Abstract. The rock avalanche that destroyed the village of Xinmo in Sichuan, China, on June 24th, 2017, brought the issue of landslide risk and disaster chain management in highly seismic regions back into the spotlight. The long-term post-seismic behaviour of mountain slopes is complex and hardly predictable. Nevertheless, the integrated use of field monitoring, remote sensing and real-time predictive modelling can help to set-up effective early warning systems, provide timely alarms, optimize rescue operations and perform secondary hazard assessments. We believe that a comprehensive discussion on post-seismic slope stability and on its implications for policy makers can no longer be postponed.

Download Full-text

Set-up and performance evaluation of a seismic rock fall observatory in the Alps

10.5194/egusphere-egu2020-14290 ◽

2020 ◽

Author(s):

Anne Schöpa ◽

Niels Hovius ◽

Jens Turowski

Keyword(s):

Data Storage ◽

Meteorological Data ◽

A Priori ◽

Monitoring Network ◽

Rock Fall ◽

Environmental Drivers ◽

Rock Falls ◽

Alpine Valley ◽

Mountain Environments ◽

Set Up

Rock falls are important agents of erosion shaping the topography of bedrock slopes. Despite the considerable attention rock falls get when causing damage we still lack detailed information about the triggers, lag times, seasonal and elevation-dependent rock fall occurrence. This is due to the difficulty in observing rockfalls directly as the mobilisation of rock masses occurs rapidly, infrequently and distributed at a priori unknown locations. To identify seasonal and elevation-dependent rock fall activities and characteristics and their environmental drivers and triggers in an alpine setting, we have operated a monitoring network to detect and classify rock falls in the Reintal valley, German Alps, since 2014. The Reintal is an Alpine valley in the Wetterstein massif close to the Zugspitze, Germany&#8217;s highest mountain. The Reintal observatory produces nearly continuous datasets of seismic, meteorological and camera data. To our knowledge, these datasets are one of a few that permit a systematic study of rockfall patterns and their controls over a period of several years in an alpine setting.In this contribution, we present the layout of the observatory and the instrumental network. Six seismometers record the motion of the ground; different types of seismic signals are shown and their sources discussed. This is done in combination with the meteorological data of the two weather stations in the valley and the images of the optical and infrared cameras of the observatory. We evaluate the performance, limitations and capabilities of the observatory. In addition, we discuss how we dealt with challenges such as power consumption of the instruments in the field, data storage and data loss. Our experience with the set-up and maintenance of the observatory can help guide the design and construction of other observatories in mountain environments.

Download Full-text

Proxies, drivers, and impact of macrofaunal transport in sediment of the southern North Sea.

10.5194/egusphere-egu2020-20810 ◽

2020 ◽

Author(s):

Andreas Neumann ◽

Justus van Beusekom ◽

Annika Eisele ◽

Kay-Christian Emeis ◽

Jana Friedrich ◽

...

Keyword(s):

Exchange Rates ◽

Pore Water ◽

German Bight ◽

Benthic Communities ◽

Coastal Sediments ◽

Ex Situ ◽

Environmental Drivers ◽

Sediment Type ◽

Physical Parameters ◽

Water Control

Coastal sediments play an important role in the nutrient cycling, and the intensities of exchange processes between bottom water and pore water control the balance between sequestration and recycling of nutrients. Pore water advection as one major exchange mechanism is determined by physical parameters and thus well describable with models. By contrast, biotransport (bioirrigation, bioturbation) as the other major transport mechanism is much more complex and observational data are often scarce to quantify these processes.We present ex-situ observations of oxygen and nutrient fluxes, sediment characteristics, and fauna composition over the past six years from all benthic provinces of the German Bight, which enable us to describe the spatial and seasonal variability of the benthic- pelagic coupling. We employ this dataset to detect environmental drivers of the observed variability and to test several proxies of faunal activity.Our results show that abiotic parameters (sediment type, local primary production) explain the spatial variability while the dynamics of temperature and faunal activity explain the temporal variability. Effects of the complex benthic communities on benthic exchange rates can be parameterized by surprisingly simple proxies, which may help to improve benthic exchange models. By comparing in-situ measurements of pore water advection with ex situ observations, we conclude that biotransport approximately doubles the benthic- pelagic exchange rates in the German Bight.

Download Full-text