Search

The use of remote sensing for forest health monitoring has increased in popularity over the years, with improved quality in spatial and spectral resolutions. However, satellite revisit times are ge...

Frost is a sporadic meteorological event affecting the productivity of commercial forests in South Africa. Severe frost occurrences may cause irreversible damage to forest stands, slowing down tree...

Accurate spatial information on the location of invasive alien plants (IAPs) in riparian environments is critical to fulfilling a comprehensive weed management regime. This study aimed to automatic...

Developing models for estimating aboveground biomass (AGB) in naturally growing forests is critical for climate change modelling. AGB models developed using satellite imagery varies with study area, depending on the complexity of vegetation and landscape structure, which affects the upwelling radiance. We assessed the potential of SPOT-6 imagery in predicting AGB of trees planted at different time periods, using image texture combinations. Image texture variables were computed from the SPOT6 pan-sharpened image data, which is characterised by a 1.5 m spatial resolution. In addition, we incorporated the minimal variance technique to select the optimum window sizes that best captures AGB variation in our study area. The results showed that image texture was able to detect AGB for both mature and young trees, however, models detecting mature trees were more superior, with accuracies of R2 = 0.70 and 0.25 for 2009–2011 and 2011–2013 plantation phases, respectively. In addition, our results showed that the three band texture ratios yielded the highest accuracy (R2 = 0.88 and RMSE = 54.54 kg m−2) compared to two texture (R2 = 0.85 and RMSE = 60.65 kg m−2) and single texture band combinations (R2 = 0.64 and RMSE = 94.13 kg m−2). A frequency analysis was also run to determine which bands appeared more frequently in the selected texture band models. The frequency analysis revealed that both the red and green bands appeared more frequently on the selected texture band variables, indicating that they were more sensitive to the variation of AGB in our study area. The results showed high variation in AGB within the Buffelsdraai reforestation site, especially due to varying tree plantation phases as well as topography. In essence, the study demonstrated the possibility of image texture combinations computed from the SPOT-6 image in estimating AGB.

Commercial forests are expanding globally, with great potential to absorb carbon and mitigate climate change. However, whereas the role of natural forests in carbon sequestration has been widely investigated, there is a paucity in the literature on the role of commercial forests in carbon assimilation. Hence, understanding the role of commercial forests in carbon storage is essential for quantifying local, regional or global carbon balances, which is valuable for climate change mitigation. Soil carbon is known to be the largest pool within any forest landscape, and is controlled by a wide range of physical and climatic factors. However, the relationship between soil organic carbon (SOC) and topo-climatic variables controlling its distribution within commercial forests is still poorly understood. Due to the limitations encountered in traditional systems of SOC determination, particularly at large spatial extents, geospatial techniques have recently emerged as a viable alternative for mapping soil properties. Therefore, this study sought to map SOC stocks variability within the commercial forest landscape, using landscape topo-climatic variables. A total of 81 soil samples was analysed for SOC concentrations and 31 topographic and climatic variables were used as predictors to SOC variability. To reduce multicollinearity, these variables were reduced to 11 using stepwise backward elimination and the maximum entropy (Maxent) algorithm was used for regression analysis to determine the relationship between SOC and the selected topo-climatic variables. Good accuracies were obtained for both training (area under the curve = 0.906) and test (area under the curve = 0.885) data sets, and demonstrate the effectiveness of selected topo-climatic variables and the Maxent algorithm in predicting SOC stocks. This study provides a framework for monitoring the status of soil carbon in commercial forest compartments and provides a viable approach for local, national or regional carbon accounting – valuable for climate change mitigation. Significance: Rainfall and temperature, as well as topographic variables (such as slope, elevation and topographic wetness index) are effective in mapping SOC distribution. The model developed is useful in predicting SOC occurrence and yielded an effective framework for continuous monitoring and assessment of SOC. The method developed in this study is cost-effective and suggests the use of other readily available climatic and topographic information for the prediction of SOC under commercial forestry in South Africa and indeed globally. Results from this study are important to achieve the national carbon accounting objective and are also valuable to forest managers, ecologists and relevant stakeholders in understanding the spatial distribution of SOC.

The sustainability of the South African forestry industry is under threat due to the increasing number of biotic stressors in a changing environment. Bark-stripping baboons are impacting th...

Remote sensing has revolutionized forest management and has been widely employed to model canopy gaps. In this study, a canopy height model (CHM) and an intensity raster (IR) derived from light det...

Industrial zones in eThekwini, South Africa, such as the South Durban Basin (SDB) are often characterised by extremely poor air quality owing to industrial emissions. This study investigated the effects of SO 2 pollution on selected stress biomarkers in the leaves of Trichilia dregeana to assess the utility of this species as a bioindicator of this pollutant. Leaves were sampled from trees growing at three industrial (treatment) sites (Prospecton, Ganges and Southern Works) within the SDB and from greenhouse-located trees ( ex situ control). Sampling accommodated for directional and seasonal wind effects and yielded a sample size of n = 24 for all four seasons. Ground-level SO 2 concentrations ([SO 2 ]) measured at each site were positively correlated with leaf sulphate contents and both [SO 2 ] and leaf sulphate levels of the control were lower than the treatment sites. Values for the various biomarkers did not differ significantly for leaves from different cardinal points within sites, but seasonal variations were evident. Except for leaf chlorophyll fluorescence, all biomarkers could discriminate between treatment and control leaves. Though seasonal data for a number of these biomarkers were significantly correlated with leaf sulphate content, none of the biomarkers were sensitive enough to reflect differences in leaf sulphate levels across the treatment sites. Furthermore, the results suggest that leaf area and relative chlorophyll content should be measured in combination with each other. Leaf sulphate content is a reliable proxy for SO 2 pollution. All biomarkers, except for chlorophyll fluorescence, can be used to establish T. dregeana leaves as a bioindicator of SO 2 pollution.

Swiss chard nutrients were estimated under trial conditions using hyperspectral data.Partial least squares (PLS1 and PLSR2) and sparse PLS1 and PLS2 regressions performance was compared.SPLS-based regression methods significantly reduced the dimensionality in hyperspectral data.Macronutrients estimation models were more accurate than micronutrients estimation ones. Timely information on crop foliar nutrient content provides a measure of crop nutritional and vitality status. Growers and farm managers use such information for precision crop management such as an appropriate fertilizer application to correct for any crop nutrient deficiencies at identified hotspots. Foliar heavy nutrient content could also be a direct indicator of crops having been polluted from the surroundings, which may be a result of heavy metals absorbed from, among others, contaminated soils and waste water. In the present study, we explored the potential use of four partial least squares (PLS)-based regression algorithms for estimating foliar Swiss chard macro- and micronutrient concentrations using ground-based hyperspectral data under three treatments; i.e. rainwater+fertilizer (R+F), tap water+fertilizer (T+F), and treated wastewater (W). Swiss chard canopy-level hyperspectral measurements under these three treatments were collected using a handheld spectroradiometer 2.5months after sowing. The reflectance spectra were normalized to their first-order derivatives. The concentrations of three Swiss chard foliar macronutrients (NPK) and three micronutrients (Zn, Cu and Fe) under the three treatments were determined. Regression models for estimating macro- and micronutrient concentrations were then derived using PLS1 and sparse PLS1 methods, while the potential simultaneous estimation of the macronutrient as well as micronutrient concentrations was explored using the PLS2 and SPLS2 regression approaches. Results showed that high variances in the macro- and micronutrient concentrations can be explained by the four regression models under the three treatments (R2train ranged between 0.73 and 0.99), except when P, Zn and Cu concentrations were estimated using the PLS2-based models under the three treatments (R2train ranged between 0.08 and 0.68) and Fe concentration using SPLSR1 under W treatment (R2train=0.64). Our results further showed that Swiss chard foliar N (RMSE=1.67%) concentration under R+F treatment and Fe (RMSE=7.83%) concentration under the T+F treatment most accurately estimated macro- and micronutrients. Our study also showed that the Swiss chard foliar macronutrient concentrations were more accurately estimated compared to micronutrient concentrations and PLS2 outperformed the PLS1 based regression model. The results of the current study pave the way for developing an effective foliar nutrient estimation routine suitable for monitoring Swiss chard nutrient status under different treatments.

Drought limits the production of plantation forests, notably in the drought-prone Zululand region of South Africa. During the last 40 years, the country has faced a series of severe droughts, however that of 2015 stands out as the most extreme and prolonged. The 2015 drought impaired forest productivity and led to widespread tree mortality in this region, but the identification of tree response to drought stress remains uncertain because of its spatial variability. To address this problem, a method that can capture drought patterns and identify trees with similar reactions to drought stress is desired. This could improve the accuracy of detecting trees suffering from drought stress which is key for forest management planning. In this study, we aimed to evaluate the utility of unsupervised mapping approaches in compartments of Eucalyptus trees with similar drought characteristics based on the Normalized Difference Water Index (NDWI) and to demonstrate the value of cloud-based Google Earth Engine (GEE) resources for rapid landscape drought monitoring. Our results showed that calculating distances between pixels using three different matrices (Random Forest (RF) proximity, Euclidean and Manhattan) can accurately detect similarities within a dataset. The RF proximity matrix produced the best measures, which were clustered using Wards hierarchical clustering to detect drought with the highest overall accuracy of 87.7%, followed by Manhattan (85.9%) and Euclidean similarity measures (79.9%), with user and producer results between 84.2% to 91.2%, 42.8% to 98.2% and 37.2% to 94.7%, respectively. These results confirm the value of the RF proximity matrix and underscore the capability of automatic unsupervised mapping approaches for monitoring drought stress in tree plantations, as well as the value of using GEE for providing cost effective datasets to resource stricken countries.

Coryphodema tristis is a wood-boring insect, indigenous to South Africa, that has recently been identified as an emerging pest feeding on Eucalyptus nitens, resulting in extensive damage and economic loss. Eucalyptus plantations contributes over 9% to the total exported manufactured goods of South Africa which contributes significantly to the gross domestic product. Currently, the distribution extent of the Coryphodema tristis is unknown and estimated to infest Eucalyptus nitens compartments from less than 1% to nearly 80%, which is certainly a concern for the forestry sector related to the quantity and quality of yield produced. Therefore, the study sought to model the probability of occurrence of Coryphodema tristis on Eucalyptus nitens plantations in Mpumalanga, South Africa, using data from the Sentinel-2 multispectral instrument (MSI). Traditional field surveys were carried out through mass trapping in all compartments (n = 878) of Eucalyptus nitens plantations. Only 371 Eucalyptus nitens compartments were positively identified as infested and were used to generate the Coryphodema tristis presence data. Presence data and spectral features from the area were analysed using the Maxent algorithm. Model performance was evaluated using the receiver operating characteristics (ROC) curve showing the area under the curve (AUC) and True Skill Statistic (TSS) while the performance of predictors was analysed with the jack-knife. Validation of results were conducted using the test data. Using only the occurrence data and Sentinel-2 bands and derived vegetation indices, the Maxent model provided successful results, exhibiting an area under the curve (AUC) of 0.890. The Photosynthetic vigour ratio, Band 5 (Red edge 1), Band 4 (Red), Green NDVI hyper, Band 3 (Green) and Band 12 (SWIR 2) were identified as the most influential predictor variables. Results of this study suggest that remotely sensed derived vegetation indices from cost-effective platforms could play a crucial role in supporting forest pest management strategies and infestation control.

The invasive weed Solanum mauritianum (bugweed) has infested large areas of plantation forests in KwaZulu-Natal, South Africa. Bugweed often forms dense infestations and rapidly capitalises on available natural resources hindering the production of forest resources. Precise assessment of bugweed canopy cover, especially at low abundance cover, is essential to an effective weed management strategy. In this study, the utility of AISA Eagle airborne hyperspectral data (393–994 nm) with the new generation Worldview-2 multispectral sensor (427–908 nm) was compared to detect the abundance of bugweed cover within the Hodgsons Sappi forest plantation. Using sparse partial least squares discriminant analysis (SPLS-DA), the best detection results were obtained when performing discrimination using the remotely sensing images combined with LiDAR. Overall classification accuracies subsequently improved by 10% and 11.67% for AISA and Worldview-2 respectively, with improved detection accuracies for bugweed cover densities as low as 5%. The incorporation of LiDAR worked well within the SPLS-DA framework for detecting the abundance of bugweed cover using remotely sensed data. In addition, the algorithm performed simultaneous dimension reduction and variable selection successfully whereby wavelengths in the visible (393–670 nm) and red-edge regions (725–734 nm) of the spectrum were the most effective.

The use of air quality monitoring stations is expensive, with pollution data being either unavailable or inaccessible. Hence, effects of atmospheric sulphur dioxide (SO2) levels on biomarkers related to environmental stress were investigated for Trichilia dregeana tree leaves, in order to assess their bioindicator potential. Leaves were sampled randomly from trees at three industrial sites within the South Durban Basin, and an ex situ control, across two seasons (n = 28, per season). Ground-level SO2 concentrations were measured daily and ranged between 1 and 25 ppb. There were significant (p

The accurate detection and mapping of plant invasions is important for an effective weed management strategy in forest plantations. In this study, the utility of WorldView-2 was investigated to automatically map the occurrence of Solanum mauritianum (bugweed) found as an anomaly in forest margins, open areas and riparian zones. The unsupervised methodology developed, proved to be an effective and an accurate framework in detecting and mapping the invasive alien plant (IAP). Using the random forest (RF) proximity matrix, similarity measures between pixels were successfully transformed into scores (Eigen weights) for each pixel using eigenvector analysis. Neighbourhood windows with minimum variance revealed the most important information from localized surrounding pixels to detect potential anomalous pixels. Bugweed occurrence in forest margins, open areas and riparian zones were successfully mapped at accuracies of 91.33%, 85.08%, and 67.90%, respectively. This research has demonstrated the unique capability of using an automated unsupervised RF approach for mapping IAPs using new generation multispectral remotely sensed data.

Hyperspectral data collected using a field spectroradiometer was used to model asymptomatic stress in Pinus radiata and Pinus patula seedlings infected with the pathogen Fusarium circinatum. Spectral data were analyzed using the random forest algorithm. To improve the classification accuracy of the model, subsets of wavebands were selected using three feature selection algorithms: (1) Boruta; (2) recursive feature elimination (RFE); and (3) area under the receiver operating characteristic curve of the random forest (AUC-RF). Results highlighted the robustness of the above feature selection methods when used in conjunction with the random forest algorithm for analyzing hyperspectral data. Overall, the Boruta feature selection algorithm provided the best results. When discriminating F. circinatum stress in Pinus radiata seedlings, Boruta selected wavebands (n = 69) yielded the best overall classification accuracies (training error of 17.00%, independent test error of 17.00% and an AUC value of 0.91). Classification results were, however, significantly lower for P. patula seedlings, with a training error of 24.00%, independent test error of 38.00%, and an AUC value of 0.65. A hybrid selection method that utilizes combinations of wavebands selected from the three feature selection algorithms was also tested. The hybrid method showed an improvement in classification accuracies for P. patula, and no improvement for P. radiata. The results of this study provide impetus towards implementing a hyperspectral framework for detecting stress within nursery environments.

South Africa has a long history of recurrent droughts that have adversely affected its economic performance. The recent 2015 drought has been declared the most serious in 26 years and impaired key agricultural sectors including the forestry sector. Research on the forests&rsquo, responses to drought is therefore essential for management planning and monitoring. The effects of the latest drought on the forests in South Africa have not been studied and are uncertain. The study reported here addresses this gap by using Moderate Resolution Imaging Spectroradiometer (MODIS)-derived normalized difference vegetation index (NDVI) and precipitation data retrieved and processed using the JavaScript code editor in the Google Earth Engine (GEE) and the corresponding normalized difference infrared index (NDII), Palmer drought severity index (PDSI), and El Ni&ntilde, o time series data for KwaMbonambi, northern Zululand, between 2002 and 2016. The NDVI and NDII time series were decomposed using the Breaks for Additive Seasonal and Trend (BFAST) method to establish the trend and seasonal variation. Multiple linear regression and Mann&ndash, Kendall tests were applied to determine the association of the NDVI and NDII with the climate variables. Plantation trees displayed high NDVI values (0.74&ndash, 0.78) from 2002 to 2013, then, they decreased sharply to 0.64 in 2015. The Mann&ndash, Kendall trend test confirmed a negative significant (p = 0.000353) trend between 2014 and 2015. This pattern was associated with a precipitation deficit and low NDII values during a strong El Ni&ntilde, o phase. The PDSI (&minus, 2.6) values indicated severe drought conditions. The greening decreased in 2015, with some forest remnants showing resistance, implying that the tree species had varying sensitivity to drought. We found that the plantation trees suffered drought stress during 2015, although it seems that the trees began to recover, as the NDVI signals rose in 2016. Overall, these results demonstrated the effective use of the NDVI- and NDII-derived MODIS data coupled with climatic variables to provide insights into the influence of drought on plantation trees in the study area.

Hyperspectral datasets contain spectral noise, the presence of which adversely affects the classifier performance to generalize accurately. Despite machine learning algorithms being regarded as robust classifiers that generalize well under unfavourable noisy conditions, the extent of this is poorly understood. This study aimed to evaluate the influence of simulated spectral noise (10%, 20%, and 30%) on random forest (RF) and oblique random forest (oRF) classification performance using two node-splitting models (ridge regression (RR) and support vector machines (SVM)) to discriminate healthy and low infested water hyacinth plants. Results from this study showed that RF was slightly influenced by simulated noise with classification accuracies decreasing for week one and week two with the addition of 30% noise. In comparison to RF, oRF-RR and oRF-SVM yielded higher test accuracies (oRF-RR: 5.36%–7.15%; oRF-SVM: 3.58%–5.36%) and test kappa coefficients (oRF-RR: 10.72%–14.29%; oRF-SVM: 7.15%–10.72%). Notably, oRF-RR test accuracies and kappa coefficients remained consistent irrespective of simulated noise level for week one and week two while similar results were achieved for week three using oRF-SVM. Overall, this study has demonstrated that oRF-RR can be regarded a robust classification algorithm that is not influenced by noisy spectral conditions.

Accurate, reliable, and up-to-date forest stand volume information is a prerequisite for a detailed evaluation of commercial forest resources and their sustainable management. Commercial forest responses to global climate change remain uncertain, and hence the mapping of stand volume as carbon sinks is fundamentally important in understanding the role of forests in stabilizing climate change effects. The aim of this study was to examine the utility of stochastic gradient boosting SGB and multi-source data to predict stand volume of a Eucalyptus plantation in South Africa. The SGB ensemble, random forest RF, and stepwise multiple-linear regression SMLR were used to predict Eucalyptus stand volume and other related tree-structural attributes such as mean tree height and mean diameter at breast height DBH. Multi-source data consisted of SPOT-5 raw spectral features four bands, 14 spectral vegetation indices, rainfall data, and stand age. When all variables were used, the SGB algorithm showed that stand volume can be accurately estimated R2 = 0.78 and RMSE = 33.16 m3 ha−1 23.01% of the mean. The competing RF ensemble produced an R2 value of 0.76 and a RMSE value of 37.28 m3 ha−1 38.28% of the mean. SMLR on the other hand, produced an R2 value of 0.65 and an RMSE value of 42.50 m3 ha−1 42.50% of the mean. Our study further showed that Eucalyptus mean tree height R2 = 0.83 and RMSE = 1.63 m 9.08% of the mean and mean diameter at breast height R2 = 0.74 and RMSE = 1.06 7.89% of the mean can also be reasonably predicted using SGB and multi-source data. Furthermore, when the most important SGB model-selected variables were used for prediction, the predictive accuracies improved significantly for mean DBH R2 = 0.81 and RMSE = 1.21 cm 6.12% of the mean, mean tree height R2 = 0.86 and RMSE = 1.39 m 7.02% of the mean, and stand volume R2 = 0.83 and RMSE = 29.58 m3 ha−1 17.63% of the mean. These results underscore the importance of integrating multi-source data with remotely sensed data for predicting Eucalyptus stand volume and related tree-structural attributes.

Detecting and mapping plant invaders using hyperspectral remote sensing is necessary in mitigating the extensive ecologic and economic damage these alien plants induce on our forest ecosystems. Using AISA Eagle image data, this study investigated the capability of two unsupervised classification methods for the detection and mapping of Solanum mauritianum located within commercial forestry ecosystems. The existing random forest (RF) outlier detection method when used in conjunction with Anselins Moran’s I produced a detection rate (DR) of 89% with a false positive rate (FPR) of 9.26%. In comparison, the newly developed methodology which is based on the decomposition of the RF proximity matrix using principal component analysis (PCA) resulted in a DR of 95% with a lower FPR (6.39%). Overall, this research has demonstrated the potential of utilizing an unsupervised and accurate RF framework for the detection and mapping of alien invasive plants.

In commercial forestry, regular terrestrial enumerations of the growing stock are required for the valuation, sustainable management and planning of current and future timber supplies. In this study we examined whether the combination of synthetic aperture radar (ALOS PALSAR) and optical satellite (SPOT 4) image data can accurately predict the timber volume of even-aged Eucalyptus plantations located in South Africa. Results from this study show that the combination of ALOS PALSAR and SPOT 4 produces a R 2 value of 0.68 for the planted model, whereas the coppiced model produced a R 2 value of 0.55. However, by including stand age as an independent variable in the stepwise model, there was a 15% improvement for the planted model, whereas the coppiced model produced a 27% improvement. The final model developed in this study produced a R 2 value of 0.83 and a RMSE of 31.71 m3 ha −1 for planted stands, whereas the model for coppiced stands produced a R 2 value of 0.82 and a RMSE of 27.70 m3 ha −1 . As it is not practical or financially feasible for commercial forestry companies to carry out terrestrial enumerations for all plantations on an annual basis, the model developed in this study presents an alternative and accurate method to calculate timber volume for even-aged Eucalyptus plantations. Keywords: commercial forestry, Eucalyptus species, optical remote sensing, saturation, synthetic aperture radar, timber volume

To ensure sustainable planning and management within a commercial forest plantation, forest inventory data that is up to date has become increasingly essential and necessary. Data for these measurements may be collected in the form of traditional field based approaches or using remote sensing techniques. The aim of this study was to examine the utility of the partial least squares regression (PLSR), random forest (RF) and a PLSR-RF hybrid machine learning approach for the prediction of four forest structural attributes: (basal area, volume, dominant tree height and mean tree height) within a commercial Eucalyptus forest plantation using a combination of spectral and textural information of high spatial resolution (0.15m) remote sensing data. The best model for this study was produced for mature E. dunii species for dominant tree height using the PLSR-RF hybrid model (R 2 = 0.82 and RMSE = 207m). The results of this study highlight the robustness and potential of the PLSR-RF hybrid model for the prediction of forest structural attributes using high resolution imagery within a commercial Eucalyptus forest plantation.

The quantification of aboveground biomass using remote sensing is critical for better understanding the role of forests in carbon sequestration and for informed sustainable management. Although remote sensing techniques have been proven useful in assessing forest biomass in general, more is required to investigate their capabilities in predicting intra-and-inter species biomass which are mainly characterised by non-linear relationships. In this study, we tested two machine learning algorithms, Stochastic Gradient Boosting (SGB) and Random Forest (RF) regression trees to predict intra-and-inter species biomass using high resolution RapidEye reflectance bands as well as the derived vegetation indices in a commercial plantation. The results showed that the SGB algorithm yielded the best performance for intra-and-inter species biomass prediction; using all the predictor variables as well as based on the most important selected variables. For example using the most important variables the algorithm produced an R2 of 0.80 and RMSE of 16.93 t·ha−1 for E. grandis; R2 of 0.79, RMSE of 17.27 t·ha−1 for P. taeda and R2 of 0.61, RMSE of 43.39 t·ha−1 for the combined species data sets. Comparatively, RF yielded plausible results only for E. dunii (R2 of 0.79; RMSE of 7.18 t·ha−1). We demonstrated that although the two statistical methods were able to predict biomass accurately, RF produced weaker results as compared to SGB when applied to combined species dataset. The result underscores the relevance of stochastic models in predicting biomass drawn from different species and genera using the new generation high resolution RapidEye sensor with strategically positioned bands.

There is an increasing demand for fresh vegetables such as Swiss chard in cognisance of their nutritive value. Early prediction of Swiss chard yield provides a valuable knowledge base for product management decisions like pre-harvest planning, post-harvest handing, food policy, and marketing. Consequently, the objective of the present study was to investigate the use of hyperspectral data in predicting yield of Swiss chard grown under different irrigation water sources. Swiss chard ground-based hyperspectral data were collected at canopy level using a handheld spectroradiometer at 2 and 2.5months after planting. Reflectance spectra were transformed to their first-order derivative and partial least squares (PLS) and sparse PLS (SPLS) regressions (R) were used for data analysis. Results showed that 95% and 97% of Swiss chard fresh and dry yields variation, respectively could be explained. SPLSR outperformed PLSR models for predicting Swiss chard fresh and dry yields. Results further showed that models developed using data collected when the crop was 2.5months old were more accurate than models derived using a 2-month old crop data, except the SPLSR model for predicting dry yield. Fresh yield estimates could be accurately modelled (root mean square error: RMSE=23.97% of the mean, Nash-Sutcliffe efficiency: NSE=0.93). However, Swiss chard dry yield could not be reliably predicted (at minimum RMSE=35.00% of the mean and a maximum NSE of 0.60 were obtained). This study demonstrates the potential of hyperspectral data in predicting Swiss chard fresh yield using combined irrigation treatments data sets. The study offers insight to the potential of large-scale prediction and estimation of Swiss chard yield using space borne and/or airborne hyperspectral data.

The visual progression of sirex ( Sirex noctilio ) infestation symptoms has been categorized into three distinct infestation phases, namely the green, red and grey stages. The grey stage is the final stage which leads to almost complete defoliation resulting in dead standing trees or snags. Dead standing pine trees however, could also be due to the lightning damage. Hence, the objective of the present study was to distinguish amongst healthy, sirex grey-attacked and lightning-damaged pine trees using AISA Eagle hyperspectral data, random forest (RF) and support vector machines (SVM) classifiers. Our study also presents an opportunity to look at the possibility of separating amongst the previously mentioned pine trees damage classes and other landscape classes on the study area. The results of the present study revealed the robustness of the two machine learning classifiers with an overall accuracy of 74.50% (total disagreement = 26%) for RF and 73.50% (total disagreement = 27%) for SVM using all the remaining AISA Eagle spectral bands after removing the noisy ones. When the most useful spectral bands as measured by RF were exploited, the overall accuracy was considerably improved; 78% (total disagreement = 22%) for RF and 76.50% (total disagreement = 24%) for SVM. There was no significant difference between the performances of the two classifiers as demonstrated by the results of McNemar’s test (chi-squared; χ 2 = 0.14, and 0.03 when all the remaining ASIA Eagle wavebands, after removing the noisy ones and the most important wavebands were used, respectively). This study concludes that AISA Eagle data classified using RF and SVM algorithms provide relatively accurate information that is important to the forest industry for making informed decision regarding pine plantations health protocols.

Accurate estimates of papyrus Cyperus papyrus biomass are critical for an efficient papyrus swamp monitoring and management system. The objective of this study was to test the utility of random forest RF regression and two narrow-band vegetation indices in estimating above-ground biomass AGB for complex and densely vegetated swamp canopies. The normalized difference vegetation index NDVI and enhanced vegetation index EVI were calculated from field spectrometry data and fresh AGB was measured in 82 quadrats at three different areas in the iSimangaliso Wetland Park, South Africa. NDVI was calculated from all possible band combinations of the electromagnetic spectrum 350 and 2500 nm, while EVI was calculated from possible band combinations in the blue, red, and near infrared of the spectrum. Backward feature elimination and RF regression were used as variable selection and modelling techniques to predict papyrus AGB. Results showed that the effective portions of electromagnetic spectrum for estimation AGB of papyrus swamp were located within the blue, red, red-edge, and near-infrared regions. The three best selected EVIs were computed from bands located at i 445, 682, and 829 nm, ii 497, 676, and 1091 nm, and iii 495, 678, and 1120 nm. These indices produced better predictive accuracies R2 = 0.90; root mean square error of prediction RMSEP = 0.289 kg m−2; 7.99% of the mean than the best selected NDVIs R2 = 0.85; RMSEP = 0.343 kgm−2; 9.49% of the mean that were calculated from bands located at i 739 and 829 nm, ii 739 and 814 nm, iii 744 and 789 nm, and iv 734 and 909 nm. The results of the present study demonstrate the utility of narrow-band vegetation indices EVI and NDVI and RF regression in estimating papyrus AGB at high density, a previously challenging task with broadband satellite sensors.

The invasive plant known as bugweed (Solanum mauritianum) is a notorious invader of forestry plantations in the eastern parts of South Africa. Not only is bugweed considered to be one of five most widespread invasive alien plant (IAP) species in the summer rainfall regions of South Africa but it is also one of the worst invasive alien plants in Africa. It forms dense infestations that not only impacts upon commercial forestry activities but also causes significant ecological and environment damage within natural areas. Effective weed management efforts therefore require robust approaches to accurately detect; map and monitor weed distribution in order to mitigate the impact on forestry operations. The main objective of this research was to determine the utility of support vector machines (SVMs) with a 272-waveband AISA Eagle image to detect and map the presence of co-occurring bugweed within mature Pinus patula compartments in KwaZulu Natal. The SVM when utilized with a recursive feature elimination (SVM-RFE) approach required only 17 optimal wavebands from the original image to produce a classification accuracy of 93% and True Skills Statistic of 0.83. Results from this study indicate that (1) there is definite potential for using SVMs for the accurate detection and mapping of bugweed in commercial plantations and (2) it is not necessary to use the entire 272-waveband dataset because the SVM-RFE approach identified an optimal subset of wavebands for weed detection thus enabling improved data processing and analysis.

Narrowband normalized difference spectral indices (SI) have found wide application in vegetation studies. Consequently, several studies have investigated the utility of optimized SI for targeted applications. The objective of this study is to statistically develop optimized two-band normalized difference SI from a subset of hyperspectral bands derived using the Boruta wrapper algorithm. These indices are applied to model Fusarium circinatum stress in Pinus radiata seedlings. The performance of our developed optimized indices was compared with a selection of widely used existing SI ( n = 111) noted in the literature. Analyses were undertaken within a univariate (using the Jeffries–Matusita distance) and a multivariate (using the random forest algorithm) framework. Our results clearly demonstrate the improved accuracies using optimized SI (overall accuracy ranged from 76% to 96%) compared with using existing indices (overall accuracy ranged from 83% to 90%). Additionally, our results show that a multivariate approach yields superior results compared with a univariate approach. Overall, the results demonstrate the operational potential of optimized two-band normalized difference SI within a multivariate paradigm.

Defoliation induced by the weevil Gonipterus scutellatus is causing significant damage to South Africa’s eucalyptus plantations. Therefore, the ability of remote sensing to detect and map G. scutellatus defoliation is essential for monitoring the spread of the weevil so that precautionary measures are set in place. In our study, an integrated approach using image texture in various processing combinations and an artificial neural network (ANN) were developed to detect and map G. scutellatus induced vegetation defoliation. A 0.5-m WorldView-2 pan-sharpened image was used to compute texture parameters from the gray-level occurrence matrix and gray-level co-occurrence matrix, using optimal moving windows for specific levels of G. scutellatus induced vegetation defoliation. In order to improve the accuracy of detecting and mapping G. scutellatus induced vegetation defoliation, a method involving a three-band texture processing combination was proposed and tested. Using a sequential forward selection algorithm allowed for the selection of optimal texture combinations, which were subsequently input into a backpropagation ANN. The results showed an improvement in detecting vegetation defoliation using single texture bands [R2 = 0.82, root mean square error (RMSE) = 0.95 (1.82% of the mean measured defoliation)] when compared to single spectral reflectance bands [R2 = 0.60, RMSE = 1.79 (3.43% of the mean measured defoliation)], two-band spectral reflectance combination model [R2 = 0.74, RMSE = 1.48 (2.83% of the mean measured defoliation)], and the three-band spectral reflectance combination model [R2 = 0.80, RMSE = 1.35 (2.59% of the mean measured defoliation)]. Further improvements were obtained using the two-band texture combination model [R2 = 0.85, RMSE = 1.05 (2.01% of the mean measured defoliation)] and the most promising result was obtained using the proposed three-band texture combination model [R2 = 0.90, RMSE = 0.85 (1.63% of the mean measured defoliation)]. Overall, our study highlights the potential of image texture combinations in improving the detection and mapping of vegetation defoliation.

The aim of this study was to investigate the potential use of leaf-level hyperspectral data to predict the density of sugarcane thrips Fulmekiola serrata (Kobus). A hand-held spectroradiometer was used to make the spectral measurements on spindle leaves of 4- to 5-month-old plants of sugarcane cv. N19 growing in commercial fields near Umfolozi, South Africa. A random forest algorithm followed by partial least squares regression was used for the analysis. Developed models were adequate to predict nymph numbers in December and adult numbers in March, but different models were needed for the thrips life stage assessed and the season when the estimation took place.