Mapeo De Los Bosques Tipo Varillal Utilizando Imágenes De Satélite Rapideye en La Provincia Maynas, Loreto, Perú

Abstract

Los bosques tipo varillal albergan una diversidad biológica endémica y varias especies monodominantes, cuando estos bosques se ubican en los pantanos acumulan gran cantidad de carbono en el suelo. El objetivo del presente estudio fue elaborar el mapa de distribución espacial de los varillales sobre arena blanca y sobre pantano en la provincia Maynas, Loreto, a escala 1/100,000, utilizando imágenes de satélite RapidEye y aplicando el algoritmo clasificador de Máxima Verosimilitud (MVS) en el programa ArcGIS versión 10.1. Usamos siete categorías de clasificación en tres escenas RapidEye y definimos las firmas espectrales respectivas basadas en 32 áreas de entrenamiento de muestreo en campo. Se clasificaron las 206 escenas restantes utilizando el algoritmo MVS. El mapa fue validado utilizando 57 puntos de control entre varillales y otros tipos de bosque. Se determinaron 51,025 ha de varillales que representan 0.67% de la superficie de la provincia Maynas, distribuidas principalmente entre las cuencas de los ríos Napo y Nanay y entre la ciudad de Iquitos y la boca del Río Amazonas. Obtuvimos un 91.2% de exactitud en la clasificación de los varillales. Las bandas borde rojo (BR) e ifrarrojo cercano (IRC) de las imágenes RapidEye diferenciaron mejor los varillales de los otros bosques. Se reportan nuevas áreas de bosque de varillales sobre arena blanca y pantanos. En conclusión, las imágenes RapidEye son útiles para clasificar y diferenciar los varillales de otros tipos de bosque, siendo el mapa de los varillales útil para la exploración de nuevos sitios en el estudio de almacenamiento de carbono en el suelo.

Imaging spectroscopy predicts variable distance decay across contrasting Amazonian tree communities

“…Patches of white‐sand forests at this site are frequently smaller than 800 m across, and ecological similarity is likely to be higher within a patch than between patches. In this way, white‐sand forest tree communities may be functioning as meta‐communities, separated by terra firme forests (Adeney, Christensen, Vicentini, & Cohn‐Haft, ; Palacios et al, ). The lack of declining similarity with increasing distance beyond 800 m is consistent with some published distance decay curves for white‐sand forests in this region (Draper, Honorio Coronado, et al, ), while others that have been developed for much broader spatial scales appear to show a more constant decay (García‐Villacorta, Dexter, & Pennington, ; Guevara et al, ), presumably because they include several compositionally distinct floras.…”

Dominant tree species drive beta diversity patterns in western Amazonia

“…To do this, we estimated that the different habitat types account for the following proportions of forested land area in Loreto: Terra firme forest 60%, seasonally flooded forest 20%, palm-swamp forest 10%, white-sand forest 5%, peatland pole forest 5%, based on the best maps of habitat types available in Loreto (Josse et al 2007, Draper et al 2014, Palacios et al 2016, Asner et al 2017. To do this, we estimated that the different habitat types account for the following proportions of forested land area in Loreto: Terra firme forest 60%, seasonally flooded forest 20%, palm-swamp forest 10%, white-sand forest 5%, peatland pole forest 5%, based on the best maps of habitat types available in Loreto (Josse et al 2007, Draper et al 2014, Palacios et al 2016, Asner et al 2017.…”

“…Because some habitat types (e.g., white-sand forests) were overrepresented in our plot data set compared to their distribution in Loreto, it was necessary to standardize our data set so that our list of dominant species was as representative as possible of the wider Loreto region rather than our environmentally biased data set. To do this, we estimated that the different habitat types account for the following proportions of forested land area in Loreto: Terra firme forest 60%, seasonally flooded forest 20%, palm-swamp forest 10%, white-sand forest 5%, peatland pole forest 5%, based on the best maps of habitat types available in Loreto (Josse et al 2007, Draper et al 2014, Palacios et al 2016, Asner et al 2017. We then created a correction factor (actual proportion of habitat type area in plot data set/predicted proportion of habitat type area).…”

Influence of flooding variability on the development of an Amazonian peatland

“…Peatland palm swamps ( aguajales ) represent the majority of the peatlands in the PMFB (43 338 km 2 ) and are dominated by the palm Mauritia flexuosa L.f. (Arecaceae), with an above‐ground carbon (AGC) storage of 75.61 ± 8.44 Mg C ha −1 , below‐ground carbon (BGC) of 22.50 ± 2.41 Mg C ha −1 and soil organic carbon (SOC) of 647.76 ± 80.37 Mg C ha −1 (Honorio Coronado et al, 2021). Peatland pole forests ( varrillal hidromorfico ) account for 7540 km 2 and are characterized by low‐stature forests with limited diversity and thin‐stemmed trees, often harbouring monodominant species such as Pachira nitida (Palacios et al, 2016; Draper et al, 2014; López Gonzales et al, 2020), with an AGC of 78.24 ± 13.17 Mg C ha −1 , a BGC of 21.18 ± 3.08 Mg C ha −1 and SOC of 1,033.75 ± 91.57 Mg C ha −1 (Honorio Coronado et al, 2021). So‐called ‘open peatlands’ ( pantano abierto de turbera , or pantano herbaceo‐arbustivo ), which are named as such due to their significantly more open canopies compared with peatland palm swamps and pole forests, occupy a comparatively smaller area of 4888 km 2 (Draper et al, 2014; Draper, 2015; Hastie et al, 2022) and have an AGC of 41.26 ± 22.71 Mg C ha −1 , a BGC of 15.59 ± 8.72 Mg C ha −1 and SOC of 628.27 ± 131.85 Mg C ha −1 (Honorio Coronado et al, 2021).…”