Vegetation community development on a chronosequence of abandoned gold and rehabilitated nickel tailings around the Great Dyke region in Zimbabwe

Thesis (PhDConsEcol)--Stellenbosch University, 2020.

ENGLISH ABSTRACT: Ecosystem development on unique anthropogenic sites such as mine tailings is poorly understood and there is much controversy on whether the trajectory of vegetation development is towards restoration of pristine conditions or towards formation of novel plant ecosystems. While traditional theories assert that vegetation development is towards restoration of natural conditions which may be accelerated by active rehabilitation or reclamation, emerging theories propose a divergent trajectory. However, studies focusing on the trajectory of natural and assisted revegetation and their abiotic determinants on rehabilitated and unrehabilitated mine tailings are limited. The traditional and emerging theories therefore remain untested on such unique anthropogenic sites. This has led to adoption of expensive mine restoration programs that are not informed by scientific empirical evidence, often without success. In order to contribute to closure of this gap, the study was carried out primarily at Freda Rebecca gold mine and at Trojan nickel mine in Bindura district in the northern part of savanna Zimbabwe, and secondarily at several other ancillary mines situated around the great Dyke of Zimbabwe. The sites have mine tailings of different ages subjected to spontaneous revegetation, in the case of gold mine tailings while nickel tailings were rehabilitated by planting leguminous woody species and left for subsequent self-sustained revegetation. The main aim of the study was to establish the trajectory of vegetation structural and compositional attributes on a chronosequence of abandoned gold and nickel tailings with reference to adjacent natural vegetation matrices and to identify physicochemical substrate correlates. A secondary aim was to test the contribution of local site factors such as topography and climatic gradients and their interaction with substrate age on revegetation of the mine tailings. The study aimed to answer the following questions: (1) Do trajectories of vegetation succession on unrehabilitated gold and rehabilitated nickel tailings resemble convergence to pristine conditions or divergence towards formation of novel ecosystems?, (2) How do physicochemical properties of gold and nickel tailings vary over time and how do they relate to vegetation patterns?, (3) Are the occurrence patterns of vegetation affected by tailings age, type, local topographic factors and or climatic gradients?, and (4) What are the occurrence patterns of biological crusts (BCs) on gold tailings and what are their relationship with substrate and vegetation attributes? Unexpectedly, vegetation composition for unrehabilitated gold tailings appeared to gradually follow a convergence trajectory towards natural reference conditions while vegetation composition on rehabilitated nickel mine tailings moved away from reference conditions, likely forming novel ecosystems. However, vegetation structural attributes largely converged towards references for both rehabilitated and unrehabilitated mine tailings. Substrate physicochemical properties showed differential trajectories with some (total N and available P) increasing towards reference properties, some fluctuating (exchangeable bases, texture and EC), and some declining (e.g. pH, mineral N and bulk density) thereby yielding multiple contrasting end-points relative to reference sites. Overall, there was a significant effect of age and local site position or topography on total N, available P, OC, and EC and interactive effect of tailings age and position on EC. Areas covered by BCs had enhanced total N, OC, exchangeable bases and higher pH than areas without. Based on substrate physicochemical properties, the gold tailings were distinct from reference conditions suggesting more than 110 years may be needed to restore soil properties or restoration is not possible. On both rehabilitated and unrehabilitated mine tailings, vegetation attributes were mainly correlated to dynamics in total N, OC, available P, pH, EC and bulk density. Total BC cover significantly increased with tailings age, particularly lichens and mosses showing significant increases while cyanobacteria and algae significantly decreased. Ordination analysis showed a substrate age dependent correlation between BC cover and vegetation cover, species richness and densities, particularly, a positive relationship on 10-82 year-old-dumps and negative correlation thereafter. Moreover, vegetation composition and structure varied with BC cover gradients. Vegetation structure was significantly affected by local site topography while composition was not significantly affected. A gradient analysis of plant structural and compositional attributes on gold tailings along climatic gradients showed significant increase in vegetation structural attributes (cover, density, richness) with increase in rainfall, particularly for forbs and grass species except for woody species. Life cycle appeared to affect plant response to climatic gradients with annuals significantly decreasing with increase in rainfall, while perennial plants increased. However, plant species composition was generally similar on gold tailings across varying climatic gradients except for those with above 300 mm difference. Rehabilitation, appeared to break the influence of climate on vegetation composition as some unrehabilitated sites in contrasting climatic regions separated while those rehabilitated grouped showing high compositional similarity for rehabilitated tailings. When contextualized against existing revegetation theories, the results provide evidence that no single trajectory of vegetation composition development can be generalized for mine tailings. However, results confirm general convergence of structural attributes regardless of substrate type and management. Temporal changes in vegetation attributes are correlated to increases in fertility, pH dynamics and a decrease in substrate compaction and salinity. Results also suggest that mine tailings support specific plant species assemblages owing to their uniqueness and that may largely override the effect of climatic gradients. Evidence also suggest BCs and the planted woody species create microsites (nutrient hot spots) most likely due to their soil ecological engineering capabilities, significantly influencing vegetation patterns on mine tailings. Overall, when determining revegetation trajectories on anthropogenic sites, the following factors are important: (1) substrate age, (2) the starting point, reclamation versus no reclamation, (3) substrate type, (4) specific vegetation variables and growth forms, (5) topography, (5) the role of BCs, (7) and (8) to a lower extent climatic variables.

