Comparison of multiple statistical techniques to predict soil phosphorus

Hosseini, Mehdi; Agereh, Sareh Rajabi; Khaledian, Yones; Zoghalchali, H Jafarzadeh; Brevik, Eric C.; Naeini, S. A. R. Movahedi

doi:10.1016/j.apsoil.2017.02.011

Cited by 38 publications

(17 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…, Hosseini et al. ) have emphasized smaller quantitative differences in predictive power, not the stark qualitative differences in variable importance and directions of effects we see here. In spite of calls for more careful consideration, MRMI continues to be used in an off‐the‐shelf manner by many ecologists, perhaps because tools such as PLSR that can be applied where MRMI is inappropriate, e.g., data sets with high collinearity, are still relatively unfamiliar.…”

Section: Discussionmentioning

confidence: 43%

“…Simulation studies have shown that PLSR is more robust than MRMI, in the sense that it is better at identifying the correct causal variables, especially when sample sizes are low and/or predictor variables are correlated (Selwood et al 2015). However, past comparisons of PLSR and MRMI (Carrascal et al 2009, Hosseini et al 2017) have emphasized smaller quantitative differences in predictive power, not the stark qualitative differences in variable importance and directions of effects we see here. In spite of calls for more careful consideration, MRMI continues to be used in an off-the-shelf manner by many ecologists, perhaps because tools such as PLSR that can be applied where MRMI is inappropriate, e.g., data sets with high collinearity, are still relatively unfamiliar.…”

Section: Discussionmentioning

confidence: 70%

See 1 more Smart Citation

Why are monarch butterflies declining in the West? Understanding the importance of multiple correlated drivers

Crone

Pelton

Brown

et al. 2019

Ecological Applications

View full text Add to dashboard Cite

Understanding the factors associated with declines of at‐risk species is an important first step in setting management and recovery targets. This step can be challenging when multiple aspects of climate and land use are changing simultaneously, and any or all could be contributing to population declines. We analyzed population trends of monarch butterflies in western North America in relation to likely environmental drivers. Unlike the larger eastern monarch population, past analyses of western monarchs have only evaluated the importance of climate (i.e., not land use) factors as drivers of abundance. We used partial least squares regression (PLSR) to evaluate the potential importance of changes in land use and climate variables. Trends in western monarch abundance were more strongly associated with land use variables than climate variables. Conclusions about importance of climate and land use variables were robust to changes in PLSR model structure. However, individual variables were too collinear to unambiguously separate their effects. We compared these conclusions to the more widely used technique of multiple regression, followed by multi‐model inference (MRMI). Naïve interpretation of MRMI results could be misleading, if collinearity were not taken into account. MRMI was also highly sensitive to variation in model construction. Our results suggest a two‐pronged approach to monarch conservation, specifically, starting efforts now to restore habitat, while also using experiments to more clearly delineate separate effects of climate and land use factors. They also demonstrate the utility of PLSR, a technique that is growing in use but is still relatively under‐appreciated in conservation biology.

show abstract

Section: Discussionmentioning

confidence: 43%

Section: Discussionmentioning

confidence: 70%

Why are monarch butterflies declining in the West? Understanding the importance of multiple correlated drivers

Crone

Pelton

Brown

et al. 2019

Ecological Applications

View full text Add to dashboard Cite

show abstract

“…There are a number of studies on statistical methods to estimate soil parameters and assist in environmental management (i.e., McBratney and Odeh, 1997;Chen et al, 2014;Khaledian et al, 2016a;Hosseini et al, 2017) including CEC using general linear models (Seybold et al, 2005), multiple linear regression (Shabani and Norouzi, 2015;Khaledian et al, 2016b), PLS and stepwise regression (Khaledian et al, 2016a), adaptive network-based fuzzy inference system (ANFIS) and artificial neural networks (ANNs) , and genetic expression programming (GEP) and multivariate adaptive regression splines (MARS) (Emamgolizadeh et al, 2015). However, these studies have not modeled CEC using advanced statistical methods such as algorithms (particle swarm optimization (PSO) and monotone regression tool (MONANOVA)) and compared the performance of these methods, which would represent an advance in the current knowledge.…”

