Assessing the success of soil reclamation programs can be costly and time-consuming due to the cost of traditional soil analytical techniques. One cost-effective tool that has been successfully used to efficiently analyze a range of soil parameters is reflectance spectroscopy. We used reflectance data to analyze natural and reclaimed soils in the field, examining three key soil parameters: soil organic carbon (SOC), total nitrogen (TN), and soil pH. Continuous wavelet transforms combined with machine learning models were used to predict these parameters. Based on the root mean square error (RMSE), R 2 value, and the ratio of performance to deviation (RPD), the Cubist model produced the most accurate models for SOC, TN, and pH. The RMSE, R 2 , and RPD values for SOC were 0.60%, 0.80, and 2.2, respectively. The TN model results were 0.05%, 0.81 and 2.5, and pH model results were 0.44, 0.69 and 1.8. Overall, the optimal model can be used to predict SOC and TN accurately, and improvements in the pH model are needed as pH values less than 6.5 were consistently overpredicted.Key words: reflectance spectroscopy, reclamation, carbon, nitrogen.Résumé : Le coût des techniques classiques d'analyse du sol rend parfois l'efficacité des programmes de restauration du sol aussi onéreuse que laborieuse à évaluer. La spectroscopie par réflectance est une technique rentable dont on s'est servi pour analyser efficacement un éventail de paramètres du sol. Les auteurs ont recouru à la réflectance pour analyser des sols naturels et restaurés sur le terrain, notamment les trois grands paramètres que sont la teneur en carbone organique, la concentration d'azote total et le pH. Ils ont prévu ces paramètres en combinant les transformées d'ondelettes continues et des modèles d'apprentissage machine. D'après l'écart-type, le coefficient R 2 et le ratio performance/écart (RPD), le modèle cubiste est celui qui donne les résultats les plus précis pour le carbone organique du sol, la concentration totale d'azote et le pH. L'écart-type, le coefficient R 2 et le RPD du carbone organique du sol s'établissent respectivement à 0,60 %, 0,80 et 2,2. Les résultats du modèle de l'azote total se chiffrent à 0,05 %, 0,81 et 2,5, tandis que ceux du modèle du pH sont de 0,44, 0,69 et 1,8. En règle générale, on peut se servir du modèle optimal pour prévoir avec précision la concentration de carbone organique et celle d'azote total dans le sol; il faudrait améliorer le modèle du pH, car il y a toujours surestimation des prévisions quand le pH est inférieur à 6,5. [Traduit par la Rédaction]

No abstract

Since 1963 approximately 100 000 oil and gas well pads have been officially certified as reclaimed in Alberta, yet follow-up monitoring programs have not been established to assure provincial environmental desired outcomes such as equivalent land capability are met at these certified well pads. The Ecological Recovery Monitoring Project was designed to determine long-term recovery of environmental quality at certified, reclaimed oil and gas well pads. Our study used soil data from 25 oil and gas well pads that were certified as reclaimed between 1964 and 2011. The objective of this study was to determine if well pad reclamation had long-term legacy effects on soil parameters and land capability as measured with the land suitability rating system (LSRS). We used multivariate covariance generalized linear models to determine effects of recovery time and reclamation on well pad soil parameters pH, bulk density (Db), total organic carbon (TOC), and electrical conductivity relative to reference soils. We also determined effects of reclamation on capability ratings using the LSRS. We concluded that soil quality and LSRS indices on well pads were lower than references due to adverse effects on pH, Db, and TOC. Long-term legacy effects on cultivated soils are significant and not well aligned with intended provincial outcomes. Further monitoring of reclaimed sites is strongly recommended.

No abstract

Human actions drive landscape heterogeneity, yet most ecosystem classifications omit the role of human influence. This study explores land use history to inform a classification of forestland of the Tayal Mrqwang indigenous people of Taiwan. Our objectives were to determine the extent to which human action drives landscape heterogeneity. We used interviews, field sampling, and multivariate analysis to relate vegetation patterns to environmental gradients and human modification across 76 sites. We identified eleven forest classes. In total, around 70 % of plots were at lower elevations and had a history of shifting cultivation, terrace farming, and settlement that resulted in alder, laurel, oak, pine, and bamboo stands. Higher elevation mixed conifer forests were least disturbed. Arboriculture and selective harvesting were drivers of other conspicuous forest patterns. The findings show that past land uses play a key role in shaping forests, which is important to consider when setting targets to guide forest management.