https://orcid.org/0000-0002-1492-6029
Figures
Abstract
Burullus Lagoon, situated in the Nile Delta of Egypt, is a Ramsar-listed wetland of high ecological importance, particularly in relation to its floristic diversity. This study presents a field-based hyperspectral characterization of wetland vegetation with the objective of establishing a reference spectral library to support biodiversity assessment and environmental monitoring. Hyperspectral reflectance measurements were obtained for 41 plant species selected from a total of 63 floristically identified taxa, based on ecological dominance, spatial recurrence across sampling sites (≥3 stands), and suitability for reliable field spectral acquisition. Spectroscopic data were collected from 44 stands representing lagoon shores, islets, and open-water habitats using an ASD FieldSpec spectroradiometer covering the 350–2500 nm spectral range. A set of vegetation indices was applied to evaluate key biophysical and biochemical properties associated with plant vitality, water status, and biomass. The results indicate that the red and near-infrared regions provide the highest discriminatory capability among species, whereas the shortwave infrared region exhibits more limited discriminatory capability. Dominant taxa, including Phragmites australis and Atriplex halimus, displayed elevated near-infrared reflectance, consistent with differences in canopy structure and biochemical composition. Most species showed vegetation index responses broadly indicative of healthy physiological conditions, although interspecific variability suggests contrasting stress responses among taxa. Overall, the study demonstrates the applicability of field-based hyperspectral data for species-level discrimination in wetland environments and delivers a curated spectral library to support biodiversity conservation and long-term ecosystem management at Burullus Lagoon.
Citation: Khdery GA, Shokr MS, Rebouh NY, El-Amier YA (2026) Field-based hyperspectral characterization of wetland plant diversity and vitality in Burullus Lagoon (Nile Delta, Egypt). PLoS One 21(3): e0341891. https://doi.org/10.1371/journal.pone.0341891
Editor: Przemysław Mroczek, Maria Curie-Sklodowska University: Uniwersytet Marii Curie-Sklodowskiej, POLAND
Received: May 6, 2025; Accepted: January 13, 2026; Published: March 4, 2026
Copyright: © 2026 Khdery et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Wetlands are ecologically significant environments due to their unique hydrological characteristics and their role as ecotones linking terrestrial and aquatic ecosystems [1]. Wetlands occur in a wide range of geomorphological settings, including river deltas, coastal and inland lagoons (lakes), intertidal zones, river floodplains, inland depressions, and flat terrains [2]. Egyptian coastal lakes are among the most productive wetland ecosystems worldwide. Ecologists around the world have observed and documented changes affecting these critical environments, which led to the establishment of the Ramsar Convention on Wetlands in 1971. One of the Ramsar locations, Burullus Lagoon Egypt’s deltaic Mediterranean coast, was designated as a natural protectorate in 1998 [3,4]. Burullus Lagoon, located in the Nile Delta of Egypt, is among the country’s most ecologically significant wetland systems. It serves as a vital habitat for numerous bird species and aquatic life and supports local fishing communities [5]. Plants are the primary biological components of lake ecotones. Lake ecotones often host a diverse range of plant species, including aquatic, emergent, and terrestrial vegetation [6]. Examining these plants is crucial for understanding their current status and trends, whether they are improving or deteriorating. It is essential to study these relationships to effectively manage and enhance the productivity of this area [7,8]. Over the years, human activities, pollution, and climate change have posed increasing threats to the lake’s biodiversity. To effectively study, monitor, and manage these challenges, Geographic Information System (GIS) technology has emerged as a critical tool [9]. Previous studies have shown that GIS and Remote Sensing data play significant role in assessing the ecological status of Burullus Lagoon [10–13].
Remote sensing is an efficient method for assessing plant parameters due to its global reach, cost-effectiveness, and non-destructive nature [14]. It acquires data by analyzing the Earth’s reflected and emitted waves, providing valuable information on soil, vegetation, water, and other environmental parameters with minimal effort. Remote sensing enables efficient mapping of vegetation and environmental resources, optimizing the economic potential of soils and vegetation [15]. However, most conventional remote sensing sensors acquire broadband spectral data, which may obscure narrow spectral features. Meaning that valuable information available in narrower spectral bands may be lost [16]. Previous studies have reported the use of remote sensing to assess environmental conditions in Lake Burullus Lagoon [17]. Dewidar [10] analyzed land use and land cover changes and assessed potential future changes following the construction of the international coastal road, which crosses the study area, using a Landsat image from 1984 and 1997. Dewidar and Khedr [11] determined the relationships between water quality parameters and radiance data from Landsat Thematic Mapper (TM) in Burullus Lagoon.
By analyzing multispectral or hyperspectral data, researchers can identify different types of vegetation, track their health, and assess the impact of human activities or climate change on plant cover. A hyperspectral signature refers to the unique spectral profile of an object or material, captured across a wide range of wavelengths, typically beyond the visible spectrum. Hyperspectral imaging collects detailed information by dividing the electromagnetic spectrum into hundreds of narrow bands. The hyperspectral signature of a material reveals how it absorbs, reflects, and emits light at different wavelengths, which can be used to identify and distinguish between different substances, even those that appear similar in the visible spectrum. Previous research has demonstrated the importance of hyperspectral data in providing critical information on the biochemical and biophysical characteristics of vegetation [18–23]. Advances in hyperspectral remote sensing have enhanced the accuracy of information regarding vegetation’s structural, biochemical, and physiological properties [24].
The development of new hyperspectral indices has facilitated significant progress in this field, particularly in measuring biophysical and biochemical parameters [25–27]. Spectral indices, which are mathematical transformations of spectral reflectance, are used to enhance the vegetation signal [28,29]. These hyperspectral indices offer the potential for distinguishing plant species or communities with varying canopy structures and/or biochemical compositions [30]. The vegetation in and around Burullus Lagoon plays a crucial role in maintaining the ecological balance.
Spectral analysis offers a nondestructive way to quantify plant biochemical and structural traits and to discriminate species and plant functional types [31,32]. In floristic and biodiversity studies, imaging spectroscopy links community composition with canopy chemistry and structure, supporting monitoring and conservation applications [32,33]. Vegetation indices derived from reflectance—such as NDVI, SR, and red-edge metrics—are widely used as proxies for canopy vigor, photosynthetic status, and pigment composition [34]. Recent advances also show that reflectance spectroscopy can robustly predict leaf traits across functional groups, reinforcing the physiological basis for index selection [35]. The central aim of this study is to establish a field-based hyperspectral library for Burullus Lagoon species. This library provides a baseline reference for biodiversity assessment, vegetation health evaluation, and long-term environmental monitoring. Specifically, the study investigates the spectroscopic characteristics and spectral reflectance patterns of 41 species, selected out of 63 floristically identified, and examines vegetation indices to evaluate plant vitality and canopy vigor. Accordingly, the main research question addressed in this work is whether hyperspectral signatures can reliably discriminate Burullus Lagoon plant species and their physiological conditions, thereby supporting accurate monitoring and conservation efforts.
2. Materials and methods
2.1. Study area
Burullus Lagoon, with an area of approximately 460 km2, is located in the Kafr El-Sheikh Governorate (30° 22’ – 31° 35’N; 30° 33’ – 31° 08’E). It lies along the eastern bank of the Nile’s Rosetta branch and is situated in the heart of the Nile Delta’s Mediterranean coast (Fig 1). The satellite image was downloaded from the USGS EarthExplorer platform (https://earthexplorer.usgs.gov) for Sentinel-2 imagery acquired in March 2022. Materials from plants, animals, or other natural environments are not collected for the study. Additionally, field site access and research operations did not require special licensing because the study was conducted to take spectroscopic measurements of plants and did not include interactions with restricted species or ecosystems. After Manzala Lake, it is Egypt’s second largest natural lake. The northern part of Kafr El-Sheikh Governorate includes El-Burullus (Baltim) district to the east, El-Hamool, El-Riad, and Sidi Salim districts to the south, and Motobas district to the west [36]. The Egyptian Environmental Affairs Agency has established and manages a network of protected sites throughout Egypt. Burullus Lagoon is one of them. It is a Ramsar site, and it has been designated as an Important Bird Area (IBA) by Birdlife International [7]. A total of forty-four geo-referenced sites were selected to represent the different habitat types (Table S1 in S2 File). The distribution of these sites is from 1 to 22 for the Lake Shores habitat, 23–33 for the Islets habitat, and 34–44 for open water (Fig 1). Fieldwork and spectral measurements were conducted between 10 and 22 March 2022, during late-winter/early-spring conditions. All sampling days experienced clear-sky weather with no rainfall events. Air temperature during measurements ranged between 22 and 28°C and relative humidity between 55 and 65%. All measurements were taken between 10:00 a.m. and 2:00 p.m. local time to minimize variation due to solar angle.
2.2. Floristic composition
Throughout Burullus Lagoon, 44 stands were randomly selected across three habitats: the lake shores, the lake islets, and the open water. These habitats were chosen because they represent the main ecological gradients of Burullus Lagoon —contrasting shoreline reedbeds, insular vegetation, and pelagic zones—thereby capturing the lake’s hydrological and salinity variability, in line with earlier floristic and habitat classifications [37,38]. In each stand, all plant species were recorded in five plots of 25 m2 each, and one representative stand, defined as a relatively homogeneous vegetation patch within each habitat [39] was used to determine species abundance (frequency, IV = 100) according to Westhoff and Van der Maarel [39] and Ellenberg [40]. The taxonomic nomenclature and identification were determined, and chorotypes were defined as the geographical distribution patterns of the recorded taxa, following the classification of Boulos (1999–2005) [41,42] and Tackholm [43]. However, life forms were identified according to Raunkiaer [44]’s system, which categorizes plants based on the position of perennating buds relative to the soil surface. In each stand, five plots of 25 m2 were surveyed, and species frequency and abundance were calculated as the average across all plots within the stand to ensure replication. This stratified random design captured the ecological heterogeneity of the lagoon, minimized sampling bias, and enhanced the statistical robustness of the floristic dataset [45]. Species abundance was calculated by averaging values from five plots within each stand to obtain representative values for frequency and abundance, ensuring that all plots contributed equally to the final calculation [39].
2.3. Spectroscopic measurements
Out of the 63 plant species identified during the floristic survey, a subset of 41 species was selected for hyperspectral reflectance measurements. Species selection was based on ecological relevance, spatial abundance (occurrence in ≥3 stands), and structural distinctiveness, including leaf size, canopy density, and accessibility for reliable spectroradiometric acquisition. Submerged species and morphologically indistinct taxa were excluded due to difficulties in obtaining consistent and accurate field spectra. This stratified selection ensured representative coverage of dominant and spectrally distinguishable vegetation types in Burullus Lagoon, consistent with recommendations for species-level spectral discrimination by Cochrane [46] and Clark et al., [47] on species-level spectral discrimination.
Spectral reflectance measurements were acquired using an ASD FieldSpec spectroradiometer covering the full optical spectral range from 350 to 2500 nm. The instrument provides a nominal spectral resolution of 1 nm, with sampling intervals of 1.4 nm in the 350–1050 nm range and 2 nm in the 1000–2500 nm range. All measurements were conducted under clear-sky conditions between 10:00 and 14:00 local time to minimize variations in solar angle and illumination conditions. The sensor was positioned approximately 50 cm above the canopy using a pistol grip, and measurements were restricted to mature, fully expanded, sunlit leaves to ensure spectral consistency.
For each species, three individuals were sampled to provide biological replication. For each individual, ten successive spectral scans were recorded and averaged to reduce sensor noise. Calibration was performed every 15 minutes using a Spectralon white reference panel. Raw spectra were pre-processed by removing noisy spectral regions at the edges of the spectrum (<400 nm and >2400 nm) and major atmospheric water absorption bands (1350–1450 nm and 1800–1950 nm). A two-step averaging procedure was applied, first at the scan level and subsequently at the individual level, resulting in a single representative spectrum for each species. This approach minimized within-species variability and enhanced spectral reliability for subsequent analyses.
Statistical analyses were conducted to evaluate interspecific spectral differences and classification performance. One-way analysis of variance (ANOVA) followed by Tukey’s Honestly Significant Difference (HSD) test was applied to identify spectral regions exhibiting significant interspecific variation, thereby directly testing the hypothesis of spectral separability among species. Linear Discriminant Analysis (LDA) was employed to assess species-level classification accuracy and to validate the potential of hyperspectral data for discriminating wetland vegetation in Burullus Lagoon. All statistical analyses were performed using JMP Pro software (version 14.0, SAS Institute, Cary, NC, USA).
Vegetation and sediment indices were calculated from the processed spectra using established formulations (Tables 1 and 2). Vegetation indices were selected based on their widespread use in hyperspectral studies and their ability to capture key biophysical and biochemical properties, including canopy water content, senescence, pigment composition, and red-edge dynamics. Sediment indices were included to evaluate the influence of soil salinity and water status on plant spectral responses, thereby integrating vegetation condition with habitat characteristics. The combined use of vegetation and sediment indices has been widely recommended for improving ecological interpretation and species-level discrimination in hyperspectral analyses [65].
This workflow (Fig 2) provides a stepwise summary of the methods applied, linking field sampling, spectral data acquisition, vegetation and soil indices calculation, and statistical analyses for species discrimination and ecological interpretation.
3. Results
3.1. Floristic composition
The floristic diversity in the study area includes 63 species (19 annuals, 2 biennials, and 42 perennials) spread across 44 sites in various habitats of Burullus Lake. These species belong to 50 genera and 30 families (Table 3). The lakeshore habitat is the most floristically diverse, with 51 species, consisting of 31 perennials, 2 biennials, and 18 annual species. The flora of the lake islets habitat comprises 30 species, including 23 perennials, 1 biennial, and 6 annuals. In the open water habitat, 14 species were recorded (12 perennials and 2 annuals). Sampling was conducted during the rainy season, which favors the blooming of numerous annual species. The primary plant families include Chenopodiaceae (13 species), Asteraceae (7 species), and both Poaceae and Cyperaceae (4 species each), collectively accounting for 28 species, or 44.4% of the total species recorded. Juncaceae and Polygonaceae are represented by 3 species each, while the other 24 families were monospecific, collectively contributing 46.03% of the total species (Fig 3).
The broad ecological range of Chenopodiaceae and Asteraceae, spanning saline coastal regions, deserts, temperate grasslands, and disturbed sites, is due to several adaptive traits that enable them to thrive in diverse and often challenging environments (Fig 3). Chenopodiaceae species are typically halophytic, with a high salinity tolerance, allowing them to colonize areas unsuitable for many other plant families [66]. Asteraceae species are highly drought-resistant, and their effective wind dispersal of diaspores aids their survival [67]. Poaceae have developed traits that help them endure and even benefit from grazing and drought, such as an extensive shallow root system that efficiently absorbs soil moisture [68]. Cyperaceae species are particularly adapted to wetlands, marshes, and waterlogged areas, with many sedges thriving in both aquatic and terrestrial environments, making them common in areas with varying water levels [69].
Based on their percentage presence, the recorded species were classified into three main categories: 1) the wide-range distribution class (presence > 50%), consisting of three species: Phragmites australis (95.45%), Typha domingensis (56.82%), and Eichhornia crassipes (54.55%); 2) the moderate distribution class (presence between 25–50%), which includes 11 species such as Echinochloa stagnina and Potamogeton pectinatus (both at 40.91%), Halocnemum strobilceum (34.09%), Ceratophyllum demersum and Cynanchum acutum (both at 31.82%); 3) the narrow-range distribution class (presence < 25%), which includes 49 species, such as Saccharum spontaneum (20.45%), Lemna gibba and Potamogeton crispus (both at 15.91%), Atriplex prostrata (11.36%), and Alternanthera sessilis (6.82%) (Table 3).
The life-form spectrum in the study area and its various habitats is predominantly composed of cryptophytes, which include helophytes, geophytes, and hydrophytes, with additional representation from therophytes, chamaephytes, and hemicryptophytes. Nanophanerophytes are the least represented group across all habitats in the study area (Fig 3b). Cryptophytes account for 32% of the life forms in the overall study area, with values of 23%, 12%, and 16% in lake shores, lake islets, and open-water habitats, respectively. This life form contributed about 20.5%, 20.9%, 26.3%, and 28.57% in the studies of Al-Sodany [70], and, Shaltout et al., [71] respectively. Therophytes represent approximately 20% of the species in the study area and are distributed as 20%, 7%, and 1% in lake Shores, lake Islets, and open water, respectively. Therophytes dominate due to topographical changes and human or animal disturbances [72]. Their ability to survive as seeds during dry periods makes them well-adapted to the region’s arid conditions. These results are consistent with plant life forms in arid environments across the Middle East [73]. Moreover, Chamaephytes make up 10% of the species in the study area, with 6% in Lake Shores and 9% in Lake Islets. Many chamaephytes possess adaptations that allow them to tolerate saline conditions, which are common in coastal and brackish environments like Burullus Lake. Their ability to survive in areas with high salinity provides them a competitive advantage over other plant forms [74].
3.2. Chorological affinities
Egypt serves as a crossroads for floristic components from four phytogeographical regions: African Sudano-Zambian, Asiatic Irano-Turanian, Afro-Asian Sahro-Sindian, and Euro-Afro-Asian Mediterranean [75]. The entire country falls within the Saharo-Arabian belt of the Holarctic floristic region. The chorological analysis of the area indicates that 35 species, approximately 55.55% of the total recorded species, are Mediterranean taxa (Fig 3. 2c). These include 18 plurinational species (28.57%), 16 biregional species (25.39%), and 6 mono regional species (9.52%). Additionally, 23 species, around 36.50% of the total, are worldwide species, categorized as 10 cosmopolitan (15.87%), 4 palaeotropical (6.35%), 6 pantropical (9.52%), and 3 neotropical (4.76%) (Table 3). Other floristic categories are underrepresented, with each chorotype consisting of only a few species. In terms of habitats, Mediterranean elements are most abundant in lake shores (27 taxa), followed by lake islets (20 taxa), and then open water (4 taxa) (Table 3).
3.3. Spectral analysis
3.3.1. Spectroscopic parameters and analysis.
Analysis of the spectroscopic parameters revealed that all plant species exhibit a consistent spectral shape across various spectral zones, particularly in the visible and infrared regions [18,20,22]. The spectral signature shows that the highest reflectance was observed in the NIR zone between (944–1076 nm) while the lowest reflectance was in SWIRII zone from (2400–2462 nm) (Supplementary Data S1 in S1 File). Moderate reflectance was found in visible region from (555–614 nm) and SWIRI zone from (1538–1670 nm) (Fig 4). Phragmites australis and Atriplex halimus exhibited higher spectral reflectance than the other species in all spectral zones, whereas Phragmites australis showed close spectral reflectance pattern to Atriplex halimus. Mesembryanthemum crystallinum showed the lowest reflectance in SWIRII while Juncus rigidus showed the lowest reflectance in NIR. In the visible region, reflectance was largely convergent among species, with limited interspecific differentiation. In visible zone, Cakile maritima showed a relatively different reflectance pattern [76].
3.3.2. Statistical analyses including Tukey’s HSD and LDA.
To address our aim of discriminating taxa and identifying the most informative wavebands, we used band-wise ANOVA/Tukey to test interspecific separability across spectral regions and LDA to quantify species-level classification accuracy. Prior to these analyses, the standard deviation (STDV) of reflectance was computed for each species from its three field replicates to evaluate within-species spectral variability. The complete STDV dataset for all 41 taxa is presented in Table S2 in S2 File and was used to confirm the consistency of spectral replicates and ensure that interspecific differences detected by ANOVA and LDA reflected true taxonomic separation rather than within-species variability. A one-way ANOVA followed by Tukey’s Honestly Significant Difference (HSD) post-hoc tests was conducted across the major spectral regions blue, green, and red (Fig 5) and NIR, SWIR I, and SWIR II (Fig 6) to assess interspecific spectral difference. The results showed that the Red (600–700 nm) and NIR (750–900 nm) bands provided the strongest discriminatory power among species, with highly significant F-values (p < 0.001). In contrast, reflectance differences in the green region (500–600 nm) were generally non-significant, indicating limited value for species-level separation. This non-significance in the green band is expected because reflectance near ~550 nm is dominated by broad chlorophyll absorption and multiple scattering, which reduces interspecific contrast compared with the red edge and NIR. Tukey’s grouping further indicated that dominant taxa such as Phragmites australis, Typha domingensis, and Atriplex halimus formed significantly distinct clusters (p < 0.05), whereas halophytic species such as Suaeda spp. displayed overlapping groups, suggesting partial spectral convergence under saline conditions (Table 4).
Effect sizes (η²) were calculated for each spectral region to complement the ANOVA results and quantify the magnitude of interspecific spectral differences. The Red and NIR regions exhibited very large effect sizes (η² = 0.89 and 0.92), indicating strong discriminatory power, while SWIR-I and SWIR-II showed moderate-to-high effects (η² = 0.68 and 0.86). In contrast, the Green region exhibited a weaker, non-significant effect (η² = 0.49). These values are included in the revised ANOVA Table 4. Linear Discriminant Analysis (LDA) explained 89.4% of the total variance within the first two canonical discriminant functions. The overall classification accuracy reached 86.5% at the species level. Dominant species such as Phragmites australis, Typha domingensis, and Eichhornia crassipes were correctly separated with >90% accuracy, while species with subtler spectral differences (Juncus rigidus, Suaeda monoica) exhibited moderate misclassification. The confusion matrix (Table 5) highlights that dominant taxa were consistently well-separated, whereas morphologically and spectrally similar species showed partial overlap. Misclassifications mainly occurred among morphologically and biochemically similar halophytes (e.g., Juncus spp., Suaeda–Spergularia), consistent with convergent stress-adapted pigment signatures. Validation was performed using leave-one-out cross-validation (LOOCV), which produced stable classification performance with only a slight reduction in overall accuracy. Producer’s accuracy values ranged from 74% to 93%, while user’s accuracy ranged from 71% to 95%, indicating strong separability for dominant species and moderate overlap among morphologically similar halophytes. The confusion matrix further confirmed these trends by showing high correct-classification rates for emergent macrophytes and limited confusion among stress-adapted chenopodiaceous species.
3.3.3. Wavelength of plants samples.
The findings highlight the optimal wavebands suitable for identifying the studied plant species (Table 6). Overall, 12 species were found to have only a single narrow spectral zone (e.g., Arthrocnemum macrostachyum, Cakile maritima, Cyperus alopecuroides, and Tamarix nilotica), which limits their separability to specific wavelengths. In contrast, two species, Atriplex halimus and Spergularia marina, exhibited exceptionally broad spectral ranges that extended across the visible, NIR, and SWIR domains, demonstrating their high diagnostic potential.
Among the spectral regions, the red (600–700 nm) and NIR (750–900 nm) zones were the most informative, covering the largest number of distinctive taxa (23 and 19 species, respectively). By contrast, the SWIR-I and SWIR-II ranges, though less represented overall, were critical for discriminating halophytic species such as Mesembryanthemum crystallinum and Spergularia marina, which showed unique reflectance beyond 2000 nm. The Green zone (500–600 nm) contributed only marginally, with fewer taxa showing separability.
Empty cells in Table 6 indicate species for which no unique diagnostic wavelengths could be identified, reflecting either overlapping spectral patterns with other taxa (e.g., Chenopodium murale, Saccharum spontaneum, Solanum nigrum, Zygophyllum album) or insufficient distinctiveness in their reflectance curves.
These results emphasize that while red and NIR regions remain the most reliable for species-level discrimination in Burullus wetland vegetation, the SWIR bands provide valuable complementary information for halophytes and broad-spectrum species.
3.3.4. Hyperspectral vegetation indices.
A series of vegetation indices derived from spectrometric analysis were employed to assess the vitality of the plants, as detailed in Table 7. The interpretation of vegetation indices provided critical insights into plant physiological status. NDVI values below 0.5 are widely accepted as indicators of stressed or sparse vegetation, while values above 0.6 reflect dense and healthy canopies [77].In this study, most species, such as Atriplex halimus, Tamarix nilotica, and Suaeda pruinosa recorded low NDVI values (<0.5), reflecting moderate stress and reduced biomass. By contrast, Eichhornia crassipes (0.85) and Chenopodium murale (0.76) indicated vigorous growth, consistent with healthy vegetation thresholds [78].
PSRI values between –0.1 and 0.2 are generally associated with healthy vegetation, with higher values reflecting increasing pigment degradation or senescence [49,58]. The species examined here exhibited PSRI values within this healthy range (–0.04 to 0.14). However, species such as Spergularia marina (0.14) and Atriplex halimus (0.12) displayed relatively higher values, suggesting the onset of pigment degradation, consistent with early senescence.
The Moisture Stress Index (MSI) is positively correlated with canopy water stress, with values >0.7 indicating limited leaf water content and higher stress levels [65]. In the present dataset, species such as Tamarix nilotica (0.84), Malva parviflora (0.79), and Phragmites australis (0.77) showed evidence of water stress. In contrast, low MSI values in Suaeda vera (0.23) confirmed a healthy water status.
The Simple Ratio (SR) is another widely used broadband index, where values above 3 are indicative of good canopy condition and higher chlorophyll concentration. In this study, most species recorded SR > 3, confirming generally good physiological condition. Exceptions included Atriplex halimus (1.61) and Tamarix nilotica (2.15), which exhibited lower SR values consistent with their elevated MSI and reduced NDVI, highlighting multi-index agreement in identifying stressed vegetation.
To strengthen ecological interpretation, species-level examples were incorporated based on established spectral thresholds. For instance, high NDVI and low PSRI values in Phragmites australis indicate dense, vigorous reed stands; moderate NDVI in Typha domingensis reflects partial senescence; and low NDVI combined with higher stress indices (MSI, PSRI) in Juncus acutus is consistent with salinity-induced stress and sparse canopy structure [79].
Together, these indices demonstrate that vegetation in Burullus Lagoon is highly heterogeneous, with halophytes such as Atriplex halimus and Tamarix nilotica showing clear signs of stress, while aquatic macrophytes such as Eichhornia crassipes maintained vigorous growth. The integration of NDVI, PSRI, MSI, and SR allowed for cross-validation of stress signals, strengthening the interpretation of hyperspectral measurements. Ecologically, these multi-index patterns indicate spatial mosaics of productivity and stress across habitats, supporting targeted monitoring of invasive hydrophytes and salt-tolerant stands along the lagoon’s salinity and water-level gradients.
Because field measurements were taken under clear-sky, midday conditions, most taxa were sampled during early to mid-growing stages; thus, higher NDVI and chlorophyll-related indices in hydrophytes (e.g., Eichhornia crassipes) and lower values in halophytes (e.g., Atriplex halimus, Tamarix nilotica) likely reflect phenology interacting with water availability and irradiance. This seasonal context helps explain why red-edge indices (RENDVI, REPI) were especially diagnostic, capturing chlorophyll build-up in actively growing canopies while flagging pigment loss in stress-prone taxa.
Analysis of pigment-related indices provided additional biochemical insights into species’ physiological responses. Pigment indices (ARI, CRI) add biochemical context to structural indices: high ARI suggests stress-linked anthocyanins, while high CRI indicates carotenoid-based photoprotection (Table 8). The Anthocyanin Reflectance Index (ARI) revealed notable variation across taxa. Most species recorded ARI1 values close to zero, consistent with limited anthocyanin accumulation. However, Limbarda crithmoides (ARI1 = 3.80), Spergularia marina (3.06), and Mesembryanthemum nodiflorum (1.31) exhibited markedly high ARI values, indicating anthocyanin accumulation as a stress-response mechanism. Elevated ARI is widely recognized as a marker of photoprotection and abiotic stress adaptation [80,81].
The Carotenoid Reflectance Indices (CRI1 and CRI2) highlighted interspecific variability in carotenoid–chlorophyll balance. High CRI values were found in Eichhornia crassipes (CRI1 = 9.07, CRI2 = 8.82) and Limbarda crithmoides (5.87), suggesting robust carotenoid activity, linked to photoprotection under high light stress [78,79]. Conversely, Atriplex halimus (CRI1 = 0.36, CRI2 = 0.68) displayed minimal carotenoid accumulation, consistent with its concurrently low NDVI and high MSI values, indicating vulnerability to stress.
Chlorophyll a & b indices further supported these findings. Species such as Juncus acutus (0.49) and Chenopodium murale (0.55) showed relatively high chlorophyll content, consistent with vigorous growth, whereas Suaeda pruinosa (0.15) and Zygophyllum album (0.18) had very low values, suggesting reduced photosynthetic capacity. Such differences align with known ecological strategies of halophytes, where pigment modulation contributes to salt and drought tolerance [65,82].
Overall, the integration of ARI, CRI, and chlorophyll indices indicates that while several species maintain balanced pigment profiles under Burullus Lagoon conditions, others exhibit pigment-level stress signatures. These biochemical indicators provide a complementary layer of interpretation to structural indices (e.g., NDVI, MSI), highlighting species-specific physiological strategies for stress resilience.
Pigment-related indices also revealed distinct physiological strategies among the surveyed species. High CRI values in Eichhornia crassipes reflected its efficient photoprotective capacity and rapid biomass accumulation typical of floating hydrophytes. In contrast, the markedly elevated ARI values observed in Spergularia marina and Mesembryanthemum nodiflorum indicated enhanced anthocyanin production, a known mechanism supporting tolerance to salinity and intense radiation in coastal and saltmarsh habitats. Species such as Suaeda pruinosa and Zygophyllum album exhibited notably low chlorophyll concentrations, suggesting limited photosynthetic capacity and resource-conservative strategies characteristic of stress-tolerant halophytes. Together, these pigment-based responses provide a coherent ecological interpretation of species performance across the Burullus Lagoon environment.
Table 9 compiles red-edge indices, which provide sensitive indicators of chlorophyll concentration and pigment dynamics across species. The Red Edge Normalized Difference Vegetation Index (RENDVI) was effective in capturing chlorophyll content variation across species (Table 9). Based on Gitelson et al. (2001), healthy vegetation is typically associated with RENDVI values between 0.2–0.9. In our dataset, most species (e.g., Chenopodium murale, 0.55; Echinochloa stagnina, 0.50) fell within this range, indicating adequate chlorophyll and canopy structure. However, species such as Suaeda pruinosa (0.15), Tamarix nilotica (0.18), and Spergularia marina (0.14) showed values below 0.2, reflecting reduced photosynthetic activity and pigment degradation [79,83].
The Modified Red Edge NDVI (MRENDVI) was designed to enhance sensitivity to subtle chlorophyll differences [48]. Thirteen species (e.g., Juncus acutus, 0.37; Chenopodium murale, 0.38; Symphyotrichum squamatum, 0.34) displayed positive MRENDVI values within the 0.2–0.7 healthy range, confirming good canopy condition. Negative values in Atriplex halimus (–0.41) and Atriplex portulacoides (–0.07) suggest reduced pigment absorption, consistent with stress-related physiological adjustments under salinity. These patterns are consistent with the findings of [82],who highlighted the sensitivity of MRENDVI to reduced chlorophyll under stress conditions.
The Modified Red Edge Simple Ratio (MRESR or mSR705) further differentiated taxa. According to Gitelson & Merzlyak [65], values between 2–8 are indicative of healthy photosynthetic function. Higher values in Echinochloa stagnina (6.97) and Cyperus alopecuroides (6.36) indicate strong photosynthetic capacity, whereas species such as, Spergularia marina (1.45), Malva parviflora (1.91), and Mesembryanthemum nodiflorum (1.78) fell below 2, highlighting reduced photosynthetic efficiency.
The Red Edge Position Index (REPI) quantifies shifts in the red edge inflection point, directly linked to leaf chlorophyll content [83]. In this study, REPI values ranged between 664–695 nm, with Spergularia marina showing the lowest value (664 nm) and Cyperus alopecuroides the highest (695 nm). Lower REPI values (e.g., Spergularia marina, 664 nm) indicate reduced chlorophyll content, whereas higher REPI values (e.g., Cyperus alopecuroides, 695 nm) correspond to greater chlorophyll accumulation and photosynthetic potential.
Taken together, the red-edge indices provide ecologically meaningful differentiation among species. Chlorophyll-rich species such as Chenopodium murale and Echinochloa stagnina demonstrated consistently high red-edge values, reflecting vigorous growth and efficient pigment functioning. In contrast, halophytic taxa such as Suaeda pruinosa and Atriplex halimus exhibited reduced red-edge responses, highlighting stress-linked pigment reduction consistent with their drought- and salinity-tolerant strategies. These species-level patterns illustrate how red-edge indices capture variations in photosynthetic efficiency, pigment regulation, and adaptive responses across the Burullus wetland [82] (Table 9).