AFRIKAANSE OPSOMMING: Ekosisteemontwikkeling op unieke antropogeniese substrate, soos mynuitskot, word nie goed verstaan nie en daar is baie kontroversie oor of die trajek van plantegroei is na die herstel van ongerepte toestande of die vorming van nuwe (‘novel’) plantekosisteme. Terwyl tradisionele teorieë stel voor dat plantegroei ontwikkel na die natuurlike toestande is en deur aktiewe rehabilitasie of reklamasie versnel kan word, stel nuwe teorieë 'n alternatiewe trajek voor. Ondersoeke wat fokus op die trajek van natuurlike en ge-assisteerde hergroei en hul abiotiese determinante op gerehabiliteerde en ongerehabiliteerde mynuitskot is egter beperk. Die tradisionele en ontluikende teorieë bly dus ongetoets op sulke unieke antropogeniese substrate. Dit het gelei tot die aanvaarding van duur mynherstelprogramme wat nie deur wetenskaplike empiriese bewyse ingelig word nie, dikwels sonder sukses. Om hierdie leemte te vul, is die studie hoofsaaklik by die Freda Rebecca-goudmyn en by die Trojan-nikkelmyn in die Bindura-distrik in die noordelike deel van savanna Zimbabwe uitgevoer, en tweedens by verskeie ander myne wat rondom die groot Dyke van Zimbabwe geleë is. Die studieterreine het uitskotte van verskillende ouderdomme wat aan spontane plantegroei herstelling onderworpe is, in die geval van goudmynuitskot, terwyl nikkeluitskot gerehabiliteer is deur houtagtige peulplant-spesies te plant wat gelos is vir selfonderhoudende herbeplanning plantegroei herstel. Die hoofdoel van die studie was om die trajek van strukturele en komposisionele eienskappe van plantegroei vas te stel op 'n chronosekwensie van verlate goud- en nikkeluitskot-areas met as verwysing die aangrensende natuurlike plantegroei-matrikse, en om substraat fisiko-chemiese korrelate te identifiseer. 'n Sekondêre doel was om die bydrae van plaaslike terreinfaktore soos topografie en klimaatgradiënte en die interaksie daarvan met die substraatouderdom op plantegroei-ontwikkeling van die mynuitskot te toets. Die studie was daarop gerig om die volgende vrae te beantwoord: (1) Is die trajek van plantsuksessie op ongerehabiliteerde goud en gerehabiliteerde nikkeluitskot die van konvergensie na ongerepte toestande of divergensie na die vorming van nuwe (‘novel’) ekosisteme? (2) Hoe wissel fisiko-chemiese eienskappe met verloop van tyd en hoe hou hulle verband met plantegroeipatrone ? (3) Word die voorkomspatrone van plantegroei beïnvloed deur die uitskot-ouderdom, tipe, plaaslike topografiese faktore en / of klimaatgradiënte? en (4) Wat is die voorkomspatrone van biologiese korse? (BK's) op gouduitskot, en wat is die verband met die substraat- en planteienskappe? Onverwags blyk dit dat die plantesamestelling van ongerehabiliteerde gouduitskot geleidelik 'n trajek van konvergensie na natuurlike verwysingstoestande volg, terwyl plantesamestelling op gerehabiliteerde nikkeluitskot wegbeweeg van verwysingsomstandighede, en waarskynlik nuwe (‘novel’) ekosisteme vorm. Strukturele eienskappe van plantegroei konverteer grootliks egter na die verwysingsareas vir sowel gerehabiliteerde as ongerehabiliteerde mynuitskot. Substraat fisiko-chemiese eienskappe het differensiële trajekte getoon, met sommige (totale N en beskikbare P) wat toeneem na die verwysingsareas, sommige was wisselend (uitruilbare basisse, tekstuur en EC), en sommige wat daal (bv. pH, minerale N en bulkdigtheid), wat meervoudige kontrasterende eindpunte relatief tot die verwysingsareas opgelewer het. In die algemeen was daar 'n beduidende effek van ouderdom en plaaslike posisie of topografie op die totale N, beskikbare P, organiese koolstof en elektriese konduktiwiteit en die interaktiewe effek van die ouderdom en posisie op elektriese konduktiwiteit. Areas gedek deur BK's het 'n verbeterde totale N, organiese koolstof, uitruilbare basisse en hoër pH gehad as gebiede sonder. Gegrond op die fisiko-chemiese eienskappe van die substraat, is die gouduitskot onderskei van die verwysingareas, wat daarop dui dat meer as 110 jaar nodig mag wees om grondeienskappe te herstel, of dat herstel nie moontlik is nie. Op beide gerehabiliteerde en ongerehabiliteerde mynuitskot, is plantegroei-eienskappe hoofsaaklik gekorreleer met dinamika in totale N, organiese koolstof, beskikbare P, pH, EC en bulkdigtheid. Totale BK-dekking het aansienlik toegeneem met die ouderdom van die uitskot, veral ligene en mosse wat aansienlike toenames toon, terwyl cyanobakterieë en alge wat aansienlik afgeneem het. Ordeningsanalise het getoon dat 'n substraat-ouderdomsafhanklike korrelasie tussen BC-dekking en plantegroeibedekking, spesierykheid en –digtheid bestaan, met veral 'n positiewe verwantskap op 10-82 jaar oue uitskotareas en negatiewe korrelasie daarna. Verder het die samestelling en struktuur van plantegroei gewissel met BK-dekgradiënte. Die struktuur van die plantegroei is beduidend beïnvloed deur die plaaslike topografiese posisie, terwyl die samestelling nie beduidend beïnvloed is nie. 'n Gradiënt-analise van plantstruktuur- en samestellings-eienskappe op gouduitskot oor klimaatgradiënte het 'n beduidende toename in strukturele eienskappe van die plantegroei (bedekking, digtheid, rykheid) getoon met toename in reënval, veral vir ‘forb’ - en grasspesies, behalwe vir houtagtige spesies. Dit wil voorkom asof lewenssiklus die reaksie van plante op klimaatgradiënte beïnvloed, met eenjariges wat aansienlik daal met ‘n afname in reënval, terwyl meerjarige plante toegeneem het. Plantspesiesamestelling was egter oor die algemeen soortgelyk op gouduiskot oor verskillende klimaatsgradiënte, behalwe vir dié met 'n verskil van meer as 300 mm. Rehabilitasie het blykbaar die invloed van die klimaat op plantegroeisamestelling afgebreek, aangesien sommige ongerehabiliteerde terreine in kontrasterende klimaatstreke uitgeskei het, terwyl die gerehabiliteerde groepe 'n hoë komposisie-ooreenkoms vir gerehabiliteerde uitskot toon. As dit gekontekstualiseer word teen bestaande plantegroei-onwikkelingsteorieë, lewer hierdie resultate 'n bewys dat daar nie 'n enkele trajek van die ontwikkeling van plantegroeisamestelling veralgemeen kan word vir mynuitskot-areas nie. Resultate bevestig egter algemene konvergensie van strukturele eienskappe, ongeag die substraat-tipe en bestuur. Tydelike veranderinge in plantegroei-eienskappe is gekorreleer met toename in vrugbaarheid, pH-dinamika en 'n afname in substraatverdigting en soutgehalte. Resultate dui ook daarop dat mynuitskot-areas spesifieke plantspesiesamestellings het as gevolg van hul uniekheid en dat dit die effek van klimaatsgradiënte grootliks kan oorheers. Bewyse dui ook daarop dat BC's en die aangeplante houtagtige spesies mikroareas (voedingstof-‘hotspots’) skep, waarskynlik as gevolg van hul ekologiese ingenieursvermoë in die grond, wat die plantegroeipatrone op mynuitskotareas aansienlik beïnvloed. In die algemeen is die volgende faktore belangrik by die bepaling van plantegroei-onwikkelings trajekte op antropogeniese terreine: (1) substraatouderdom, (2) die beginpunt, herwinning teenoor geen herwinning, (3) substraattipe, (4) spesifieke plantegroei-veranderlikes en groeivorme, (5) topografie, (5) die rol van BK's, (7) en (8) in 'n kleiner mate klimaatveranderlikes.

Doctoral