Lichens as spatially transferable bioindicators for monitoring nitrogen pollution

Delves, Jay; Lewis, Jason; Ali, Niaz; Asad, Saeed A; Chatterjee, Sudipto; Crittenden, P. D.; Jones, Matthew R.; Kiran, Aysha; Pandey, Bishnu Prasad; Reay, Dave; Sharma, Subodh; Tshering, Dendup; Weerakoon, Gothamie; Dijk, Netty van; Sutton, Mark A.; Wolseley, Patricia W; Ellis, Christopher J.

doi:10.1016/j.envpol.2023.121575

Cited by 4 publications

(1 citation statement)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent decades, ecological and climate changes in the caribou habitat, such as the expansion of shrublands and managed forests, have reduced the abundance of lichen in caribou ranges [4][5][6]. The increasing frequency, pollution, severity, and spatial extent of forest fires in the boreal region have also impacted the availability of lichen, as lichen requires decades to fully recover [3][4][5]7]. Mapping and monitoring lichen availability are thus critical for informing sustainable land management policy and caribou recovery plans.…”

Section: Introductionmentioning

confidence: 99%

Dense neural network outperforms other machine learning models for scaling-up lichen cover maps in Eastern Canada

Richardson,

Knudby,

Chen

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

Lichen mapping is vital for caribou management plans and sustainable land conservation. Previous studies have used random forest, dense neural network, and convolutional neural network models for mapping lichen coverage. However, to date, it is not clear how these models rank in this task. In this study, these machine learning models were evaluated on their ability to predict lichen percent coverage in Sentinel-2 imagery in Québec and Labrador, Canada. The models were trained on 10-m resolution lichen coverage (%) maps created from 20 drone surveys collected in July 2019 and 2022. The dense neural network achieved a higher accuracy than the other two, with a reported mean absolute error of 5.2% and an R2 of 0.76. By comparison, the random forest model returned a mean absolute error of 5.5% (R2: 0.74) and the convolutional neural network had a mean absolute error of 5.3% (R2: 0.74). A regional lichen map was created using the trained dense neural network and a Sentinel-2 imagery mosaic. There was greater uncertainty on land covers that the model was not exposed to in training, such as mines and deep lakes. While the dense neural network requires more computational effort to train than a random forest model, the 5.9% performance gain in the test pixel comparison renders it the most suitable for lichen mapping. This study represents progress toward determining the appropriate methodology for generating accurate lichen maps from satellite imagery for caribou conservation and sustainable land management.

show abstract

Section: Introductionmentioning

confidence: 99%

Dense neural network outperforms other machine learning models for scaling-up lichen cover maps in Eastern Canada

Richardson,

Knudby,

Chen

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

show abstract

Changes in epiphytic lichen diversity along the urban-rural gradient before, during, and after the acid rain period

Gauslaa

2024

Biodivers Conserv

View full text Add to dashboard Cite

Spatial variations in epiphytic macrolichen richness in the city of Oslo were assessed annually 1973–2019. These observations were contrasted with earlier published data from 1930 to examine how long-term changes in species richness of functional groups track different stages of air pollution regimes. From 1930 to the 1970’s, representing the early surge and late peak of S-deposition, the lichen-deficient urban area remained largely unaltered. Epiphytic macrolichen richness in the surrounding zone declined and changed from a mix of nitrophytes and acidophytes in 1930 when agriculture was still present to a dominance of acidophytes in the 1970’s shortly after the acid rain peak. The subsequent 1980-2019-period marked by significantly lower S-emissions, and weakly decreasing N-deposition, experienced a shift from acidophytes to nitrophytes, following the successful control of acid rain. This underscores the role of pH as a contributing determinant of the strong nitrophyte recolonization. While successive pollution regimes shaped functional group-specific changes in lichen richness over the past 90 years, continuous rain in autumn 2000 led to sudden temporal lichen dieback across the urban-to-rural gradient, delaying lichen recovery after the acid rain period by approximately 5 years for nitrophytes and over 15 years for acidophytes. Epiphytic lichen richness never returned to the high levels seen in 1930, even in the outer parts of the urban-rural gradient and despite the reduction in S-deposition. Excess N impedes effective establishment of acidophytic lichens and prevents full recovery of the former diversity.

show abstract

Effect of host tree traits on the abundance of epiphytic lichens in a tropical Atlantic rainforest

de Freitas,

Ribeiro

2024

COMMUNITY ECOLOGY

View full text Add to dashboard Cite

Lichens as spatially transferable bioindicators for monitoring nitrogen pollution

Cited by 4 publications

References 75 publications

Dense neural network outperforms other machine learning models for scaling-up lichen cover maps in Eastern Canada

Dense neural network outperforms other machine learning models for scaling-up lichen cover maps in Eastern Canada

Changes in epiphytic lichen diversity along the urban-rural gradient before, during, and after the acid rain period

Effect of host tree traits on the abundance of epiphytic lichens in a tropical Atlantic rainforest

Contact Info

Product

Resources

About