Publication:
Cartografía regional de suelos salinos y sódicos en el Valle de Lajas, suroeste de Puerto Rico

dc.contributor.advisor Sotomayor Ramírez, David
dc.contributor.author Castro Chacón, José Pablo
dc.contributor.college College of Agricultural Sciences en_US
dc.contributor.committee Pérez Alegría, Luis R.
dc.contributor.committee Martínez Rodríguez, Gustavo
dc.contributor.department Department of Crops and Agro-Environmental Sciences en_US
dc.contributor.representative Ramírez Durand, Lillian
dc.date.accessioned 2021-07-14T18:31:08Z
dc.date.available 2021-07-14T18:31:08Z
dc.date.issued 2021-07
dc.description.abstract Three methods of prediction and data extrapolation were developed: (i) simple and (ii) multiple linear regression models and (iii) artificial neural networks, to create regional saline and sodic soil maps from effective soil electrical conductivity (ECe) and sodium adsorption ratio (SARe) maps, measured at regional scale by Bonnet and Brenes (1958) and at field scale by Álvarez (2021). The predictor variables were (i) elevation, (ii) slopes, (iii) terrain curvature, (iv) direction, and (v) flow of surface water, (vi) mean annual precipitation, (vii) age of parent material, (viii)geomorphology, (ix) groundwater flows, (x) land uses, and (xi) soil series maps, multispectral images from (xii) Landsat 5, (xiii) Landsat 8, (xiv) Sentinel 2A, and (xv) aerial photos. The artificial neural networks method was the best ECe and SARe predictive model. A total of 7878 ha were modeled in 2020, at depth 0 to 60 cm, 69% were normal soils, 26.7% were saline soils, and 4.3% were saline-sodic soils, according to USSL (1954) classification. A tendency of salt and sodium accumulation was observed from lower to higher elevations in the alluvial plains. Between 1958 and 2020 at depth 0 to 60 cm, the data tendency showed an increase in normal soils from 52.1% to 68.9% and in saline soils from 9.9% to 26.7%, but a decrease in saline-sodic soils from 29.7% to 4.32%. Sodic soils were identified in 7% of the empirical samples for the year 2020, but the models did not show sodic soils at a scale of 1:20,000. Twelve soil patterns distributed in 751 ha located in different positions of the landscape were identified: (i) normal tumors, (ii) pond tumors, (iii) melon hole tumors, (iv) stony tumors, (v) soils with vertical cracking, (vi) depressions, (vii) hay affected by salts, (viii) wetlands, (ix) springs, (x) surface salts, (xi) outcrops plants, (xii) and an anthropic sulfur deposit. en_US
dc.description.abstract Se desarrollaron tres métodos de predicción y extrapolación de datos: (i) modelos de regresión lineal simple, (ii) múltiple y (iii) redes neuronales artificiales, para crear mapas regionales de suelos salinos y sódicos a partir de la conductividad eléctrica efectiva del suelo (ECe) y la relación de adsorción de sodio (SARe), medidos a escala regional por Bonnet y Brenes (1958) y a escala de campo por Álvarez (2021). Las variables predictoras fueron (i) elevación, (ii) pendientes, (iii) curvatura del terreno, (iv) dirección y (v) flujo de agua superficial, (vi) precipitación media anual, (vii) edad del material parental, (viii) geomorfología, (ix) flujos de agua subterránea, (x) usos de la tierra, (xi) mapas de series de suelos, imágenes multiespectrales de (xii) Landsat 5, (xiii) Landsat 8, (xiv) Sentinel 2A y (xv) fotos aéreas. El mejor modelo predictivo de ECe y SARe se obtuvo al utilizar redes neuronales artificiales. Se modelaron 7878 ha en 2020, a una profundidad de 0 a 60 cm. El 69% de los suelos resultaron normales, el 26.7% fueron suelos salinos y el 4.3% fueron suelos salino-sódicos, según la clasificación por USSL (1954). En las llanuras aluviales se observó mayor acumulación de sal y sodio conforme se desciende en altitud. Entre 1958 y 2020 a una profundidad de 0 a 60 cm, la tendencia de los datos mostró un aumento en suelos normales de 52.1% a 68.9% y en suelos salinos de 9.9% a 26.7%, pero una disminución en suelos salino-sódicos de 29.7% a 4.32 %. Se identificaron suelos sódicos en el 7% de las muestras empíricas para el año 2020. Sin embargo, los modelos cartográficos no mostraron suelos sódicos en una escala de 1: 20.000. Se identificaron doce patrones de suelo distribuidos en 751 ha ubicadas en diferentes posiciones del paisaje: (i) tumores normales, (ii) tumores de estanques, (iii) tumores de agujero de melón, (iv) tumores rocosos, (v) suelos con agrietamiento vertical, (vi) depresiones, (vii) heno afectado por sales, (viii) humedales, (ix) manantiales, (x) sales superficiales, (xi) afloramientos de plantas, (xii) y un depósito antrópico de azufre. en_US
dc.description.graduationSemester Spring en_US
dc.description.graduationYear 2022 en_US
dc.description.sponsorship United States Department of Agriculture (USDA); USDA-NCRS-Mayaguez MLRA Soil Survey Office; National Institute of Food and Agriculture (NIFA); Hispanic Service Institutions (HSI) Grant Program, Award No. 2016-38422- 25542 North Dakota State University and farm staff for the help and tools provided during the investigation en_US
dc.identifier.uri https://hdl.handle.net/20.500.11801/2806
dc.language.iso es en_US
dc.rights Attribution-NonCommercial 4.0 International *
dc.rights.holder (c) 2021 José Pablo Castro Chacón en_US
dc.rights.uri http://creativecommons.org/licenses/by-nc/4.0/ *
dc.subject Digital Soil Mapping en_US
dc.subject Sodic soils en_US
dc.subject remote sensing en_US
dc.subject geographic information systems en_US
dc.subject Soil Salinity en_US
dc.subject saline soils en_US
dc.subject.lcsh Soil mapping - Puerto Rico en_US
dc.subject.lcsh Soils, Salts in - Puerto Rico en_US
dc.subject.lcsh Soils - Sodium content - Puerto Rico en_US
dc.subject.lcsh Digital soil mapping - Puerto Rico en_US
dc.subject.lcsh Geographic information systems en_US
dc.title Cartografía regional de suelos salinos y sódicos en el Valle de Lajas, suroeste de Puerto Rico en_US
dc.title.alternative Regional mapping of saline and sodic soils in the Lajas Valley, southwestern Puerto Rico en_US
dc.type Thesis en_US
dspace.entity.type Publication
thesis.degree.discipline Soil Sciences en_US
thesis.degree.level M.S. en_US
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
AGSM_CastroChaconJP_2021.pdf
Size:
15.44 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description: