No description available.

No description available.

There is growing evidence that the achievement of biodiversity conservation objectives at the regional scale will have to take into account not only habitats within reserves, but also the contribution of semi-natural habitats in production landscapes. This requires knowledge of the conditions that maintain or enhance biodiversity in these suboptimal habitats. I focused on a system of forest patches and hedgerows in an agricultural landscape as habitats for dispersal-limited native forest herbs to (1) identify species-specific traits that appear to either limit or facilitate survival or colonization of forest herbs in hedgerows compared to forest patches, and (2) to evaluate the relative contribution of landscape and local variables in explaining patterns of forest herb species richness, abundance, and diversity in hedgerows. A fourth-corner analysis revealed that traits such as early spring flowering, ant dispersal or a lack of vegetative propagation can reduce the likelihood of native forest species to benefit from hedgerows. Conditional autoregressive models showed that landscape variables, specifically hedgerow age and neighboring forest area, largely account for species diversity in hedgerows.

Thesis (MScFor)--University of Stellenbosch, 2000. / ENGLISH ABSTRACT: The windbreak species near Wolseley (WoSl and WoS3) was Casuarina cunninghamiana. The windbreak species near Saron (SaSl) and Villiersdorp (ViSl) was Eucalyptus cladocalyx and Pinus radiata respectively. A shelter effect was indicated at ViSI (from northerly winds in winter/early spring 1999), and WoS3 (from southerly winds in late spring 1999). Wind speeds were consistently reduced in the leeward sheltered zone ofViSl and WoS3. To improve the correlation of the shelter effect, mild contaminating winds (comprising approximately 28% of the total data recorded at each site) were excluded for the prediction equations of the shelter effect at each site. Compared to the more exposed wind speeds at 1.0 H to the windward side, wind speeds at ViS] were reduced by 32% at 3.0 H. Compared to 11.0 H, wind speeds at ViSl were reduced by 49 and 46% at 3.0 and 1.0 H respectively. With r2 values above 93%, accurate linear prediction equations were produced. The early barley damage assessments indicated that damage was absent or negligible at 11.1 H to the lee, and significantly highest (X < 0.05) at 13.7 H to the lee and beyond. The crop shelter effect significantly increased (p < 0.05) barley head number, transformed proportion of total barley head mass from above-ground mass, potato tuber mass, potato above-ground mass total potato tuber number, small tuber number, and disproportionate tissue damage to the windward section of each plant. For WoS3, the shelter effect became pronounced with severe south-easterly winds. The average hourly wind speeds at 3.0 H dropped from 0.9 m / s (in the previous and less windy sampling period) to 0.8 m / s, despite the sharp increase in exposed wind speeds. Compared to 13.0 H, wind speeds were reduced by 73 and 32% at 3.0 and 7.0 H respectively. With r2 values above 94%, accurate linear prediction equations were produced. The crop shelter effect significantly increased (p < 0.05) sub-sample mass of lOO grains. At WoS3, strong and sustained wind speeds caused leeward soil temperature increases of up to 4°C at 3.0 H, compared to 11.0 H. Brief strong winds (characteristic of winds at ViSl) had little effect on the soil temperature differences. At ViSl, a deviation of soil moisture content between 3.0 and 11.0 H, following periods of recharge, indicated a potential soil-moisture conservation effect in the sheltered zone. This did not occur at WoS3, due partly to a very low soil moisture content that had little scope for variation; From the crop variations and the microclimate variations at both ViS 1 and WoS3, the maximum shelter effect extended to