TY - JOUR
T1 - Compost Cation Exchange Capacity via Portable X-Ray Fluorescence (PXRF) Spectrometry
AU - Li, Bin
AU - Chakraborty, Somsubhra
AU - Sosa, Maria Fernanda Godoy
AU - Kusi, Nana Yaw O.
AU - Weindorf, David C.
N1 - Publisher Copyright:
© 2018, © 2018 Taylor & Francis Group, LLC.
PY - 2018/10/2
Y1 - 2018/10/2
N2 - Compost is a valuable organic amendment which affords substantive fertility to soils where applied. A common component of compost fertility is cation exchange capacity (CEC), which has traditionally been determined via standard wet chemistry laboratory methods. This research utilized portable X-ray fluorescence (PXRF) spectrometry to evaluate 74 compost samples from the USA and Canada. PXRF elemental data were used for predicting compost CEC via random forest (RF) regression. Comparison between laboratory-determined vs. PXRF predicted CEC produced the following relationships: R 2 =0.90, RMSE = 5.41 meq 100 g −1 (model calibration) and R 2 =0.60, RMSE = 8.07 meq 100 g −1 (model validation). A key advantage of this technique is that the same data used for CEC prediction can also yield insight into other compost parameters of interest such as heavy metal content, plant essential nutrient content, salinity, and pH. Taken collectively, the PXRF approach can provide rapid, on-site analysis of compost which was previously not feasible with conventional methods. Our initial study has established the viability of PXRF for compost CEC determination, with further development on a wider array of feedstocks suggested for future study.
AB - Compost is a valuable organic amendment which affords substantive fertility to soils where applied. A common component of compost fertility is cation exchange capacity (CEC), which has traditionally been determined via standard wet chemistry laboratory methods. This research utilized portable X-ray fluorescence (PXRF) spectrometry to evaluate 74 compost samples from the USA and Canada. PXRF elemental data were used for predicting compost CEC via random forest (RF) regression. Comparison between laboratory-determined vs. PXRF predicted CEC produced the following relationships: R 2 =0.90, RMSE = 5.41 meq 100 g −1 (model calibration) and R 2 =0.60, RMSE = 8.07 meq 100 g −1 (model validation). A key advantage of this technique is that the same data used for CEC prediction can also yield insight into other compost parameters of interest such as heavy metal content, plant essential nutrient content, salinity, and pH. Taken collectively, the PXRF approach can provide rapid, on-site analysis of compost which was previously not feasible with conventional methods. Our initial study has established the viability of PXRF for compost CEC determination, with further development on a wider array of feedstocks suggested for future study.
UR - http://www.scopus.com/inward/record.url?scp=85055114226&partnerID=8YFLogxK
U2 - 10.1080/1065657X.2018.1522280
DO - 10.1080/1065657X.2018.1522280
M3 - Article
AN - SCOPUS:85055114226
SN - 1065-657X
VL - 26
SP - 271
EP - 278
JO - Compost Science and Utilization
JF - Compost Science and Utilization
IS - 4
ER -