4. Discussion
Moreover, the ecological patterns observed in the floristic and life-form analyses provide a broader context for interpreting the spectral behavior of the surveyed taxa. The dominance of therophytes and chamaephytes across much of the Burullus wetland reflects a community structure shaped by recurrent disturbances, seasonal water fluctuations, and spatially heterogeneous salinity gradients [69,72,73]. These functional groups typically exhibit rapid life cycles, shallow rooting systems and opportunistic resource use, traits that are consistent with their moderate chlorophyll signals and relatively variable NDVI values. In contrast, the presence of perennial halophytic shrubs within the central and eastern basins—such as Atriplex, Tamarix and Suaeda—indicates long-term adaptation to stable yet stressful saline regimes, which aligns with the lower photosynthetic activity and higher stress-related spectral indices reported in this study. Chorological patterns further support this ecological interpretation: the predominance of Saharo-Arabian and Mediterranean–Irano-Turanian elements suggests that many of the dominant taxa are pre-adapted to arid, saline and nutrient-poor conditions, explaining their conservative water-use traits and distinctive red-edge signatures. Spatial differences between northern, southern and islet habitats also reinforce the role of environmental heterogeneity in driving species assemblages, with hydrophytes occupying deeper, nutrient-rich waters and generating high-chlorophyll spectral profiles, while xerohalophytes colonize elevated saline margins where spectral stress indicators increase noticeably [74].
Spectral indices extend beyond physiological assessment to provide clear ecological meaning for species strategies and Wetland dynamics. For example, species with low NDVI and high MSI values such as Atriplex halimus and Tamarix nilotica reflect stress-tolerant halophytic strategies that persist under salinity and water deficit, contributing to patchy vegetation cover and reduced net primary productivity [71]. Conversely, the very high NDVI recorded for Eichhornia crassipes illustrates vigorous growth, which ecologically translates to invasive dominance that suppresses native biodiversity and alters ecosystem functioning. Pigment indices such as ARI and CRI further highlight adaptive mechanisms, where anthocyanin accumulation (Suaeda monoica) or carotenoid investment (Mesembryanthemum nodiflorum) indicate photoprotective and antioxidant roles under extreme radiation and salinity [84]. Similarly, red-edge indices (RENDVI, REPI) capture subtle variation in chlorophyll content and canopy structure, which ecologically corresponds to differences in photosynthetic efficiency, competitive ability, and community assembly. Collectively, these results emphasize that spectral indices not only quantify plant condition but also serve as proxies for ecological strategies, stress tolerance, and species interactions within the Burullus Lagoon ecosystem. Furthermore, linking these spectral patterns with species-level ecological traits clarifies how dominant taxa maintain physiological performance under contrasting habitat conditions—for instance, hydrophytes with high chlorophyll and carotenoid signals (e.g., Eichhornia crassipes) exhibit competitive dominance, while halophytes with reduced NDVI and elevated stress indices (e.g., Atriplex halimus, Tamarix nilotica) reflect resource-conservative strategies adapted to salinity and water limitation. Recent hyperspectral studies also confirm that integrating red-edge and pigment indices improves the detection of subtle physiological stress under salinity and drought, supporting species-level discrimination in wetland ecosystems [79,82].
5. Conclusion
This study developed a spectral library for Burullus Lagoon by integrating floristic surveys with hyperspectral measurements, covering 41 species across key habitats. The results demonstrate the utility of hyperspectral indices for assessing plant health, stress responses, and ecological dynamics in wetland ecosystems. The spectral library provides a valuable baseline for monitoring vegetation changes and supports conservation and sustainable management of this Ramsar-listed wetland. Future applications can build on this dataset to enhance wetland monitoring under increasing anthropogenic and climate pressures.
Acknowledgments
The authors would like to acknowledge the National Authority of Remote Sensing and Space Science (NARSS) in Cairo, Egypt and this publication has been supported by the RUDN University Scientific Projects Grant System.
References
- 1. Walton CR, Zak D, Audet J, Petersen RJ, Lange J, Oehmke C, et al. Wetland buffer zones for nitrogen and phosphorus retention: Impacts of soil type, hydrology and vegetation. Sci Total Environ. 2020;727:138709. pmid:32334232
- 2. Ferreira CSS, Kašanin-Grubin M, Solomun MK, Sushkova S, Minkina T, Zhao W, et al. Wetlands as nature-based solutions for water management in different environments. Curr Opin Environ Sci Health. 2023;33:100476.
- 3. Bridgewater P. A New Context for the Ramsar Convention: Wetlands in a Changing World. Rev Eur Commun Int Environ Law. 2008;17(1):100–6.
- 4.
Ghoraba SMM, Halmy MWA, Salem BB, Badr NBEJRS i MS. Application of IUCN Red List of Ecosystems to assess the ecological status of marine bar ecosystems of Burullus wetland: A Ramsar site. 2021;45:101844.
- 5. Nasser Mohamed Eid A, Olatubara CO, Ewemoje TA, Farouk H, El-Hennawy MT. Coastal wetland vegetation features and digital Change Detection Mapping based on remotely sensed imagery: El-Burullus Lake, Egypt. Int Soil Water Conserv Res. 2020;8(1):66–79.
- 6.
Castello M, Poldini L, Altobelli AJPS. The aquatic and wetland vegetation of Lake Doberdò: an analysis for conservation value assessment of a disappearing lake of the Classical Karst (North East Italy). 2021;58:75–106.
- 7. Shaltout KH, Al-Sodany YM. Vegetation analysis of Burullus Wetland: a RAMSAR site in Egypt. Wetlands Ecol Manage. 2008;16(5):421–39.
- 8. Keshta AE, Shaltout KH, Baldwin AH, Sharaf El-Din A, Eid EM. Variation in Plant Community Composition and Biomass to Macro and Micronutrients and Salinity across Egypt’s Five Major Coastal Lakes. Sustainability. 2022;14(10):6180.
- 9.
Jalhoum ME, Abdellatif MA, Mohamed ES, Kucher DE, Shokr MJH. Multivariate analysis and GIS approaches for modeling and mapping soil quality and land suitability in arid zones. 2024;10(5).
- 10. Dewidar KHM. Detection of land use/land cover changes for the northern part of the Nile delta (Burullus region), Egypt. Int J Remote Sens. 2004;25(20):4079–89.
- 11. Dewidar Kh, Khedr AA. Remote Sensing of Water Quality for Burullus Lake, Egypt. Geocart Int. 2005;20(3):43–9.
- 12.
El-Amier YA, Abd El-Azim H, El-Alfy MA. Spatial assessment of water and sediment quality in Burullus Lake using GIS technique. 2016;6(1):1–16.
- 13. El-Amier YA, Elnaggar AA, El-Alfy MA. Evaluation and mapping spatial distribution of bottom sediment heavy metal contamination in Burullus Lake, Egypt. Egypt J Basic Appl Sci. 2017;4(1):55–66.
- 14. Kganyago M, Adjorlolo C, Mhangara P, Tsoeleng L. Optical remote sensing of crop biophysical and biochemical parameters: An overview of advances in sensor technologies and machine learning algorithms for precision agriculture. Comput Electron Agric. 2024;218:108730.
- 15. Kingra PK, Majumder D, Singh SP. Application of remote sensing and gis in agriculture and natural resource management under changing climatic conditions. Agric Res J. 2016;53(3).
- 16. Weiss M, Jacob F, Duveiller GJR. Remote sensing for agricultural applications: A meta-review. R Soc. 2020;236:111402.
- 17.
Maged MA. Remote sensing of some environmental conditions at Lake Burullus. Alexandria: Department of Environmental Studies, Institute of Graduate Studies and Research, Alexandria University; 1994.
- 18.
Khdery GA, Abdel-Hameed UK, Aboelghar MA, Arafat SMJ. Diversity and taxonomic implication of angiosperms in Sinai peninsula as revealed by hyperspectral remote sensing. 2014;16:18.
- 19. Yones MS, Aboelghar MA, Khdery GA, Farag E, Ali AM, Salem NH, et al. Spectral measurements for monitoring of sugar beet infestation and its relation with production. Asian J Agri Biol. 2019;7(3):386–95.
- 20. Khdery GA, Farg E, Arafat SM. Natural vegetation cover analysis in Wadi Hagul, Egypt using hyperspectral remote sensing approach. Egypt J Remote Sens Space Sci. 2019;22(3):253–62.
- 21. A. Basheer M, B. El Kafrawy S, Mekawy A. Identification of mangrove plant using hyperspectral remote sensing data along the Red Sea, Egypt. Egypt J Aquat Biol Fish. 2019;23(1):27–36.
- 22. Khdery GA, Helal NM, Saleh HA. Using hyperspectral indices for investigation polluted plants and soil in Abu Zabaal area, Egypt. Acta Ecol Sin. 2023;43(5):776–84.
- 23.
Khdery GA. Remote sensing technology and its applications in pathology. In: Singh KP, Jahagirdar S, Sarma BK, editors. Emerging trends in plant pathology. Singapore: Springer; 2020.
- 24. Blackburn GA. Quantifying chlorophylls and caroteniods at leaf and canopy scales: An evaluation of some hyperspectral approaches. Remote Sens Environ. 1998;66(3):273–85.
- 25. Chappelle EW, Kim MS, McMurtrey JEJ III. Ratio analysis of reflectance spectra (RARS): an algorithm for the remote estimation of the concentrations of chlorophyll a, chlorophyll b, and carotenoids in soybean leaves. Rsoe. 1992;39(3):239–47.
- 26. Carter GA. Ratios of leaf reflectances in narrow wavebands as indicators of plant stress. Int J Remote Sens. 1994;15(3):697–703.
- 27. Gitelson AA, Merzlyak MN. Remote estimation of chlorophyll content in higher plant leaves. J Appl Remote Sens. 1997;18(12):2691–7.
- 28. Huete AR, Jackson RD. Soil and atmosphere influences on the spectra of partial canopies. Remote Sens Environ. 1988;25(1):89–105.
- 29. Qi J, Moran M, Cabot F, Dedieu GJ. Normalization of sun/view angle effects using spectral albedo-based vegetation indices. Remote Sens Environ. 1995;52(3):207–17.
- 30. Nagendra H, Lucas R, Honrado JP, Jongman RHG, Tarantino C, Adamo M, et al. Remote sensing for conservation monitoring: Assessing protected areas, habitat extent, habitat condition, species diversity, and threats. Ecol Indicat. 2013;33:45–59.
- 31. Ustin SL, Gamon JA. Remote sensing of plant functional types. New Phytol. 2010;186(4):795–816. pmid:20569415
- 32. Asner GP, Martin RE, Anderson CB, Knapp DEJR. Quantifying forest canopy traits: Imaging spectroscopy versus field survey. Remote Sens Environ. 2015;158:15–27.
- 33. Seeley M, Asner GP. Imaging Spectroscopy for Conservation Applications. Remote Sens. 2021;13(2):292.
- 34. Xue J, Su B. Significant Remote Sensing Vegetation Indices: A Review of Developments and Applications. J Sensor. 2017;2017:1–17.
- 35. Kothari S, Beauchamp-Rioux R, Blanchard F, Crofts AL, Girard A, Guilbeault-Mayers X, et al. Predicting leaf traits across functional groups using reflectance spectroscopy. New Phytol. 2023;238(2):549–66. pmid:36746189
- 36.
Maswada H. Ecological studies on Lake Borollus protected area, Egypt. Egypt: Faculty of Agriculture, Tanta University; 2004.
- 37. Keshta AE, Riter JCA, Shaltout KH, Baldwin AH, Kearney M, Sharaf El-Din A, et al. Loss of Coastal Wetlands in Lake Burullus, Egypt: A GIS and Remote-Sensing Study. Sustainability. 2022;14(9):4980.
- 38. El-Alfy MA, Serag MS, Basiony AI, Fathi M, Darwish DH. Influence of land cover indices and surface temperature on the metals bioaccumulation by three macrophytes in Lake Burullus, Egypt. J Coast Conserv. 2023;27(1):5.
- 39.
Westhoff V, Van Der Maarel E. The braun-blanquet approach. Springer; 1978.
- 40.
Ellenberg D, Mueller-Dombois D. Aims and methods of vegetation ecology. New York: Wiley; 1974.
- 41.
Boulos L. Flora of Egypt. Checklist. 1995.
- 42.
Boulos L. Flora of Egypt (Alismataceae-Orchidaceae). Cairo, Egypt: Al Hadara Publishing; 2005.
- 43.
Tackholm V, Boulos L. Students’ flora of Egypt. 1974.
- 44.
Raunkiaer C. The life forms of plants and statistical plant geography; being the collected papers of C. Raunkiær. 1934.
- 45.
Kent M. Vegetation description and data analysis: a practical approach. John Wiley & Sons; 2011.
- 46. Cochrane MA. Using vegetation reflectance variability for species level classification of hyperspectral data. Int J Remote Sens. 2000;21(10):2075–87.
- 47. Clark ML, Roberts DA, Clark DBJR. Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales. Remote Sens Environ. 2005;96(3–4):375–98.
- 48. Huntjr E, Rock B. Detection of changes in leaf water content using Near- and Middle-Infrared reflectances⋆. Remote Sens Environ. 1989;30(1):43–54.
- 49. Merzlyak MN, Gitelson AA, Chivkunova OB, Rakitin VY. Non‐destructive optical detection of pigment changes during leaf senescence and fruit ripening. JPp. 1999;106(1):135–41.
- 50. Xu L, Myneni RB, Chapin FS III, Callaghan TV, Pinzon JE, Tucker CJ, et al. Temperature and vegetation seasonality diminishment over northern lands. Nature Clim Change. 2013;3(6):581–6.
- 51.
Rouse Jr JW, Haas RH, Deering D, Schell J, Harlan JC. Monitoring the vernal advancement and retrogradation (green wave effect) of natural vegetation. 1974.
