The effect of residue management on the nutrient cycle in the production of rooibos (Aspalathus linearis) at Nieuwoudtville, Northern Cape

Nieuwoudt, Stephanus Francois (2017-03)

Thesis (MScAgric)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Rooibos (Aspalathus linearis) is a sensitive fynbos species with a large genetic variation, adapted to acid, nutrient poor soils, and can only be grown in certain parts of the Western- and Northern-Cape. Rooibos yields are getting poorer with increasing age of the field and the lifespan of rooibos are also just a maximum of 5 years commercially produced. A lot of medicinal/health research has been done and published on rooibos, but not much on the agricultural production aspects of rooibos. In order to increase the production and lifespan of rooibos tea further research needs to be done to better understand the rooibos plant in its cultivated environment. The role of leaf litter in fynbos, particularly in the cultivated rooibos nutrient cycle is still a grey area that could open up key management principles regarding plant water availability and plant nutrition style. The hypothesis is that the method of harvesting the rooibos seed by removing the litter layer without returning it under the plant can have a negative impact on the nutrient pools and cycle and thus lead to a shorter lifespan. The main aim was thus to look at the effect of different residue treatments on the rooibos plant nutrient cycle (uptake and nutrient pools). Four sites across the Nieuwoudtville Bokkeveld region were selected with all the rooibos plants being ± 2 years old. Four different mulch treatments; a bare soil (leaf residue removed) treatment imitating seed harvesting (A), an added rooibos mulch (B), a natural leaf mulch (C) and an enriched rooibos mulch (D) were prepared at 4 sites. The chemical properties of soil and plant tissue from rooibos plants were tested. 5TE soil probes were used to measure the volumetric water content, EC and temperature at two soil depths of each treatment. All measurements were also duplicated using near-infrared spectroscopy (NIRS), to generate a database for future reference and to build calibrations that will be able to predict the nutrient content in the soils and plants. It was found that soil chemical properties including P (mg kg-1), Na (cmolc kg-1), K (cmolc kg-1), Ca (cmolc kg-1), Mg (cmolc kg-1), Zn (mg kg-1), Mn (mg kg-1), C (%) and % Na (at pH 7); and plant chemical properties including Na (%) and plant N (%), P (%), K (%), Al (mg kg-1) and Fe (mg kg-1) all had a significant effect of the regrowth models using multiple regression analysis. Soil P, Mg and K had the biggest positive influences on the regrowth models. During this process it was found that the N:P ratio in soil plays an important role in the uptake of N and growth. Only at treatment D, with the lowest soil N:P ratio, plant N (%) had a positive influence on the regrowth multiple regression model. Plant N and P had a moderate positive correlation (R2=0.56). Nutrient uptake by the rooibos plant was very high from July 2015 to September 2015. These nutrients included N, P, K, Ca, Mg, Zn, Mn, Fe and Al. From September 2015 to January 2016 however the uptake was lower for all the nutrients, but for K and Mg the uptake was higher compared to the other nutrients. The decrease in plant nutrient concentration from September 2015 to January 2016 is a result of less nutrient uptake and nutrient dilution following rapid growth of plant. There was an increase in soil exchangeable Mg and Ca from July 2015 to September 2015 due to increase in soil pH during this time. Soil exchangeable Ca (R2=0.49) and Mg (R2=0.61) correlated positively with pH, thus the increase in soil exchangeable Ca and Mg can be due to the increase in pH. For all the treatments there was a total decrease in soil N (significant for A and B) and plant N over the one-year period. The plant Al and Zn for all the treatments also increased over the one-year period. The increase in plant Al was not significant and the increase in plant Zn was significant for all treatments. From July 2015 to January 2016 there were differences in growth between the treatments. Treatment A resulted in the lowest estimated dry matter increase during this period and for treatment B it was the highest. The difference in estimated dry matter increase between these two treatments was also statistically significant. The estimated dry matter increase for treatment C and D was higher than treatment A but it was not significant. The mulch treatments, especially treatment B, resulted in higher P, K and Mg uptake. For all the treatments, except treatment D, the soil P decreased over the one-year period. For treatment B and D the plant P increased significantly compared to treatment A and C where the increase was not significant. The mulch treatments showed an increase in plant K, but it was only significant for treatment B over the one-year period. For all the treatments there was an increase in plant Mg, but only for treatment A the increase was not significant. All the mulch treatments also conserved more water for longer compared to treatment A. The combination of nutrient leaching from the mulch (Mg and K) and the conservation of more soil water may be the reason for the higher nutrient uptake by treatment B and thus the better growth. From the NIRS study it was found that for these sandy soils it was only exchangeable Mg that could be predicted with good accuracy (RPD>2). The soil chemical properties pH, H (cmolc kg-1), K (mg kg-1), Ca (cmolc kg-1), Fe (mg kg-1) and C (%) models showed satisfactory predictability. For plant samples NIRS predicted P (%) and Mg (%) with good accuracy. The prediction models for N (%), K (%), Ca (%) and Na (mg kg-1) were only satisfactory and for the rest it was unreliable. From these results it was thus not possible to quantitatively predict all the chemical properties in the soil and plant samples but there is potential for better calibrations in the future. Differences in growth and vigour can also be attributed to location. The micro conditions in which a single plant grows, related to the impact of normal agricultural practices, was found to also determine the success of rooibos production. The hypothesis was supported by treatment C (the plant where natural mulch was not removed) having a higher estimated dry matter increase compared to treatment A (bare soil), but this difference was not significant. Treatment B (added mulch) however showed to have a significant impact. Rooibos production systems are far from being optimized and the amounts of unknown impacts were narrowed down by this work. This research indicates that the rooibos plant is genetically unrefined and that agricultural practices should focus more towards the natural state of fynbos nutrient availability and growth.

AFRIKAANSE OPSOMMING: Rooibos (Aspalathus linearis) is ʼn sensitiewe fynbos spesie met groot genetiese variasie, aangepas vir suur, voedingstof arme gronde en kan net verbou word in sekere gedeeltes van die Wes- en Noord-Kaap. Rooibos opbrengste raak al hoe swakker met toenemende ouderdom van die veld en die lewensduur van rooibos is ook net ʼn maksimum van 5 jaar (kommersieel verbou). Baie medisinale en gesondheidsnavorsing is al gedoen en gepubliseer, maar min op die landbouproduksie aspek van rooibos. Om die produksie en lewensduur van rooibostee te verhoog en te verleng word verdere navorsing benodig om die rooibos plant in sy kommersieel verboude omgewing beter te verstaan. Die rol van ʼn blaar deklaag in fynbos, veral in die rooibos voedingstof siklus, is nog steeds ʼn onbekende area en kan lei tot baie belangrike bestuurs praktyke , met betrekking tot plant water beskikbaarheid en plantvoeding styl. Die hipotese is dat die manier waarop die rooibos saad geoes word deur die verwydering van die natuurlike blaar deklaag sonder om dit terug te sit onder die plant ʼn negatiewe impak op die voedingstof poele en siklus het en dus lei tot laer opbrengste en ʼn korter rooibos lewensduur. Die hoofdoel van die studie was dus om te kyk na die effek van verskillende rooibos blaar deklae op die rooibos plantvoedingstof siklus (opname en voedingstof poele). Vier eksperimentele areas (± 2 jaar oud rooibostee) oor die Nieuwoudtville Bokkeveld Plato streek is gekies waar die proewe uitgevoer sou word. Vier verskillende deklaag behandelings; ʼn kaal grond (A) (nabootsing van waar saad verwyder is), ʼn ekstra rooibos blaar deklaag (B), ʼn natuurlike blaar deklaag (C) en ʼn verrykte rooibos blaar deklaag (D) is voorberei by elkeen van die uitgesoekte areas. Die chemiese eienskappe (voedingstof status) van die grond en rooibos plantweefsel is ontleed. 5TE grond sensore is gebruik om die volumetriese waterinhoud, temperatuur en EC te meet op twee dieptes, een plant van elke behandeling. Al die grond en plant monster ontledings is ook gedupliseer met behulp van NIR spektroskopie om ʼn databasis vir toekomstige verwysing te genereer en om kalibrasies, wat ons in staat sal stel om die voedingstofinhoud in die grond en plante te kan voorspel. Deur gebruikmaking van veelvuldige regressie ontledings is dit bevind dat die grond chemie, [P (mg kg-1), Na (cmolc kg-1), K (cmolc kg-1), Ca (cmolc kg-1), Mg (cmolc kg-1), Zn (mg kg-1), Mn (mg kg-1), C (%) en % Na (by pH 7)] en plant chemiese eienskappe [Na (%) en plant N (%), P (%), K (%), Al (mg kg-1) en Fe (mg kg-1)] almal ʼn beduidende effek gehad het op die hergroei modelle. Deur hierdie proses van uitkenning t.o.v. die belangrikste voedingstowwe, is bevind dat die N:P verhouding in die grond ʼn belangrike rol speel met die opname van N en ook groei. Net by behandeling D, met die laagste N:P verhouding, het plant N (%) ʼn positiewe effek gehad in die hergroei model. Plant N en P het ʼn positiewe korrelasie getoon (R2=0.56). Voedingstof opname deur die rooibos plante was baie hoog vanaf Julie 2015 tot September 2015. Hierdie voedingstowwe sluit in N, P, K, Ca, Mg, Zn, Mn, Fe en Al. Van September 2015 na Januarie 2016 was dit laer vir al die voedingstowwe, maar vir K en Mg was die opname hoër in vergelyking met die ander voedingstowwe. Die afname in plant voedingstof konsentrasie vanaf September 2015 na Januarie 2016, is ʼn gevolg van minder voedingstof opname en verdunning a.g.v. die drastiese toename in groei. Daar was ʼn toename in grond uitruilbare Mg en Ca van Julie 2015 na September 2015 a.g.v. die toename in grond pH. Grond uitruilbare Ca (R2=0.49) en Mg (R2=0.61) het positief gekorreleer met grond pH, dus die gevolglike toename in uitruilbare Ca en Mg met toename in pH. Vir al die behandelings is daar 'n totale afname in die grond N bespeur (belangrik vir A en B) en plant N oor die tydperk van een jaar. Die plant Al en Zn vir al die behandelings het ook toegeneem oor die tydperk van een jaar. Die toename in die plant Al was nie betekenisvol nie, maar die toename in die plant Zn was betekenisvol vir alle behandelings. Vanaf Julie 2015 tot Januarie 2016 was daar verskille in groei tussen die behandelings. Behandeling A het die laagste beraamde toename in droëmassa getoon gedurende hierdie tydperk, en vir behandeling B was dit die hoogste. Die verskil in beraamde droëmassa toename tussen hierdie twee behandelings was statisties betekenisvol. Die beraamde droëmassa toename vir behandeling C en D was hoër as behandeling A, maar was nie statisties betekenisvol nie. Die deklaag behandelings, veral behandeling B, het gelei tot hoër P, K en Mg opname. Vir al die behandelings, behalwe behandeling D, het die grond P afgeneem oor die tydperk van een jaar. Vir behandeling B en D het die plant P aansienlik toegeneem in vergelyking met behandeling A en C. Slegs die deklaag behandelings het 'n toename in die plant K getoon, maar net vir behandeling B oor die tydperk van een jaar was die toename betekenisvol. Vir al die behandelings was daar 'n toename in die plant Mg, maar slegs vir behandeling A was die toename nie betekenisvol nie. Al die deklaag behandelings het ook meer water vir langer bewaar in vergelyking met behandeling A. Die kombinasie van voedingstof loging uit die deklaag (Mg en K) en die bewaring van meer grondwater m.b.v. van deklae, kan die rede wees vir die hoër voedingstof opname veral deur behandeling B en dus beter groei. Vanuit die NIRS studie was bevind dat, vir die baie sanderige gronde, dit net uitruilbare Mg (cmolc kg-1) was wat voorspel kon word met hoë akkuraatheid (RPD>2). Die modelle vir die grond chemiese eieskappe [pH, H (cmolc kg-1), K (mg kg-1), Ca (cmolc kg-1), Fe (mg kg-1) en C (%)] het bevredigende voorspelbaarheid vertoon. By die plantmonsters was dit net die P (%) en Mg (%) wat met hoë akkuraatheid voorspel kon word. Die modelle vir N (%), K (%), Ca (%) en Na (mg kg-1) was net bevredigend en vir die res onbetroubaar. Van hierdie resultate was dit dus nie moontlik om al die chemiese eienskappe van die grond en plant monsters kwantitatief akkuraat te voorspel m.b.v. NIRS nie, maar daar is potensiaal met verdere navorsing vir 'n beter kalibrasies in die toekoms. Verskille in groeikrag kan ook toegeskryf word aan spesifieke ligging. Met die mikro omstandighede waarin ʼn enkele plant groei, wat verband hou met die impak van normale landboupraktyke, is bevind dat dit ook die sukses van rooibos produksie bepaal. Die hipotese is ondersteun deur behandeling C (die plant waar natuurlike deklaag nie verwyder) wat 'n hoër geraamde droëmateriaal toename in vergelyking met behandeling A (kaal grond) getoon het, maar die verskil was nie betekenisvol nie. Behandeling B (bygevoegde deklaag) het egter tot 'n betekenisvolle toename in groei getoon. Rooibos produksiestelsels is vêr van optimaal en die hoeveelheid van onbekende impakte is verklein deur hierdie werk. Hierdie navorsing dui daarop dat die rooibos plant geneties onverfynd is en dat landboupraktyke meer moet fokus op die natuurlike toestand van fynbos voedingstof beskikbaarheid en groei.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/101329
This item appears in the following collections: