Human related activities have influenced the rivers of the southern Western Cape since as early as the 1700’s. As there is no detailed information available on ecological status of the Holsloot River, a tributary of the Breede River, this study aimed to gain insight into the effect of impacts associated with human activities on the habitat integrity of this river. The study intended to understand how seasonal changes, catchment characteristics and events are reflected in the ecological status of habitats along the river by applying bio-monitoring and river health measurements at selected sites in the upper, middle and lower reaches of the Holsloot River and compare the results to that of an undisturbed reference site. Results obtained in this study are compared with data gathered in 2008/2009 to determine if the ecological status of the river had changed in the period between the two sampling times. The study included assessment of the ecological status of the river based on standard bio-monitoring protocol (SASS5, IHI, IHAS and VEGRAI) as well as in situ water quality analysis (pH, dissolved oxygen, electrical conductivity and total dissolved solids). The construction of the instream Stettynskloof Dam changed the configuration of the riparian zone and river channel in the upper catchment area. Agricultural- and other human related activities, with consequent water abstraction, non-point-source pollution, loss of riparian vegetation, as well as dense stands of alien invader plants influence flow patterns and affects river ecology, especially in the dry summer months. Providing sufficient stream flow and adequate water levels, human related activities can create a larger variety of habitat types available that can support larger biodiversity and higher productivity. The level of inundation and stream flow, influenced by water abstraction as well as irrigation return-flow from extensive drainage systems especially in the dry months, contribute to the loss of biodiversity in the middle and lower reaches of the river. Where the upper reaches of the river are largely natural with few modifications, the habitat integrity deteriorates in the middle reaches so much so that ecosystem functioning are collectively impaired in lower reaches due to human related impacts. Sensitive macro-invertebrates found at lower seriously impacted parts of the river however, were in all probability washed down from lower impacted upstream habitats and may expectedly be able to again occupy habitats downstream if water quality and habitat availability improves. / Environmental Sciences / M. Sc. (Environmental Science)

Plant invasions represent a major ecological and socio-economical issue and understanding the drivers as well as consequences of plant invasions is thus one of the main goals of plant ecology. It is equally important to reveal general patterns underlying plant invasions and to understand the details of biology of individual invaders. In this thesis I explored plant-soil feedback (PSF) as a possible general mechanism underlying plant invasiveness, and also focused in detail on drivers and consequences of Impatiens parviflora invasion. The aims of this thesis were to i) assess the differences in intraspecific PSF between invasive and alien non-invasive species using a large set of species; ii) explore the relationship between PSF, residence time and phylogenetic novelty of the alien species; iii) compare the importance of PSF and other plant characteristics for plant invasiveness; iv) compare PSF between invasive and native congeners of similar level of dominance in the field; v) evaluate the effect of cultivating conditions on results of PSF experiments; vi) describe invasion dynamics and determine factors affecting spread of invasive I. parviflora using a method of monitoring its natural spread in several types of habitats, and vii) assess the impact of I. parviflora on native vegetation of oak-...

Combiner les drones et l’intelligence artificielle afin de cartographier la végétation représente une nouvelle approche prometteuse susceptible d’améliorer la détection des espèces végétales exotiques envahissantes (EVEE). La résolution au sol élevée atteignable avec les drones et les innovations récentes en vision par ordinateur, notamment avec les réseaux neuronaux convolutifs, laissent penser qu'une détection précoce des EVEE serait possible, favorisant ainsi leur gestion. Dans cette étude, nous avons évalué la pertinence de cette approche pour cartographier l’emplacement du roseau commun (Phragmites australis subsp. australis) à l’intérieur d'un parc national situé dans le sud du Québec, Canada. Nous avons collecté des données à six dates distinctes durant la saison de croissance, couvrant des environnements présentant différents niveaux d'envahissement par le roseau. De façon générale, la performance du modèle était élevée pour les différentes dates et zones, surtout au niveau du rappel (moyenne globale de 0.89). Les résultats ont montré une augmentation de la performance pour atteindre un sommet à la suite de l’apparition de l’inflorescence en septembre (F1-score le plus haut à 0.98). De plus, une diminution de la résolution spatiale affectait négativement le rappel (diminution de 18% entre une résolution au sol de 0,15 cm pixel-1 et 1,50 cm pixel-1), mais n’avait pas un impact important sur la précision (diminution de 2%). Malgré des défis associés à la cartographie du roseau commun dans un contexte de gestion post-traitement, l'utilisation de drones et de l'apprentissage profond montre un grand potentiel pour la détection des EVEE lorsque appuyé par un jeu de données adapté. Nos résultats montrent que, d'un point de vue opérationnel, cette approche pourrait être un outil efficace pour accélérer le travail des biologistes sur le terrain et assurer une meilleure gestion des EVEE. / The combination of unoccupied aerial vehicles (UAVs) and artificial intelligence to map vegetation represents a promising new approach to improve the detection of invasive alien plant species (IAPS). The high spatial resolution achievable with UAVs and recent innovations in computer vision, especially with convolutional neural networks, suggest that early detection of IAPS could be possible, thus facilitating their management. In this study, we evaluated the suitability of this approach for mapping the location of common reed (Phragmites australis subsp. australis) within a national park located in southern Quebec, Canada. We collected data on six distinct dates during the growing season, covering environments with different levels of reed invasion. Overall, model performance was high for the different dates and zones, especially for recall (mean of 0.89). The results showed an increase in performance, reaching a peak following the appearance of the inflorescence in September (highest F1-score at 0.98). Furthermore, a decrease in spatial resolution negatively affected recall (18% decrease between a spatial resolution of 0.15 cm pixel-1 and 1.50 cm pixel-1) but did not have a strong impact on precision (2% decrease). Despite challenges associated with common reed mapping in a post-treatment monitoring context, the use of UAVs and deep learning shows great potential for IAPS detection when supported by a suitable dataset. Our results show that, from an operational point of view, this approach could be an effective tool for speeding up the work of biologists in the field and ensuring better management of IAPS.