Browsing by Author "Carrick, Peter"

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    Open Access
  • Thumbnail Image
    Item
    Open Access
    Exploring the relationship between restored ecosystem function and species composition: a meta-analysis
    (2013) Forsythe,Katherine Jane; Carrick, Peter; Amar, Arjun
    The idea that biodiversity confers higher levels of ecosystem functioning has been used as an argument for the conservation of biodiversity, and the relationship between these variables has been well researched. Ecosystem restoration often aims to restore former ecosystem functioning as well as specific species assemblages, not just biodiversity. In many cases monitoring programmes lack funds to measure all these aspects and often assume relationships between these variables; however, these relationships remain largely untested. In this thesis, I undertake the first ever meta-analysis of studies measuring ecological function and species composition in restored sites to test whether such a relationship exists. Bray-Curtis similarity measures were used to compare species composition for each site compared to multiple reference sites. Indicators of ecosystem function were also compared to reference sites, taking into account natural variation within reference sites. A weak relationship between ecosystem function and species composition was found, and it differed between different groups of ecosystem functions. Live plant biomass and structural framework ecosystem functions increased in similarity to reference sites as species composition also increased in similarity to reference sites. Nutrients, soil attributes and interactions between biotic component as well as litter and deadwood production, showed little association with species composition, with levels of ecosystem functioning showing little change as species composition became closer to reference sites. Variables relating to nutrients, soil attributes, and biotic interactions were always similar to intact sites regardless of the degree to which species composition had been restored. Live plant biomass, litter and deadwood production, and framework ecosystem functions, however, often did not reach reference levels of functioning, even when full species composition was restored. This analysis found that overall the shape of the relationship indicates redundancy in species composition, suggesting that increasing similarity in terms of species composition initially increases ecosystem function; however beyond a point, additional similarity does not further assist in the restoration of ecosystem function to reference levels of similarity. The main difference to biodiversity ecosystem function relationships comes with the fact that all ecosystem functions were not returned even when restored sites had similar species composition to reference sites, indicating that something, possibly time, is limiting the return of full ecosystem function to restoration sites. Finally, this research indicates that the relationship between species composition and ecosystem function is not consistent and it would be unwise to use species composition or ecosystem function as proxies for one another.
  • Thumbnail Image
    Item
    Open Access
  • Thumbnail Image
    Item
    Open Access
    Spatio-temporal distribution of grass and shrubs at the ecotone between an arid grassland and succulent shrubland : ecological interactions and the influence of soils
    (2007) Shiponeni, Ndafuda Nangula; Allsopp, Nicky; Hoffman, Timm; Carrick, Peter
    This thesis presents ecological investigations of vegetation at an ecotone between the Namaqualand shrublands and Bushmanland arid grassland at the climatic transition between the predominantly winter rainfall Succulent Karoo and the predominantly summer rainfall Nama-Karoo biomess. The work focused on the determinants of grass and succulent shrub vegetation characterising the ecotone, and on determining the current and past distribution of this vegetation. A combination of detailed community the current and past distribution of this vegetation. A combination of detailed community level investigations (both observational and experimental), and a variety of remotely sensed data and techniques were applied to capture processes and patterns at different scales. In the first set of investigations, I have investigated the role of soil and competition on vegetation patterns. Secondly, field data were combined with remote sensing technology to establish the current distribution of grasses and shrubs along the ecotone, and to determine how this vegetation has changed over the past 20 years.
  • Thumbnail Image
    Item
    Open Access
    Untangling clumps-factors influencing seedling ecology in a semi-desert, and the implications for restoration ecology
    (2010) Krüger, Raldo; Carrick, Peter
    Facilitation between plants and patch dynamics are two common ecological processes in semi-arid regions. I tested the relevance of these processes to seedling establishment of perennial species on degraded landscapes in Namaqualand, South Africa. The species were grouped into two functional guilds, Mesembryanthema (formerly the family Mesembryanthemaceae, but now part of the Aicoaceae; hereafter Mesemb), which rely on CAM to a variable degree and are shallow rooted, and non-Mesembryanthema (hereafter non-Mesemb) that utilize C3 photosynthesis and have deeper root systems.
  • Thumbnail Image
    Item
    Open Access
    Using ecological concepts to restore the drylands of Namaqualand, South Africa
    (2022) Becker, Carina Helene; Hoffman, Timm; Carrick, Peter
    This thesis investigated the impact of cultivation, and the efficiency of passive and active restoration of fallow fields, in Namaqualand's Hardeveld and Renosterveld vegetation communities. The core theory of this thesis lies in the ecology of semi-arid environments and the concept of patch dynamics. In such areas plants grow together, creating communities with a distinct patch/inter-patch structure. Patches of vegetation concentrate more organic matter, nutrients and moisture, in ‘islands of fertility' as well as providing protection for seedlings from harsh elements, trampling and herbivory. The objectives of this thesis are to assess the efficacy of passive restoration and why it might not work, and to explore the value of patch dynamics as a theoretical framework for developing restoration approaches. I used these ideas in three separate studies as they pertain to two main vegetation types in the Kamiesberg area of Namaqualand, South Africa. These are the Kamiesberg Mountain Renosterveld surrounding the village of Leliefontein, and Hardeveld vegetation surrounding the village of Tweeriver. Both vegetation types are located within the Leliefontein commonage, are subjected to the same disturbance patterns and occur on the same underlying geology with similar soils, but are differentiated by their altitude and rainfall. In the first study I tested the hypothesis that cultivation had removed and compromised the structure and functioning of patches, and that these features do not improve over time. In each vegetation type, I surveyed a 100 m vegetation transect on four fields in two age classes respectively. The age classes represented fallow lands that had not been cultivated for a period of 7 – 20 years (“young fields”) and 30 – 60 (“old fields”) years. This was compared to eight, 100 m transects taken from intact communities under two different grazing pressures. These are communal land which was exposed to regular, even daily, grazing and private farms which experience rotational grazing and lower grazing pressure. The transects recorded all perennial species in the growth forms of herbs, grasses, small and medium leaf succulents, stem succulents, and small and medium non-leaf succulent shrubs. The results indicated that passive restoration is not effective in either vegetation type, as fallow fields rested between 7 to 60 years were not distinguishable from each other, suggesting that there had been no improvement over time. The removal of a diverse community through ploughing leads to the establishment by pioneer plants of Dicerothamnus rhinocerotis (renosterbos) and Galenia africana (kraalbos), with few other species occurring in the landscape. The dominance of these two species persists for decades. In this first study I also tested the hypothesis that the recovery of vegetation on fallow fields in Namaqualand is hindered by a decline in soil condition and the removal of fertile islands in the landscape. In each vegetation type, on three old fields, and on three intact communities, soil samples were taken from between shrubs (three samples) and from under the dominant patchforming shrubs (three samples) respectively, to determine the role that plant patches have on the formation of fertile islands beneath their canopies. Results indicate that soil characteristics are substantially altered after cultivation. Even after decades, the levels of organic matter and nutrients in old lands of both vegetation types are lower than in the intact community. In Hardeveld there was evidence that the pioneer shrub, kraalbos, does create a fertile island under its canopies, gradually accumulating organic matter and carbon, but at very low levels. G. africana also increased the pH and phosphorus levels beyond levels found in the predisturbed system. In Renosterveld vegetation, soils on fallow fields showed no signs of fertile island formation or improvement in soil nutrients. In the second study, which was undertaken in a greenhouse, I tested the hypothesis that the recovery of fallow fields is hindered by a depleted perennial soil seed bank. In each vegetation type, seedlings that germinated from the top 5 cm of soil collected from beneath and around the edge of 12 plant patches from four land types (young and old fields, and communal and private reference farms) were identified and classed into the same growth forms as listed earlier. The results supported the hypothesis, as the seed bank on fallow fields consisted almost entirely of annuals with kraalbos and renosterbos the only perennial plants to germinate. In the final study I set up factorial restoration experiments in the field at three sites in each of the two vegetation types, and tested the efficacy of different restoration interventions on perennial species over a three year period. Each treatment contained 45 replicates. The species were grouped into four growth form types: grasses, herbs, leaf succulent shrubs and non-leaf succulent shrubs. In Hardeveld, succulent shrubs and grasses established well, whilst in Renosterveld it was grasses and herbs that were most successful. The experiment used pioneer plants to establish the original community. The thesis hypothesized that pioneer plants create a fertile island under their canopies and that this could be used to improve seedling establishment success. This was done by sowing seeds in open control plots and plots where the above ground portion of kraalbos and renosterbos were removed. The results did not support this hypothesis as there was no noticeable benefit to seedlings growing on a fertile island plot as compared to an open area. I also hypothesized that adding fertilizer (nutrients) would increase the success of seedling establishment because this would re-create the more nutrient rich patch environment seedlings would be growing in. Results show that it was only succulent shrubs (Hardeveld) and grasses (Renosterveld) that responded significantly to nutrient addition. The other growth forms had no noticeable benefit. The most unexpected result was that this treatment resulted in a decline of herb seedlings in Renosterveld. To test the hypothesis that nurse plants or shelters play an important role in restoration success, unsheltered control plots were compared to plots sheltered with brush packs, square boxes or a pioneer nurse plant. Results supported this hypothesis, as box and brush pack treatments were successful in increasing seedling numbers. In the Hardeveld, boxes were very effective for non-succulent shrubs, succulent shrubs, and grasses. In Renosterveld, there was a clear preference by all growth forms for brush pack shelters, whilst the open boxes did not have much influence. Finally, I hypothesized that a nurse plant would facilitate seedling establishment, if there is no competition between the seedling and the nurse plant. Kraalbos, as a nurse plant, provided no noticeable benefit to seedling establishment as they had similar results to unsheltered controls for all growth forms. The only indication of facilitation by a nurse plant was with renosterbos and herb seedlings. In this thesis, I contribute conceptually and theoretically to our understanding of patch dynamics, by analysing their role in intact systems and how these dynamics recover over time after degradation. Passive recovery of Renosterveld and Hardeveld has not occurred in the vegetation types and areas that I surveyed. Without active re-introduction of perennial seeds to fallow lands seed will not reach these fields. Although soil condition was seen to be improving under kraalbos, the high pH and phosphorous restrict plant growth of the original community. In Renosterveld, the fertile island does not re-form. Results from this study show that different treatments have different outcomes in different vegetation types. Restoration practitioners should carefully consider the environment they are working in, as well as consider the requirements of different growth forms in restoration planning. Some growth forms establish well initially and focus should be to return these first, in order to improve cost effectiveness. In the Hardeveld succulent shrubs and grasses established well over the three years and in Renosterveld it was grasses and herbs.