A multi-scale feature selection approach for predicting benthic assemblages

Nemani, Shreya; Côté, David; Misiuk, Benjamin; Edinger, Evan; Mackin-McLaughlin, Julia; Templeton, Adam; Shaw, John; Robert, Katleen

doi:10.1016/j.ecss.2022.108053

Cited by 17 publications

(7 citation statements)

References 79 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite being an initial study, the results here presented can play a key role in supporting the implementation of conservation/management plans for bait harvesting activities in this coastal lagoon and elsewhere, namely estuarine systems where bait harvesting is more intense and relevant from an economic, social, and ecological perspective. The use of statistical methods, such as the Boruta selection function used in this study, can also be applied in other areas of marine science, such as benthic habitat mapping [78]. The selection performed by Boruta helps to achieve more robust and accurate models, as well as to understand which explanatory variables (e.g., elements and FA) characterize each categorical response variable (e.g., location)."…”

Section: Discussionmentioning

confidence: 99%

Combined Use of Fatty Acid Profiles and Elemental Fingerprints to Trace the Geographic Origin of Live Baits for Sports Fishing: The Solitary Tube Worm (Diopatra neapolitana, Annelida, Onuphidae) as a Case Study

Ricardo,

Lopes,

Mamede

et al. 2024

Animals

View full text Add to dashboard Cite

Diopatra neapolitana Delle Chiaje, 1841 (Annelida, Onuphidae) is one of the most exploited polychaete species in European waters, particularly in Ria de Aveiro, a coastal lagoon in mainland Portugal, where the overexploitation of this resource has led to a generalized decline of local populations. In an attempt to reduce the impact of harvesting, several management actions were implemented, but illegal poaching still fuels a parallel economy that threatens the sustainable use of this marine resource. The present study evaluated the combination of fatty acid profiles and elemental fingerprints of the whole body and jaws, respectively, of D. neapolitana collected from four harvesting locations within Ria de Aveiro in order to determine if their geographic origin could be correctly assigned post-harvesting. Results showed that both fatty acid profiles and elemental fingerprints differ significantly among locations, discriminating the geographic origin with higher accuracy when combining these two natural barcodes than when employing each individually. The present work can, therefore, contribute to the implementation of an effective management plan for the sustainable use of this marine resource, making it possible to detect if D. neapolitana was sourced from no-take zones and if it was collected from the place of origin claimed by live bait traders.

show abstract

Section: Discussionmentioning

confidence: 99%

Combined Use of Fatty Acid Profiles and Elemental Fingerprints to Trace the Geographic Origin of Live Baits for Sports Fishing: The Solitary Tube Worm (Diopatra neapolitana, Annelida, Onuphidae) as a Case Study

Ricardo,

Lopes,

Mamede

et al. 2024

Animals

View full text Add to dashboard Cite

show abstract

“…A variable is deemed important if it consistently contributes more to the model than its shadow variable. Degenhardt et al, 2019 found that the Boruta algorithm generally outperformed other selection methods, and previous successful applications can be found in Li et al, 2016, Diesing andThorsnes, (2017) and Nemani et al, 2021. Variables identified as "important" or "tentative" were selected for model training here.…”

Section: Model Developmentmentioning

confidence: 97%

“…For both the assemblages and for each marine algae taxa model, a Boruta Feature Selection (Kurse and Rudnicki, 2010) algorithm was run separately to include terrain features grouped by scale (window of analysis: 3 × 3, 13 × 13, 35 × 35 cells) in order to reduce the number of candidate terrain features and promote model parsimony and support performance (Nemani et al, 2021). Important variables are identified by the Boruta wrapper as it compares the importance of a variable with a randomly shuffled version containing the same distribution of values (i.e., "shadow features").…”

Section: Model Developmentmentioning

confidence: 99%

Spatial distribution of benthic flora and fauna of coastal placentia bay, an ecologically and biologically significant area of the island of newfoundland, atlantic Canada

Mackin-McLaughlin

Nemani²,

Misiuk³

et al. 2022

Front. Environ. Sci.

View full text Add to dashboard Cite

Coastal habitats have the potential to be biodiversity hotspots that provide important ecosystem services, but also hotspots for human development and exploitation. Continued use of coastal ecosystem services requires establishing baselines that capture the present state of the benthos. This study employs habitat mapping to establish a baseline describing the spatial distribution of benthic organisms along the western coast of Placentia Bay, an Ecologically and Biologically Significant Area (EBSA) in Newfoundland, Canada. The influence of seafloor characteristics on the distribution of four dominant epifaunal assemblages and two macrophyte species were modelled using two machine learning techniques: the well-established Random Forest and the newer Light Gradient Boosting Machine. When investigating model performance, the inclusion of fine-scale (<1 m) substrate information from the benthic videos was found to consistently improve model accuracy. Predictive maps developed here suggest that the majority of the surveyed areas consisted of a species-rich epifaunal assemblage dominated by ophiuroids, porifera, and hydrozoans, as well as prominent coverage by Agarum clathratum and non-geniculate crustose coralline algae. These maps establish a baseline that enables future monitoring of Placentia Bay’s coastal ecosystem, helping to conserve the biodiversity and ecosystem services this area provides.

show abstract

“…Compared to other machine learning models, the Extreme Gradient Boosting (XGBoost) model based on Decision Trees (DT) is a flexible and versatile tool capable of solving a wide range of regression and classification problems. The non-linear properties of the estimated variables are accurately captured at a low computational cost and without over-fitting, as shown by Nemani et al (2022). The XGBoost model differs from Random Forest (RF) in that it combines both DT and Gradient Boosting (GB) to aggregate the weak learners' predictions and strengthen the strong learners through effective training strategies (Fan et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Inter‐ and intra‐annual fluctuations and attribution analysis of runoff and sediment transport in the middle and lower Yangtze River

Yu,

Du,

et al. 2024

Hydrological Processes

View full text Add to dashboard Cite

Assessing the changing regimes of runoff and sediment load in rivers, as well as analysing their underlying causes, is essential for maintaining the ecological integrity of rivers and promoting the sustainable development of water resources. Current methods of attribution analysis focus on accounting for the effects of climate change and human activities on inter‐annual variability in streamflow and sediment. However, the contribution of individual drivers to intra‐annual hydrologic variability is often overlooked. The study analysed the inter‐ and intra‐annual fluctuations in runoff and sediment load in the middle and lower Yangtze River using a water‐sediment variation attribution model, the Extreme Gradient Boosting algorithm, and the Multiple‐Factor Attribution method. The results showed that human activities were responsible for over 70% of the inter‐ and intra‐annual fluctuation in runoff and sediment load in the Yichang, Hankou, and Datong sections. Additionally, it was found that human activities were the main contributor to the reduction in both runoff and sediment load in all sections. Closer to the Yangtze estuary, human influence on intra‐annual fluctuation of runoff decreases while climate change influence increases, and human influence on both inter‐ and intra‐annual fluctuation of sediment load becomes less significant. The sensitivity of annual runoff to precipitation is greater than that of air temperature and evaporation in the Yichang, Hankou, and Datong sections. An increase of 10% in precipitation from baseline values leads to a corresponding increase of 3.13%, 2.24%, and 3.91% in runoff for the Yichang, Hankou, and Datong sections. The annual sediment load is more sensitive to changes in runoff than to precipitation, air temperature, and evaporation in each section. A 10% increase in runoff from the baseline value leads to an annual sediment load increase of 20.36%, 10.11%, and 8.78% at Yichang, Hankou, and Datong, respectively. This study provides a theoretical basis and guidance for decision‐making to protect and sustainably develop the water and sediment resources in the Yangtze River Basin.

show abstract

A multi-scale feature selection approach for predicting benthic assemblages

Cited by 17 publications

References 79 publications

Combined Use of Fatty Acid Profiles and Elemental Fingerprints to Trace the Geographic Origin of Live Baits for Sports Fishing: The Solitary Tube Worm (Diopatra neapolitana, Annelida, Onuphidae) as a Case Study

Combined Use of Fatty Acid Profiles and Elemental Fingerprints to Trace the Geographic Origin of Live Baits for Sports Fishing: The Solitary Tube Worm (Diopatra neapolitana, Annelida, Onuphidae) as a Case Study

Spatial distribution of benthic flora and fauna of coastal placentia bay, an ecologically and biologically significant area of the island of newfoundland, atlantic Canada

Inter‐ and intra‐annual fluctuations and attribution analysis of runoff and sediment transport in the middle and lower Yangtze River

Contact Info

Product

Resources

About