Ensemble projections of fish distribution in response to climate changes in the Yellow and Bohai Seas, China

Chen, Yunlong; Shan, Xiujuan; Gorfine, Harry; Dai, Fangqun; Wu, Qiang; Yang, Tao; Shi, Yaohui; Jin, Xianshi

doi:10.1016/j.ecolind.2022.109759

Cited by 13 publications

(4 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For this study, we opted for ensemble methods utilizing DT, DL, and NB models to enhance learning from imbalanced data. This study's conclusions were supported by various research works, such as Jardim et al (2021) advising on fisheries management; Alfatinah et al (2023) predicting skipjack area for each time slice; Chen et al (2023) projecting fish distribution in response to climate changes; and Khiem et al (2023) predicting the growth of abalone.…”

Section: Conclusion Acknowledgementsmentioning

confidence: 54%

An Ensemble Learning Technique for Predicting Mortality Rate in Red Tilapia (Oreochromis niloticus Linn.) Fingerlings

2024

JFE

View full text Add to dashboard Cite

Aquaculture has witnessed a gradual transformation owing to advancement in automatic and intelligent technology. Coupled with the power of high-performance computers, these innovations have given rise to machine learning technologies capable of extracting valuable insights from data. Consequently, these technologies are poised to usher smart aquaculture into a new era of efficiency and productivity. This particular study focused on enhancing the predictive accuracy of mortality rates in red tilapia (Oreochromis niloticus Linn.) fingerlings raised in outdoor earthen ponds with a recirculating aquaculture system. To achieve this, the study leveraged a voting-based ensemble learning technique based on the combination of three single predictive algorithms: decision tree, deep learning, and naïve bayes (EL-V (DT-DL-NB) ). The initial phase of the research involved the compilation of a comprehensive dataset encompassing parameters were water temperature (ºC), dissolved oxygen (mg•L -1 ), pH, total ammonia nitrogen (mg•L -1 ), nitrite-nitrogen (mg•L -1 ), transparency (cm), alkalinity (mg•L -1 ), date, month and mortality rate (fish•day -1 ). Following the collection and cleaning of the dataset, 173 samples with 12 attributes were used in this study. The outcomes of this investigation revealed that the performance of the individual predictive models was eclipsed by the proposed EL-V (DT-DL-NB) model, boasting an impressive accuracy rate of 90.85%, precision of 84.00%, recall of 77.50%, and area under curve of 0.896. These results affirm the potential utility of the proposed model for accurately forecasting the mortality rate of red tilapia fingerling in aquaculture settings, thereby contributing significantly to the optimization of aquaculture practices.

show abstract

Section: Conclusion Acknowledgementsmentioning

confidence: 54%

An Ensemble Learning Technique for Predicting Mortality Rate in Red Tilapia (Oreochromis niloticus Linn.) Fingerlings

2024

JFE

View full text Add to dashboard Cite

show abstract

“…Therefore, sea level drops during the Middle Pleistocene could have caused a historical confluence of adjacent drainages, including the Luan and Liao rivers as well as Yellow and Hai rivers, resulting in close genetic relationships between the populations of G. rivuloides within the group of the Yellow + Hai rivers or the Luan + Liao rivers. The historical exchange of freshwater fish populations among coastal drainages surrounding the Bohai Sea as a result of paleo-drainage connection during the Upper Pleistocene sea level drop has also been indicated by previous phylogeographic studies [29,34,74,75].…”

Section: Genetic Structure and Historical Demographymentioning

confidence: 56%

“…However, previous phylogeographic studies revealed the lack of phylogeographic structure or secondary contact between divergent lineages of freshwater fishes living in the river basins surrounding the Bohai Sea or across the Liao River and Amur River [18,29,34,35,74,75]. These findings are inconsistent, with a clear phylogeographic break appearing in G. rivuloides.…”

Section: Drivers Of Lineage Divergencementioning

confidence: 90%

Pleistocene Landscape Dynamics Drives Lineage Divergence of a Temperate Freshwater Fish Gobio rivuloides in Coastal Drainages of Northern China