- 52. Mascarini L, Lorenzo GA, Vilella FJJ. Leaf area index, water index, and red: far red ratio calculated by spectral reflectance and its relation to plant architecture and cut rose production. Jot Asf Hs. 2006;131(3):313–9.
- 53. Peng D, Zhang X, Zhang B, Liu L, Liu X, Huete AR, et al. Scaling effects on spring phenology detections from MODIS data at multiple spatial resolutions over the contiguous United States. ISPRS J Photogrammet Remote Sens. 2017;132:185–98.
- 54. Gitelson AA, Merzlyak MN, Chivkunova OB. Optical properties and nondestructive estimation of anthocyanin content in plant leaves. Photochem Photobiol. 2001;74(1):38–45. pmid:11460535
- 55. Gitelson AA, Stark R, Grits U, Rundquist D, Kaufman Y, Derry D. Vegetation and soil lines in visible spectral space: A concept and technique for remote estimation of vegetation fraction. Int J Remote Sens. 2002;23(13):2537–62.
- 56. Gitelson AA, Viña A, Ciganda V, Rundquist DC, Arkebauer TJ. Remote estimation of canopy chlorophyll content in crops. Geophys Res Lett. 2005;32(8).
- 57. Gitelson A, Merzlyak MN. Quantitative estimation of chlorophyll-a using reflectance spectra: Experiments with autumn chestnut and maple leaves. J Photochem Photobiol B Biol. 1994;22(3):247–52.
- 58. Sims DA, Gamon JAJ. Relationships between leaf pigment content and spectral reflectance across a wide range of species, leaf structures and developmental stages. R Soc E. 2002;81(2–3):337–54.
- 59. Olofsson P, Foody GM, Stehman SV, Woodcock CE. Making better use of accuracy data in land change studies: estimating accuracy and area and quantifying uncertainty using stratified estimation. Remote Sens Environ. 2013;129:122–31.
- 60. Curran PJ, Windham WR, Gholz HL. Exploring the relationship between reflectance red edge and chlorophyll concentration in slash pine leaves. Tree Physiol. 1995;15(3):203–6. pmid:14965977
- 61.
Guyot G, Baret F. Utilisation de la haute resolution spectrale pour suivre l’etat des couverts vegetaux. Spectral Signatures of Objects in Remote Sensing. 1988.
- 62. Gao B. NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens Environ. 1996;58(3):257–66.
- 63. Wu J, Liu Y, Wang J, He T. Application of Hyperion data to land degradation mapping in the Hengshan region of China. Int J Remote Sens. 2010;31(19):5145–61.
- 64. Kumar S, Gautam G, Saha SK. Hyperspectral remote sensing data derived spectral indices in characterizing salt-affected soils: a case study of Indo-Gangetic plains of India. Environ Earth Sci. 2014;73(7):3299–308.
- 65. Thenkabail PS, Smith RB, De Pauw E. Hyperspectral Vegetation Indices and Their Relationships with Agricultural Crop Characteristics. Remote Sens Environ. 2000;71(2):158–82.
- 66. Navarro-Torre S, Garcia-Caparrós P, Nogales A, Abreu MM, Santos E, Cortinhas AL, et al. Sustainable agricultural management of saline soils in arid and semi-arid Mediterranean regions through halophytes, microbial and soil-based technologies. Environ Experim Bot. 2023;212:105397.
- 67.
Van Oudtshoorn KVR, Van Rooyen MW. Dispersal biology of desert plants. Springer Science & Business Media; 2013.
- 68.
Stanley KE. Evolutionary trends in the grasses (Poaceae): a review: Ann Arbor (MI): Michigan Publishing, University of Michigan Library; 1999.
- 69.
Behn K, Alvarez M, Mutebi S, Becker MJP. Vegetation diversity in East African wetlands. 2022;2022(3):199–219.
- 70.
Al-Sodany Y. Vegetation analysis of the northern part of Nile Delta region. Tanta: Fac Sci, Tanta University; 1992.
- 71. Shaltout K, Hassan L, Farahat E. Vegetation-environment relationships in south Nile Delta. Taeckholmia. 2005;25(1):15–46.
- 72. El-Amier YA, El Hayyany LYJA. Floristic composition and species diversity of plant communities associated with genus Atriplex in Nile Delta coast, Egypt. J Coast Biol. 2020;9(1):11–24.
- 73. Danin A, Orshan G. The distribution of Raunkiaer life forms in Israel in relation to the environment. J Vegetat Sci. 1990;1(1):41–8.
- 74.
Brown G, Mies B. Vegetation ecology of Socotra. Springer Science & Business Media; 2012.
- 75.
El-Hadidi M. Natural vegetation. 1993;39–62.
- 76. Hestir EL, Khanna S, Andrew ME, Santos MJ, Viers JH, Greenberg JA, et al. Identification of invasive vegetation using hyperspectral remote sensing in the California Delta ecosystem. Remote Sens Environ. 2008;112(11):4034–47.
- 77. Tucker CJ, Garratt MWJA. Leaf optical system modeled as a stochastic process. Appl Opt. 1977;16(3):635–42.
- 78. Gitelson AA, Zur Y, Chivkunova OB, Merzlyak MN. Assessing carotenoid content in plant leaves with reflectance spectroscopy. Photochem Photobiol. 2002;75(3):272–81. pmid:11950093
- 79. Bartold M, Wróblewski K, Kluczek M, Dąbrowska-Zielińska K, Goliński P. Examining the Sensitivity of Satellite-Derived Vegetation Indices to Plant Drought Stress in Grasslands in Poland. Plants (Basel). 2024;13(16):2319. pmid:39204755
- 80. Gitelson AA, Buschmann C, Lichtenthaler HKJR. The chlorophyll fluorescence ratio F735/F700 as an accurate measure of the chlorophyll content in plants. Rsoe. 1999;69(3):296–302.
- 81. Zarco-Tejada PJ, González-Dugo V, Berni JAJ. Fluorescence, temperature and narrow-band indices acquired from a UAV platform for water stress detection using a micro-hyperspectral imager and a thermal camera. Remote Sens Environ. 2012;117:322–37.
- 82. Nițu A, Florea C, Ivanovici M, Racoviteanu A. NDVI and Beyond: Vegetation Indices as Features for Crop Recognition and Segmentation in Hyperspectral Data. Sensors (Basel). 2025;25(12):3817. pmid:40573704
- 83. Filella I, Penuelas JJ. The red edge position and shape as indicators of plant chlorophyll content, biomass and hydric status. J Appl Ecol. 1994;15(7):1459–70.
- 84. Gould KS. Nature’s Swiss Army Knife: The Diverse Protective Roles of Anthocyanins in Leaves. J Biomed Biotechnol. 2004;2004(5):314–20. pmid:15577195