SUNScholar will be down for routine maintenance from 2018-12-11 10:00 SAST.

Conservation of spider diversity within an agricultural mosaic : insights from the Greater Cape Floristic Region, biodiversity hotspot.

Theron, Kasselman Jurie (2017-12)

Thesis (MScConsEcol)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Agricultural expansion is one of the main drivers of habitat fragmentation and land use change which negatively impacts biological diversity. The Greater Cape Floristic Region (GCFR), a biodiverse hotspot, has been recognised as a priority for conservation as its unique endemic diversity is threatened by historic land transformation and habitat fragmentation. Private nature reserves and proclaimed protected areas alone cannot conserve all biodiversity, especially with >80% of land not formally protected. Thus we must conserve biodiversity within production landscapes. Remnant patches of natural vegetation supports a wide variety of arthropod taxa. However, little information is available on spider diversity in remnant fynbos and even less on which environmental parameters drive this diversity. Furthermore, research on how the matrix impacts adjacent remnant patches, and how spiders respond to different matrix types, are needed for protecting spider diversity and the services they provide within the GCFR mosaic. This study aims to identify environmental parameters that shape spider diversity within fynbos remnant patches, and how spiders respond to different matrix types. Here, I sampled spider diversity within remnant fynbos patches of the GCFR mosaic to identify which landscape and patch variables are important for maintaining spider diversity. Fifteen environmental variables (at landscape and patch scales) were collected at each site and analysed to determine their influence on spider species richness and assemblage structure of the whole spider assemblage, and for different functional guilds. Local patch variables best predict spider diversity, particularly soil compaction and topographic complexity which negatively influenced overall and plant dwelling spider richness. This pattern of complexity is mainly driven by common spider species. Tree species richness (mostly alien trees) negatively influenced free-living spider richness. Lastly, level of site invasion by alien trees influenced overall and epigaeic spider assemblage structure. Spider diversity was more influenced by patch scale variables, which reflects local patch management, than the landscape context. I also assess how spider diversity responds to different land-use types, the magnitude of associated edge effects on spider diversity, and identify complementary habitat elements for enhancing spider diversity within agricultural mosaics of the GCFR. Spider diversity was sampled along replicated transects covering remnant fynbos vegetation into three different matrix types: old fields, vineyards and invasive alien tree stands. Fynbos remnants had significantly higher overall spider diversity than matrix sites with higher diversity in edge locations than at patch cores. Old fields had the highest spider diversity between all land-use types, as well as the greatest assemblage similarity to remnant vegetation assemblages. Lowest diversity was recorded within vineyards. Lastly, vegetation complexity enhanced spider diversity across all land-uses. In conclusion, I demonstrate that remnant vegetation is a critical landscape element for conserving spider biodiversity in GCFR mosaics, but that old fields can play an important role in increasing functional connectivity within the landscape mosaic. Increasing native vegetation diversity within the matrix helps improve spider diversity. Additionally, this work recommends alien tree removal from fynbos remnant patches within the GCFR for biodiversity conservation. Preserving remnant patches of all sizes in production landscapes, and softening the matrix, can increase heterogeneity which benefits spider diversity within the GCFR mosaic.

AFRIKAANSE OPSOMMING: Die omskepping van natuurlike habitat vir landbou is een van die groot oorsake wat landskap fragmentasie veroorsaak, en lei to verskeie negatiewe probleme vir biodiversiteit. Die Goter Kaapse Floristiese Streek (GKFS), ‘n asemrowende biodiverse streek, is bedreig as gevolg van historiese landskap verandering en fragmentasie wat ons unieke endemiese diversiteit bedreig. Met >80% van natuurlike fynbos wat nie onder formele bewaring is nie, moet ons biodiversiteit bewaar in produksie landskappe. Oorblywende natuurlike fynbos fragmente ondersteun ‘n groot verskeidenheid van verskillende geleedpotiges. Daar is alhoewel baie min informasie beskikbaar op spinnekop diversiteit, en nog minder informasie op watter omgewings-veranderlikes hierdie patroon van spinnekop diversiteit beinvloed. Verder, navorsing op hoe die produksie landskappe aangrensende natuurlike fynbos fragmente beinvloed, en hoe spinnekoppe reageer to verskillende grondgebruik tipes, is benodig om spinnekop en hul dienste te bewaar in die produksie landskap van die GKFS. Hierdie projek mik om omgewings-veranderlikes te identifiseer wat spinnekop diversiteit binne natuurlike fynbos fragmente beinvloed, en hoe spinnekoppe reageer in verkillende grondgebruik tipes. Hier het ek spinnekop diversiteit versamel binne natuurlike fynbos fragmente in die GKFS se produksie landskap, om te sien watter landskap- en plaaslike veranderlikes belangrik is om spinnekop diversiteit te onderhou. Vyftien omgewings-veranderlikes (op die landskap en plaaslike skaal) was by elke fragment versamel en ontleed om hul invloed te bepaal op spinnekop rykheid en gemeenskap struktuur van die hele spinnekop gemeentskap, en van verskeie funksionele groepe. Plaaslike veranderlikes, veral grond kompaksie en topografiese kompleksiteit wat algehele en plant bewonende spinnekop rykheid negatief beinvloed, was die mees beduidende veranderlikes om spinnekop diversiteit te bepaal. Hierdie patroon van kompleksiteit is hoofsaaklik gedryf deur algemene spinnekoppe. Boom rykheid (meestal indringer bome) het ‘n negatiewe impak op vry-lewende spinnekop rykheid gehad. Laastens, die verspreiding van indringer bome in fynbos fragmente het algehele en grond bewonende spinnekop gemeenskappe beinvloed. Spinnekop diversiteit was meer beinvloed deur plaaslike veranderlikes, wat plaaslike bestuur weerspieël, as die konteks van die landskap. Ek het ook gekyk na hoe spinnekop diversiteit reageer in verskillende grondgebruik tipes, die skaal van geassosieerde rand effekte op spinnekop diversiteit, en om aanvullende habitat elemente te identifiseer wat spinnekop diverseteit verbeter in die GKFS produksie landskap. Spinnekoppe was versamel in natuurlike fynbos fragmented (in die kern en op die rand) en dan ook in die aangrensende grondgebruik tiepe (in die kern en op die rand). Drie verskillende grondgebruik tipes was gebruik: ou velde, wingerd en uitheemse boomstande. Natuurlike fynbos fragmente, spesifiek die fragment rand, het aansienlik hoër algehele spinnekop diversiteit gehad as al die ander grondgebruik tipes. Ou velde was die mees diverse grondgebruik tipe, en het die grootse spinnekop gemeenskap ooreenkoms gehad met natuurlike fynbos fragmente. Wingerd het die laagste spinnekop diversiteit gehad. Laastens, die kompleksiteit van natuurlike plantegroei in al die verskeie produksie landskappe, het spinnekop diversiteit verbeter. Om af te sluit, hier het ek gewys dat natuurlike fynbos fragmente ‘n belangrike landskap element is om spinnekop diversiteit in die GKFS se produksie landskap te bewaar. Verder, ou velde het die vermoë om funksioneël landskappe te verbind. Ook, om die kompleksiteit van natuurlike plantegroei binne die produksie landskap te verhoog, help om spinnekop diversiteit te bewaar. Laastens, dit word aan beveel om uitheemse bome in fynbos fragmente te verwyder vir bewaring van biodiversiteit. Om fragmente te beskerm en herstel, ongeag van grootte, en om verskeie grondtipes te versag, sal heterogeneiteit verhoog wat spinnekop diversiteit in die produksie landskap van die GKFS bevoordeel.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/102895
This item appears in the following collections: