Changes in the Berg River Basin over time

Magoba, Rozwivhona Faith (2018-12)

Thesis (PhD)--Stellenbosch University, 2019.

Thesis

ENGLISH ABSTRACT: The nature of river ecosystems is influenced by the history of activities in their basins. This dissertation investigated historic changes in the Berg River Basin and their influence on river ecosystem structure. The central assumption was that all activities in a river basin landscape contribute either directly or indirectly to the condition (physical, biological, chemical) of rivers that run through them. It was first necessary to establish what changes had taken place in the river basin over time and this was done in different ways at different spatial scales. Changes in land-use were collated and mapped across the basin since these were considered to influence the river’s flow regime and river channel structure. Predictions were made about how changes in flow and river channel habitat would influence the distribution and abundance of aquatic macroinvertebrates. A history of land-use changes over the Berg River Basin was explored between four periods, 1955-1965, 1976-1985, 1996-2005 and 2006-2015. The bulk of the dryland crop production was in the lower foothills and lowlands while the upper foothills comprised orchards, vineyards and forestry. From 1955-2015 the extent of agricultural land in the basin declined by half as dryland crops were changed to orchards and vineyards and large tracts of land were left fallow. Over the same period the area under forest declined by 73% and urban areas doubled in size as did the number of farm dams in response to the increased need for irrigation to supply the more water hungry crops. The effects of the changes in land-use, the increase in farm dams and the construction of large dams on the river’s flow regime was investigated next. Changes in flow were explored at four river gauges along the length of the Berg River up- and down-stream of the two main in channel dams; the Berg River Dam in the Upper Foothills and Misverstand Dam in the Lowlands. In general the changes were more marked at the downstream gauges and the trends were towards increased dry season flows and slightly decreased wet season flows due to release of water from, and capturing of floods by the in-channel dams to meet irrigation demand in the dry season. Flow pattern from early records was better correlated with rainfall than that from the recent record indicating that flow changes were likely to be attributable to anthropogenic effects such as land-use and water resource developments. Both land-use and water resource developments were predicted to have consequences on river channel shape and habitat that was investigated next. Changes in river channel shape, the extent and composition of the floodplain and riparian area was mapped from aerial photographs at five sites along the Berg River and at five adjacent tributaries. Each site responded differently, which was not unexpected, and reductions in the extent of the channel and riparian area were more severe along the Berg River main stem when compared to the tributaries. Along tributaries no floodplains were discernible at the scale measured, however a decreased in extent over time along the main river except downstream of the Berg River Dam where the floodplain area had increased due to the previously braided channel of 1938 changing to a single thread channel with floodplain and a greater area of sandbanks. Changes in river habitat, such as these, were predicted to effect change in the abundance and community structure of aquatic macroinvertebrates, which was investigated next. The abundance of aquatic macroinvertebrates was studied from the 1950s to 2015 and showed a reduction in simulids and baetids with an increase in the abundance of chironomids, indicating a decline in water quality. Changes in other groups indicated a decline in quality of habitat, for instance a loss of plecopterans that prefer clean gravel beds being replaced by caenids that prefer a sandy channel bottom. In 2015 there were also more groups of invertebrates that are associated with slow-flowing areas and marginal vegetation, which was presumed to have occurred in response to the clearing of woody alien trees from the river banks and the subsequent proliferation of aquatic and marginal plants along the water’s edge. Data collected for land-use, hydrology, channel and riparian changes, macroinvertebrates were synthesized using BEST (BIOENV and BVSTEP) multivariate statistics in PRIMER to search for high rank correlations between environmental and biological variables. When the environmental variables were tested against the biological variables showed that changes in macroinvertebrates were strongly related to area of plantations, area of undeveloped land, the extent of braiding, maximum 5-day average discharge in the wet season and the daily average volume in the dry season. Environmental variables were most influenced/driven by location (separated into sub-basins) while time was the driving factor for the macroinvertebrates data.

