Estimating dry matter in african stargrass (Cynodon Nlemfuensis) forage with in-field hyperspectral proximal sensing

dc.creatorBlanco Barrientos, María Gabriela
dc.creatorRojas Downing, María Melissa
dc.creatorRojas González, Alejandra María
dc.creatorElizondo Salazar, Jorge Alberto
dc.date.accessioned2025-01-20T18:24:34Z
dc.date.available2025-01-20T18:24:34Z
dc.date.issued2025
dc.description.abstractHyperspectral remote sensing, applied to forage monitoring, stands as an innovative technology promising to enhance livestock production by providing real-time results at reduced costs. Currently, forage quality testing in Costa Rica is predominantly conducted using destructive and time-consuming methods such as wet chemistry or laboratory spectrometry. This study aims to estimate the dry matter content of African Star Grass (Cynodon nlemfuensis) at the Alfredo Volio Mata Experimental Station (EEAVM) employing field spectroscopy with a multitemporal approach and a partial least squares regression (PLSR) mathematical model. Data were collected from two experimental plots managed for mowing and grazing at the EEAVM of the University of Costa Rica from June to September 2022. During this period, two growth cycles were considered for mowing management and three cycles for grazing. Forage quality tests were conducted using wet chemistry at the EEAVM bromatology laboratory to provide a reference dataset to create the model. Additionally, auxiliary data such as forage height and biomass were collected to correlate with spectral data and identify outliers in the data set. Spectral data were acquired using an ASD FieldSpec4 spectroradiometer with 1 nm resolution and a spectral range of 350-2,500 nm, corresponding to visible and near-infrared. The ChemFlow Galaxy platform was used to develop 18 dry matter estimation models for forage applying PLSR to two pretreatment combinations and different spectral eliminations on three data sets corresponding to the ones collected on the management plots (mowing, grazing, and both combined). Results revealed that dry matter did not exhibit the expected behavior due to the unusual development of forage within the experimental plot. The optimal model for obtaining dry matter quality parameter demonstrated robust statistical performance, with r2cv= 0.81, RMSECV=1.41, r2c=0.89, RMSEC=1.16, and an RPD of 2.31, classified as a highly satisfactory model. These findings have significant implications for forage monitoring and livestock production, as they provide a more efficient and cost-effective method for estimating dry matter content at the EEAVM. This model can be scaled up to a broader scale by replicating the experiment in different regions to collect more data and generate a more robust model, while taking into consideration the use of alternative remote sensing platforms.
dc.description.procedenceUCR::Vicerrectoría de Docencia::Ingeniería::Facultad de Ingeniería::Escuela de Ingeniería de Biosistemas
dc.description.sponsorshipUniversidad de Costa Rica/[737-C0-208]/UCR/Costa Rica
dc.identifier.codproyecto737-C0-208
dc.identifier.doihttps://doi.org/10.1016/j.atech.2024.100731
dc.identifier.issn2772-3755
dc.identifier.urihttps://hdl.handle.net/10669/100499
dc.language.isoeng
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 Internationalen
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/4.0/
dc.sourceSmart Agricultural Technology
dc.subjectHYPERSPECTRAL
dc.subjectNEAR-INFRARED
dc.subjectSPECTROSCOPY
dc.subjectFORAGE QUALITY
dc.titleEstimating dry matter in african stargrass (Cynodon Nlemfuensis) forage with in-field hyperspectral proximal sensing
dc.typeartículo original

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
Estimating Dry Matter in African Stargrass (Cynodon nlemfuensis) Forage with In-Field Hyperspectral Proximal Sensing
Tamaño:
1.91 MB
Formato:
Adobe Portable Document Format

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
3.5 KB
Formato:
Item-specific license agreed upon to submission
Descripción: