The effects of invasive trees in riparian zones and implications for management and restoration : insights from Eucalyptus invasions in South Africa

[摘要] ENGLISH ABSTRACT: Worldwide, invasive alien plants (IAPs) alter aspects of invaded ecosystems including geomorphology, above-ground vegetation, soil seed banks and soil nutrient regimes, thereby affecting the long-term stability of ecosystems. In many cases these invasions call for various management interventions, including restoration. Effects of alien plant invaders on native ecosystems are widely acknowledged, but compared to terrestrial ecosystems, riparian habitats are poorly studied. Riparian habitats are inherently rich in biodiversity, but are particularly prone to invasion by IAPs because of their dynamic nature and because these ecosystems are affected by many anthropogenic activities. This enhances the proliferation of IAPs, especially trees such as Tamarix spp. (e.g. in the southwestern United States), Salix spp. (e.g. in Australia), and Australian Acacia spp. and Eucalyptus camaldulensis (in South Africa). Initiatives such as the Working for Water (WfW) programme in South Africa that are focusing on clearing IAPs, particularly in riparian zones, have reported much success in the short term, but ecosystem recovery remains limited. There is a poor understanding of the nature of the impacts of invasive trees, and of the opportunities that exist for ecosystem restoration. The objective of my thesis was to investigate the effects of the widespread invasions of Eucalyptus camaldulensis in riparian zones.Methodological constraints have been identified as one reason for limited success of restoration projects. Many studies examining the effects of IAPs used a comparative approach of uninvaded (reference) and invaded sites. However, this approach often fails to separate cause from effect. An experimental removal approach has been suggested as an alternative, but it is often inappropriate for large invasive trees owing to time and cost constraints. I used a comparative gradient approach, involving sampling along a continuum from uninvaded sites, and varying E. camaldulensis invasion densities categorised as 'lightly, 'moderately and 'heavily invaded (hereafter the gradient of invasion), each with four replicates (n = 16). This approach enabled me to detect trends that could be reliably attributed to invasion. Furthermore, my study attempted to provide a comprehensive understanding of the effects of IAPs in invaded ecosystems by considering various components of the ecosystem and integrating their feedback relationships to establish the net effects on ecosystem processes. The riparian zone is a unique ecosystem which provides an opportunity for such a multifaceted study examining and integrating the effects of invasive trees on riparian geomorphology, above-ground vegetation, soil seed banks, and soil properties.I studied the effects of E. camaldulensis invasion in riparian zones along the Berg River in the Mediterranean-type climate zone of the Western Cape, South Africa. I asked the questions 1) Are environmental conditions (i.e. the riparian environment created by geomorphological processes) altered by Eucalyptus invasion and if so, are these conditions less suitable for recruitment of indigenous species? 2. Has Eucalyptus invasion altered the composition of indigenous species (potentially limiting post-clearing propagule supply)? 3: How does Eucalyptus invasion change the composition of the riparian soil seed bank; and its potential to re-initiate ecosystem recovery? 4: Do soil properties (physical and chemical) change with Eucalyptus invasion?Data for all four questions were collected in the same sites to enable cross comparisons. For the first question, I employed photogrammetry techniques in a GIS and remote sensing environment to analyse repeated aerial photographs (1938 -2010) to reconstruct the invasion history and riparian geomorphology evolutionary dynamics. For the second and third questions, I compared richness, diversity, evenness and composition of resident above-ground and seed bank vegetation between uninvaded sites and the gradient of invaded sites. For the fourth question, I compared a set of variables describing physico-chemical properties in uninvaded sites to the gradient of invaded sites seasonally.Results of the geomorphology study showed that the riparian zone has been very dynamic over the 70 years, with a 13.5% net decrease in area. There was evidence of channel narrowing, riverbank steepening and river bed incision in areas that were densely invaded. No significant trends were detected in hydrometeorological data. The above-ground vegetation study revealed that species richness, diversity and structural attributes (e.g. height, relative cover and mean basal diameter) of native species decreased consistently along the invasion gradient. Invasion also altered native and alien plant species composition, both of which showed a high preference for lightly invaded sites. The seed bank study revealed that E. camaldulensis invasion had no significant effect on total and native species richness, diversity and evenness, however the effects were bigger on native than alien species. Alien species density was significantly higher than native species density. Invasion also influenced the composition of the native seed bank. However, native seed banks were more diverse than above-ground plant assemblages with some native fynbos species occurring only in the seed bank. In the soil study, I found that soil pH levels were significantly lower in invaded sites than in uninvaded sites in all seasons. Litter cover and thickness increased significantly with invasion intensity. Soil moisture decreased consistently with invasion intensity while temperature increased in winter and spring. Total macro, micro and available nutrients did not vary significantly along the invasion gradient (p > 0.05), but exchangeable cation content was significantly higher in uninvaded than in invaded sites, especially in winter and spring.Narrowing of river channel and steepening of river banks reduces the suitability of the riparian zone to perform certain ecological functions such as soil seed storage and providing space for vegetation to grow. This reduces opportunities for seedling recruitment and consequently reduces species diversity. Conversely, the reduction of species diversity in above-ground vegetation reduces seed input into the soil leading to depauperate soil-stored seed banks. The abundance of seeds of Acacia mearnsii in the soil-stored seed bank poses a threat of secondary invasions post clearing of E. camaldulensis if this is adopted as a management action.The study was successful in addressing the stated objectives. The transition from a native-dominated riparian plant community to a monoculture of E. camaldulensis over c. 50 years has resulted in marked changes to riparian geomorphology and above-ground vegetation, but has yet to radically change the total soil seed bank and soil nutrient concentrations. The findings offer support to the WfW clearing initiatives and show that potential for recovery of native species after removal of the invasive eucalypts still exists. This knowledge not only enhances effectiveness of restoration and long term management of riparian ecosystems, but also advances the field of riparian ecology and restoration. Further experimental work is needed to establish the mechanisms responsible for the changes associated with Eucalyptus invasion. The contribution of agriculture in shaping channel morphology also needs to be investigated.

[发布机构] Stellenbosch University