Ni,

Chen,

Deng

et al. 2023

Genes

View full text Add to dashboard Cite

Understanding historical processes underlying lineage distribution patterns is a primary goal of phylogeography. We selected Gobio rivuloides (Cypriniformes: Gobionidae) as a model to improve our knowledge about how intraspecific genetic divergence of freshwater fishes arises in coastal drainages of northern China via statistical analysis using cytochrome b gene. The time-calibrated phylogeny of G. rivuloides showed the divergence of two major lineages (I and II) at ~0.98 Ma (million years ago). Lineage I can be divided into two sub-lineages (I-A and I-B) with a divergence time of ~0.83 Ma. Sub-lineage I-A inhabits the Amur River, and sub-lineage I-B lives in the Luan River and Liao River. Lineage II is distributed in the Yellow River and Hai River, with close genetic relationships between the two drainages, and can be split into two sub-lineages (II-C and II-D) with a divergence time of ~0.60 Ma. Our findings indicate that the splitting of lineages and sub-lineages could be attributed to geographic isolation caused by the formation of the Bohai Sea, river capture, and the episodic hydrologic closing of a paleolake during the late Lower–Middle Pleistocene. It is also the first report we know of displaying a clear phylogeographic break for freshwater fishes across coastal drainages in northern China.

show abstract

“…For dryland ecosystems, the assessment of ecological drought is still in its infancy. At present, vegetation index based on remote sensing is commonly used to characterize the drought situation of vegetation, such as temperature vegetation Drought Index (TVDI) 11,12 , Standardized Vegetation Difference Index (NDVI) 13,14 , improved Vegetation Index (EVI) 15 , Vegetation Condition Index (VCI) 16,17 , Vegetation Water Supply Index (VSWI) 13,18 . Although these vegetation indices can indirectly re ect the in uence of drought on vegetation and the water consumption of vegetation, they cannot directly re ect the dynamic changes in the water supply and demand balance of vegetation in the process of ecological drought.…”

Section: Introductionmentioning

confidence: 99%

Temporal and spatial characteristics of ecological drought in the Inland River Basin and its driving factors

Zhang,

Liu,

Feng

et al. 2024

Preprint

View full text Add to dashboard Cite

Most of the previous studies only used the index of vegetation growth to characterize ecological drought. This study built a Standardized Ecological Water Deficit Index (SEWDI), analysed its space-time evolution characteristics from the perspectives of time, space and spatio-temporal coupling, extractes feature variables by three-dimensional spatio-temporal clustering, analysed typical ecological drought events, the effect of teleconnection factors on ecological drought was investigated using cross wavelet analysis. The reliability of SEWDI was quantitatively evaluated by comparing the r values of Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), Standardized Soil Moisture Index(SSMI), self-calibrating Palmer Drought Severity Index (scPDSI), and SEWDI with Standardized Solar-induced Fluorescence (SSIF) at various scales in time.The results demonstrated that: (1) Ecological drought in the Inland River Basin showed a weakening trend from 1982 to 2015. Seasonal drought showed a wetting trend in the west and a drought trend in the east, and the wetting trend is most evident in spring. (2) Three-dimensional recognition method could extract more detailed characteristic variables of ecological drought events. A total of 133 ecological drought events occurred in the Inland River Basin during 1982–2015, and ecological drought events had a longer duration and intensity after the 21st century. (3) June 2005 to February 2007 was the most severe ecological drought event on record, which lasted for 21 months. The drought intensity was 44.65×105month·km2, and the drought area was 23.79×104km2. The event occurred in June 2005, intensified in September 2005, attenuated in June 2006, re-intensified in August 2006 and re-attenuated (extinction) in February 2007. The trend of drought migration spread to northeast (4) Among the teleconnecting factors, Among the teleconnection factors, ENSO has the most complex and greatest influence on ecological drought in the Inland River Basin. (5) In addition to having a high correlation with other drought indexs, the SEWDI can also more accurately reflect the effects of drought on vegetation.

show abstract

Ensemble projections of fish distribution in response to climate changes in the Yellow and Bohai Seas, China

Cited by 13 publications

References 51 publications

An Ensemble Learning Technique for Predicting Mortality Rate in Red Tilapia (Oreochromis niloticus Linn.) Fingerlings

An Ensemble Learning Technique for Predicting Mortality Rate in Red Tilapia (Oreochromis niloticus Linn.) Fingerlings

Pleistocene Landscape Dynamics Drives Lineage Divergence of a Temperate Freshwater Fish Gobio rivuloides in Coastal Drainages of Northern China

Temporal and spatial characteristics of ecological drought in the Inland River Basin and its driving factors

Contact Info

Product

Resources

About