Using a conceptual Bayesian network to investigate environmental management of vegetable production in the Lake Taihu region of China

David M Nash, David K Waters, Andres Buldu, Yuming Wu, Yaping Lin, Weiqiu Yang, Yuzhi Song, Jianhua Shu, Wei Qin, Murray C Hannah

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Vegetable farms are one of many nitrogen (N) sources adversely affecting Lake Taihu in eastern China. Given the lack of quantitative "cause and effect" relationships and data relating to these systems, we developed a conceptual Bayesian network to investigate and demonstrate causal relationships and the effects of different mitigation strategies on N exports from vegetable farms in the Lake Taihu region. Structurally, the network comprised one primary transport factor, one primary source factor and three post-mobilisation strategies, and three output factors.In general the network suggests that N exports are more sensitive to transport factors (i.e. runoff volumes) than source factors (i.e. fertiliser application rates) although the cumulative effects of excessive fertiliser were not considered. Post-mobilisation mitigations such as wetlands and ecoditches appear to be particularly effective in decreasing N exports however their implementation on a regional scale may be limited by land availability. While optimising N inputs would be prudent, the network suggests that better irrigation practice, including improved irrigation scheduling, using less imported water and optimising rainfall utilisation would be more effective in achieving environmental goals than simply limiting N supply. •There is little quantitative data describing N exports from Chinese vegetable farms.•Conceptual equations were used to develop a semi-quantitative Bayesian Network.•The Conceptual Bayesian Network is used to compare a range of mitigation strategies. © 2013 Elsevier Ltd.
Original languageEnglish
Pages (from-to)170-181
Number of pages12
JournalEnvironmental Modelling and Software
Volume46
DOIs
Publication statusPublished - 2013
Externally publishedYes

Fingerprint

Environmental management
Vegetables
Bayesian networks
Fertilizers
fertilizer application
environmental management
vegetable
Lakes
lake
mitigation
Farms
farm
mobilization
Irrigation
irrigation
Wetlands
Runoff
Rain
wetland
rate

Cite this

Nash, David M ; Waters, David K ; Buldu, Andres ; Wu, Yuming ; Lin, Yaping ; Yang, Weiqiu ; Song, Yuzhi ; Shu, Jianhua ; Qin, Wei ; Hannah, Murray C. / Using a conceptual Bayesian network to investigate environmental management of vegetable production in the Lake Taihu region of China. In: Environmental Modelling and Software. 2013 ; Vol. 46. pp. 170-181.
@article{f27459d9d09f4ca1bad46f54d82d6581,
title = "Using a conceptual Bayesian network to investigate environmental management of vegetable production in the Lake Taihu region of China",
abstract = "Vegetable farms are one of many nitrogen (N) sources adversely affecting Lake Taihu in eastern China. Given the lack of quantitative {"}cause and effect{"} relationships and data relating to these systems, we developed a conceptual Bayesian network to investigate and demonstrate causal relationships and the effects of different mitigation strategies on N exports from vegetable farms in the Lake Taihu region. Structurally, the network comprised one primary transport factor, one primary source factor and three post-mobilisation strategies, and three output factors.In general the network suggests that N exports are more sensitive to transport factors (i.e. runoff volumes) than source factors (i.e. fertiliser application rates) although the cumulative effects of excessive fertiliser were not considered. Post-mobilisation mitigations such as wetlands and ecoditches appear to be particularly effective in decreasing N exports however their implementation on a regional scale may be limited by land availability. While optimising N inputs would be prudent, the network suggests that better irrigation practice, including improved irrigation scheduling, using less imported water and optimising rainfall utilisation would be more effective in achieving environmental goals than simply limiting N supply. •There is little quantitative data describing N exports from Chinese vegetable farms.•Conceptual equations were used to develop a semi-quantitative Bayesian Network.•The Conceptual Bayesian Network is used to compare a range of mitigation strategies. {\circledC} 2013 Elsevier Ltd.",
author = "Nash, {David M} and Waters, {David K} and Andres Buldu and Yuming Wu and Yaping Lin and Weiqiu Yang and Yuzhi Song and Jianhua Shu and Wei Qin and Hannah, {Murray C}",
year = "2013",
doi = "10.1016/j.envsoft.2013.03.008",
language = "English",
volume = "46",
pages = "170--181",
journal = "Environmental Software",
issn = "1364-8152",
publisher = "Elsevier BV",

}

Using a conceptual Bayesian network to investigate environmental management of vegetable production in the Lake Taihu region of China. / Nash, David M; Waters, David K; Buldu, Andres; Wu, Yuming; Lin, Yaping; Yang, Weiqiu; Song, Yuzhi; Shu, Jianhua; Qin, Wei; Hannah, Murray C.

In: Environmental Modelling and Software, Vol. 46, 2013, p. 170-181.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Using a conceptual Bayesian network to investigate environmental management of vegetable production in the Lake Taihu region of China

AU - Nash, David M

AU - Waters, David K

AU - Buldu, Andres

AU - Wu, Yuming

AU - Lin, Yaping

AU - Yang, Weiqiu

AU - Song, Yuzhi

AU - Shu, Jianhua

AU - Qin, Wei

AU - Hannah, Murray C

PY - 2013

Y1 - 2013

N2 - Vegetable farms are one of many nitrogen (N) sources adversely affecting Lake Taihu in eastern China. Given the lack of quantitative "cause and effect" relationships and data relating to these systems, we developed a conceptual Bayesian network to investigate and demonstrate causal relationships and the effects of different mitigation strategies on N exports from vegetable farms in the Lake Taihu region. Structurally, the network comprised one primary transport factor, one primary source factor and three post-mobilisation strategies, and three output factors.In general the network suggests that N exports are more sensitive to transport factors (i.e. runoff volumes) than source factors (i.e. fertiliser application rates) although the cumulative effects of excessive fertiliser were not considered. Post-mobilisation mitigations such as wetlands and ecoditches appear to be particularly effective in decreasing N exports however their implementation on a regional scale may be limited by land availability. While optimising N inputs would be prudent, the network suggests that better irrigation practice, including improved irrigation scheduling, using less imported water and optimising rainfall utilisation would be more effective in achieving environmental goals than simply limiting N supply. •There is little quantitative data describing N exports from Chinese vegetable farms.•Conceptual equations were used to develop a semi-quantitative Bayesian Network.•The Conceptual Bayesian Network is used to compare a range of mitigation strategies. © 2013 Elsevier Ltd.

AB - Vegetable farms are one of many nitrogen (N) sources adversely affecting Lake Taihu in eastern China. Given the lack of quantitative "cause and effect" relationships and data relating to these systems, we developed a conceptual Bayesian network to investigate and demonstrate causal relationships and the effects of different mitigation strategies on N exports from vegetable farms in the Lake Taihu region. Structurally, the network comprised one primary transport factor, one primary source factor and three post-mobilisation strategies, and three output factors.In general the network suggests that N exports are more sensitive to transport factors (i.e. runoff volumes) than source factors (i.e. fertiliser application rates) although the cumulative effects of excessive fertiliser were not considered. Post-mobilisation mitigations such as wetlands and ecoditches appear to be particularly effective in decreasing N exports however their implementation on a regional scale may be limited by land availability. While optimising N inputs would be prudent, the network suggests that better irrigation practice, including improved irrigation scheduling, using less imported water and optimising rainfall utilisation would be more effective in achieving environmental goals than simply limiting N supply. •There is little quantitative data describing N exports from Chinese vegetable farms.•Conceptual equations were used to develop a semi-quantitative Bayesian Network.•The Conceptual Bayesian Network is used to compare a range of mitigation strategies. © 2013 Elsevier Ltd.

U2 - 10.1016/j.envsoft.2013.03.008

DO - 10.1016/j.envsoft.2013.03.008

M3 - Article

VL - 46

SP - 170

EP - 181

JO - Environmental Software

JF - Environmental Software

SN - 1364-8152

ER -