Section: Introductionmentioning

confidence: 99%

Modeling soil cation exchange capacity in multiple countries

Khaledian

Brevik

Pereira

et al. 2017

CATENA

Self Cite

View full text Add to dashboard Cite

A B S T R A C TCation exchange capacity (CEC), as an important indicator for soil quality, represents soil's ability to hold positively charged ions. We attempted to predict CEC using different statistical methods including monotone analysis of variance (MONANOVA), artificial neural networks (ANNs), principal components regressions (PCR), and particle swarm optimization (PSO) in order to compare the utility of these approaches and identify the best predictor. We analyzed 170 soil samples from four different nations (USA, Spain, Iran and Iraq) under three land uses (agriculture, pasture, and forest). Seventy percent of the samples (120 samples) were selected as the calibration set and the remaining 50 samples (30%) were used as the prediction set. The results indicated that the MONANOVA (R 2 = 0.82 and Root Mean Squared Error (RMSE) = 6.32) and ANNs (R 2 = 0.82 and RMSE = 5.53) were the best models to estimate CEC, PSO (R 2 = 0.80 and RMSE = 5.54) and PCR (R 2 = 0.70 and RMSE = 6.48) also worked well and the results were very similar to each other. While the most influential variables for the various countries and land uses were different and CEC was affected by different variables in different situations, clay (positively correlated) and sand (negatively correlated) were the most influential variables for predicting CEC for the entire data set. Although the MANOVA and ANNs provided good predictions of the entire dataset, PSO gives a formula to estimate soil CEC using commonly tested soil properties. Therefore, PSO shows promise as a technique to estimate soil CEC. Establishing effective pedotransfer functions to predict CEC would be productive where there are limitations of time and money, and other commonly analyzed soil properties are available.

show abstract

“…Rodrigo Comino et al (2015a;2015b), using a small portable rainfall simulator, confirmed high infiltration coefficients (near 100%) in this region and small peaks in runoff coefficient during this period with negligible suspended sediment loads values. The alteration of the natural hydrological dynamics due to soil tillage, especially in the young vineyard, can imply several problems regarding solute transport, nutrients and soil losses, landslides due to piping processes, formation of rills and ephemeral gullies, degradation of roots, and decreased productivity (Biddoccu et al, 2013;Bogunović et al, 2016;Hosseini et al, 2017).…”

Section: Discussionmentioning

confidence: 99%

“…The introduction of new plantations of vines and an increase in the occurrence of extreme rainfall events are contributing to soil degradation processes on German hillslope vineyards. Vegetation cover protects soil against erosion runoff generation ( Martínez-Casasnovas et al, 2009;Ruiz-Colmenero et al, 2013), but also it improves soil quality (Hosseini et al, 2017;Salomé et al, 2016). This is relevant as soil can act as a natural filter for water and manage biogeochemical cycles in a world affected by global changes and searching for sustainable development (Keesstra et al, 2012).…”

Section: Introductionmentioning

confidence: 99%

Temporal changes in soil water erosion on sloping vineyards in the Ruwer- Mosel Valley. The impact of age and plantation works in young and old vines

Rodrigo‐Comino

Brings

Iserloh

et al. 2017

Journal of Hydrology and Hydromechanics

Self Cite

View full text Add to dashboard Cite

Abstract:It is well known that rainfall causes soil erosion in sloping German vineyards, but little is known about the effect of age of plantation on soil erosion, which is relevant to understand and design sustainable management systems. In the Ruwer-Mosel valley, young (1-to 4-years) and old (35-to 38-years after the plantation) vineyards were selected to assess soil and water losses by using two-paired Gerlach troughs over three years (2013)(2014)(2015). In the young vineyard, the overland flow was 107 L m -1 and soil loss 1000 g m -1 in the year of the plantation, and decreased drastically over the two subsequent years (19 L m -1 ; 428 g m -1 ). In the old vineyard, soil (from 1081 g m -1 to 1308 g m -1 ) and water (from 67 L m -1 to 102 L m -1 ) losses were 1.2 and 1.63 times higher, respectively, than in the young vineyard.

show abstract

Comparison of multiple statistical techniques to predict soil phosphorus

Cited by 38 publications

References 71 publications

Why are monarch butterflies declining in the West? Understanding the importance of multiple correlated drivers

Why are monarch butterflies declining in the West? Understanding the importance of multiple correlated drivers

Modeling soil cation exchange capacity in multiple countries

Temporal changes in soil water erosion on sloping vineyards in the Ruwer- Mosel Valley. The impact of age and plantation works in young and old vines

Contact Info

Product

Resources

About