approximately 4.0 H, followed by an intermediate zone of diminishing shelter that extended to approximately 9.0 H. For ViSl, WoS3 and SaSl, a shading effect significantly reduced yields (p<0.05) at 1.0 H from the respective northern windbreak, compared to yields at 2.0 H. At ViS 1 and WoS3, soil probes did not indicate a depletion of soil moisture resulting from the respective windbreaks. Soil moisture competition was indicated on the northern side of the WoSl windbreak with drier conditions; where soil moisture levels at 1.0 and 3.0 H diverged from a negligible level to a 22% lower level (p > 0.05) at 1.0 H, compared to 3.0H. / AFRIKAANSE OPSOMMING: Die boomsoorte wat as windbreke gedien het was Casuarina cunninghamiana naby Wolseley (WoS 1 en WoS3), en Eucalyptus cladocalyx en Pinus radiata by Saron (SaS 1) en Villiersdorp (ViS 1) respektiewelik. Beskutting is aangetoon by ViS 1 (teen noordelike winde in winter/vroee lente 1999) en WoS3 (teen suidelike winde in laat lente 1999). Windspoed is konsekwent aan die lykant sones van ViS 1 en WoS3 verminder. Ten dien einde die korrelasie van die beskuttings effek te verbeter, is matige kontaminerende winde (ongeveer 28% van die totale waargenome data by elke plek) buite berekening gelaat vir die voorspellingsvergelykings van die beskuttingseffek. In vergelyking met die meer blootgestelde windspoede by 1.0 H aan die windkant, is windspoede by ViSl met 32% verminder by 3.0 H. In vergelyking met 11.0 H is windspoede by ViSl met 49 en 46% by 3.0 en 1.0 H respektiewelik verminder. Met r2 waardes hoer as 93%, is akkurate lineere voorspellingsvergelykings verkry. Die vroee gars skade opnames het aangetoon dat geen of minimale skade by 11.1 H aan die lykant aangerig is, en beduidend die hoogste (x< 0.05) by 13.7 Hen verder aan die lykant was. Die gewas beskuttins effek het beduidende toenames (p< 0.05) in garsare, getransformeerde verhouding van totale gars-are massa van bo-grondse massa, aartappelknol massa, aartappel bo-grondse massa, totale aantal aartappeiknolle, en oneweredige weefsel beskadiging aan die windkant van elke plant tot gevolg gehad. Vir WoS3 het die beskuttingseffek beduidend geword met baie sterk suidoostelike winde. Die gemiddelde uurlike windsnelhede by 3.0 H het van 0.9/ (in die vorige en minder winderige toetstydperk) tot 0.8 m/" verminder, ten spyte van die skerp toename in blootgestelde windsnelhede. In vergelyking met 13.0 H is windsnelhede met 73 en 32% by 3.0 en 7.0 H respektiewelik verrninder. Met r2 waardes hoer as 94% 15 akkurate lineere voorspellingsvergelykings verkry. Die gewas beskuttingseffek het sub-monster massa van 100 graankorrels beduidend (p< 0.05) verbeter. By WoS3 het sterk en volgehoue windsnelhede grondtemperature aan die lykant met tot 4°C by 3.0 H verhoog in vergelyking met 11.0 H. Kortstondige sterk winde (tipiese winde by ViS1) het weinig uitwerking op grondtemperatuursverskille gehad. By ViS1 het 'n afwyking van grondwaterinhoud tussen 3.0 en 11.0 H na tydperke van aanvulling, 'n potensiele grondwaterbewaringseffek in die beskutte gebied getoon. Soortgelyke tendense het nie by WoS3 voorgekom nie, gedeeltelik as gevolg van 'n baie lae grondwater inhoud wat nie veel kon varieer nie. Van die variasies wat in gewasse en mikroklimaat voorgekom het, by beide ViSl en WoS3, kan afgelei word dat maksimum beskutting tot by ongeveer 4.0 H verleen is, gevolg deur 'n intermediere sone van verminderende beskutting tot ongeveer 9.0 H. Vir ViS1, WoS3 en SaSl het 'n skadu-effek gewasopbrengs beduidend (p< 0.05) by 1.0 H verminder in vergelyking met opbrengste by 2.0 H. By ViS 1 en WoS3 het grondwater strooiingspeilers nie 'n uitputting van grondwater as gevolg van die windbreke getoon nie. Kompetisie vir grondwater is getoon aan die noordelike kant van die WoS1 windbreek met droer toestande. Grondwatervlakke by 3.0 H het minimaal maar by 1.0 H met 22% beduidend (p> 0.05) teenoor 3.0 H gedaal.

No description available.