AFRIKAANSE OPSOMMING: Die aard van rivier-ekostelsels word deur die geskiedenis van hul aktiwiteite in hul komme beïnvloed. Hierdie studie was ʼn ondersoek na historiese veranderinge in die Bergrivierkom en die invloed daarvan op rivier-ekostelselstruktuur. Die sentrale aanname was dat alle aktiwiteite in die rivierkomlandskap direk of indirek tot die toestand (fisies, biologies, chemies) van riviere wat daardeur vloei, bydra. Eerstens moet bepaal word watter veranderinge in die rivierkom met verloop van tyd plaasgevind het, wat op verskillende maniere teen verskillende ruimteskale uitgevoer is. Veranderinge in grondgebruik is noukeurig vergelyk en oor die kom gekarteer, aangesien dit beskou is as verantwoordelik vir die riviervloeistelsel en rivierkanaalstruktuur. Voorspellings is gemaak oor hoe veranderinge in vloei en rivierkanaalhabitat die verspreiding en oorvloed van watermakro-invertebrate sou beïnvloed. ʼn Geskiedenis van grondgebruikveranderinge in die Bergrivierkom in vier tydperke, naamlik 1955–1965, 1976–1985, 1996–2005 en 2006–2015, is ondersoek. Die meerderheid droëlandgewasproduksie het in die laer voetheuwels en laaglande plaasgevind, terwyl die boonste voetheuwels uit boorde, wingerde en boswêreld bestaan het. Van 1955 tot 2015 het die omvang van landbougrond in die kom met die helfte afgeneem, aangesien droëlandgewasse na boorde en wingerde verander is en groot landstreke braak gelaat is. In dieselfde tydperk het die bosgebied met 73% afgeneem en stedelike gebiede het in grootte verdubbel, en so ook die aantal plaasdamme in reaksie op die verhoogde vraag na besproeiing om aan die waterhonger gewasse te verskaf. Die gevolge van die veranderinge in grondgebruik, die toename in plaasdamme en die bou van groot damme in die rivier se vloeistelsel is hierna ondersoek. Veranderinge in vloei is by vier riviermeters teen die lengte van die Bergrivier hoër op en laer af van die twee hoofdamme in die kanaal, die Bergrivierdam in die boonste voetheuwels en Misverstanddam in die laaglande, ondersoek. In die algemeen was die veranderinge meer opvallend by die stroomafmeters en die neigings was na verhoogde droëseisoenvloei en effens verlaagde natseisoenvloei weens die vrystelling van water vanaf en opvangs van strome deur die damme in die kanaal om in besproeiingsbehoeftes in die droë seisoen te voorsien. Die vloeipatroon van vroeë rekords is noukeuriger met reënval as in die vorige rekord vergelyk, en het getoon dat vloeiveranderinge waarskynlik aan antropogeniese gevolge soos grondgebruik en waterhulpbronontwikkeling toegeskryf kan word. Daar is voorspel dat sowel grondgebruik as waterhulpbronontwikkeling gevolge vir rivierkanaalvorm en habitat inhou, wat vervolgens ondersoek is. Veranderinge in rivierkanaalvorm en die omvang en samestelling van die vloedvlakte en oewergebied is van lugfoto’s by vyf terreine langs die Bergrivier en vyf aangrensende takriviere gekarteer. Al die terreine het verskillend gereageer, wat nie onverwags was nie, en verlaging van die omvang van die kanaal en oewergebied was groter langs die Bergrivier-hoofrivier in vergelyking met die takriviere. Geen vloedvlaktes is langs die takriviere teen die gemete skaal waargeneem nie, alhoewel ʼn afname in omvang met verloop van tyd langs die hoofrivier af waargeneem is, maar wel nie laer af van die Bergrivierdam nie, waar die vloedvlakte-oppervlakte toegeneem het weens die vorige vlegkanaal van 1938, wat in ʼn enkeldraadkanaal met vloedvlakte en ʼn groter oppervlakte sandbanke verander het. Die voorspelling is gemaak dat veranderinge in rivierhabitat, soos hierdie, verandering in die oorvloed en gemeenskapstruktuur van watermakro-invertebrate teweeg sou bring, wat vervolgens ondersoek is. Die oorvloed watermakro-invertebrate vanaf die 1950’s tot 2015 is bestudeer en het ʼn afname in simuliede en baetiede getoon, met ʼn toename in die oorvloed chironomiede, wat op ʼn afname in watergehalte dui. Veranderinge in ander groepe dui op ʼn afname in habitatgehalte, byvoorbeeld ʼn verlies aan plekopterane wat skoon gruisbeddings verkies, wat met kaeniede vervang is, wat ʼn sandkanaalbodem verkies. In 2015 was daar ook meer groepe invertebrate wat met stadig vloeiende gebiede en randplantegroei geassosieer word, en die aanname is gemaak dat dit plaasgevind het in reaksie op die uitwissing van houtagtige uitheemse bome aan die rivieroewers en die gevolglike voortplanting van water- en randplante teen die waterrand. Data wat vir grondgebruik, hidrologie, kanaal- en oewerveranderinge en makro-invertebrate ingesamel is, is met behulp van BEST (BIOENV and BVSTEP) meerveranderlike statistieke in PRIMER gesintetiseer om hooggeklassifiseerde verbande tussen omgewings- en biologiese veranderlikes te vind. Met toetsing van die omgewingsveranderlikes teen die biologiese veranderlikes, is gevind dat veranderinge in makro-invertebrate sterk verband hou met die oppervlakte van plantasies, die oppervlakte van onontwikkelde grond, die omvang van omvlegting, maksimum vyfdag- gemiddelde afloop in die nat seisoen en die daaglikse gemiddelde volume in die droë seisoen. Omgewingsveranderlikes is die meeste deur ligging (in subkomme geskei) beïnvloed/aangedryf, terwyl tyd die dryffaktor vir die data oor makro-invertebrate was.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/104990
This item appears in the following collections: