Fertilisation of semi-mature Pinus elliottii and Pinus elliottii x caribaea stands on a climatic gradient in the Tsitsikamma and its effect on system nutrition and stand productivity

Gerhardus Petrus, Scheepers (2018-12)

Thesis (PhD)--Stellenbosch University, 2018.

Thesis

ENGLISH ABSTRACT: Eight field trials were established in October 2015 to test the effect of different nitrogen and phosphorous combinations on Pinus elliottii and P. elliottii x caribaea growth. Fertiliser is costly, and the financial return is governed by the degree of the response, application costs and product worth. Commercial pine plantations are commonly fertilised at two stages – at establishment, and at mid-rotation following a final stem reduction. National and international studies are uncertain regarding the economic feasibility and response of semi-mature pine plantations to fertilisation. These uncertainties can be attributed to the substantial edaphic, topographic and climatic variations under which commercial plantations and forests are grown. The formulation of site-specific fertilisation rates could allow commercial and private forest companies to increase the profitability of fertilisation and achieve optimal growth responses. The field studies were established in the Tsitsikamma in the Eastern Cape, across a water-availability gradient and on sandy loam soils with soil pH (KCl) values ranging from 2.9 to 4.0. The field trials were designed to test the interaction of six fertiliser treatment combinations (all in kg ha-1) and consisted of a control treatment of 0 N and 0 P (T0), and then treatments of 0 N and 50 P (T1), 0 N and 100 P (T2), 100 N and 50 P (T3) and 100 N and 100 P (T4), and a maximum application rate of 200 N and 100 P (T5). These application rates were based on the findings of previous softwood fertilisation projects in the Southern Cape, as well as in the Boland region of the Western Cape. Data collection was done at 0, 6, 12, 18 and 24 months after fertilisation. The primary objective of this study was to formulate site-specific fertilisation rates for the semi-mature pine plantations of the Cape forest region. To achieve this, the study was divided into four sub-studies. The first sub-study focused on the effect of water availability on stand growth. It investigated whether the water deficit estimate by Thornthwaite (1948) and Thornthwaite and Mather (1955) could be used as a reliable estimate of soil water availability, relative to other widely accepted (lesser and increasingly complex) estimates of water availability generally used in forestry and other agricultural practices. The second sub-study made use of the Soil Nitrogen Availability Predictor (SNAP) model to predict the N mineralisation rates of all field trials. The predicted N mineralisation rates were then used to determine whether the N mineralisation potential of a soil significantly affects the fertilisation response after 24 months. The third sub-study was a development of the second: the mineralisable N and P, from simple to increasingly complex, was determined in controlled laboratory conditions by means of aerobically and anaerobically incubating soil samples acquired from each field study. The relationships between (a) calculated mineralisation rates, together with the basal daily N mineralisation rate predicted by the SNAP model and (b) the growth response at 24 months after fertilisation were evaluated. The final sub-study investigated whether canopy N and P contents were significantly affected by fertilisation and whether the application of different N and P fertiliser combinations can address the potential nutrient deficiencies of semi-mature slash plantations in the Tsitsikamma. Findings: This study shows that Thornthwaite’s soil water deficit methodology is an improved and increasingly accurate estimate of plant-available water relative to mean annual precipitation, and (FAO) estimates of aridity index a moisture growing season. The methodology has fewer and more easily obtainable data requirements and paints an accurate picture of soil water availabilities at times of seasonal fluctuations and inconsistent climatic conditions. The water deficit estimate has the potential to identify sites where growth is limited by soil water availability to larger and/or smaller degrees. Soils from slash pine plantations in the Tsitsikamma region have superior water-retention capabilities relative to sites from the Knysna and Boland regions of the Western Cape. No significant correlations were observed between the predicted annual N mineralisation rate of the SNAP model and the growth responses at 24 months after fertilisation; however, the model predicted the highest annual rates for the least responsive field trial: a N mineralisation rate in the range of 149 (after subtracting the 𝑁𝐻4+ before incubation from the final 𝑁𝐻4+ pool) to 238 kg N ha-1 yr.-1 (final pool only) was predicted for field trial A, and this field trial exhibited the poorest growth response to added fertiliser over the experimental period. In addition, field trial A had a higher soil pH, and a significant interaction was observed between the N mineralisation potentials of each site and the soil pH (p = 0.040). Significant Pearson correlations were observed between the total N, anaerobic N and aerobic P estimates and the growth response at 24 months after fertilisation. Field trials with higher total N contents were less responsive to increased N applications in the presence of P. The opposite was observed for increased P applications: sites with higher total N contents were increasingly responsive to higher applications of P. Sites with inherently higher anaerobic N mineralisation rates exhibited weaker growth responses to increased N application in the presence of P. Sites with higher P mineralisation rates were less responsive to P fertilisation. The anaerobically incubated N mineralisation rates were superior to the aerobic rates in this study, although the predicted basal N mineralisation rate of the SNAP model (which relies on an estimate of aerobic N) correlated with the volume responses at 24 months after fertilisation. Canopy N and P contents differed significantly between sites (p < 0.001). The N contents were significantly affected by fertilisation, at a confidence level of 90% (p = 0.059), and the N contents increased according to the highest fertilisation rates (T0 to T5). Significant treatment differences were observed for the canopy P content (p = 0.014) after 24 months, with the highest P content observed for the highest application of phosphorus (T2), in the absence of a N source. Plant nutrient availability appeared to be primarily driven by site-specific edaphic and topographic conditions and, to some extent, by the higher N and P fertiliser combinations. The field trials were established on highly acidic soils. This finding, together with the documented volume responses, suggests that stand growth in the Tsitsikamma can be improved with moderate N and P fertiliser application rates. It also calls for further testing of micronutrient and lime additions, as low pH conditions and sub-optimal foliar micronutrient levels were associated with lower stand growth responses.

AFRIKAANSE OPSOMMING: Agt veldproewe is in Oktober 2015 uitgelê met die doel om die uitwerking van verskillende stikstof- en fosfaatkombinasies op die groeitempo van Pinus elliottii en P. elliottii x caribaea plantasies te bestudeer. Kunsmistoedienings is ’n noemenswaardige uitgawe van boskultuur en die winsgewendheid berus op die groeireaksie, toedieningskostes en die kunsmismengsel. Kommersiële denneplantasies word tipies op twee groeifases bemes: tydens vestiging en kort na ’n finale uitdunning op middeljarige ouderdom. Die winsgewendheid en uitwerking van kunsmis op die groeitempo van middeljarige denneplantasies varieer en die bevindings van nasionale en internasionale studies verskil beduidend. Die beduidende verskille kan toegeskryf word aan die reeks edafiese, topografiese en klimaatstoestande waarop plantasies gevestig word. Die formulering van groeiplek-spesifieke bemestingsvlakke stel kommersiële en private maatskappye in staat om kostes te sny en optimale groeireaksies te behaal. Die veldproewe is geleë in die Tsitsikamma in die Oos-Kaap, oor ’n reënvalgradiënt en op sanderige leemgronde met pH (KCl) waardes van ongeveer 2.9 tot 4.0. Die proewe het elk ses kunsmisvlakke getoets (in kg ha-1): 0 N en 0 P (T0), 0 N en 50 P (T1), 0 N en 100 P (T2), 100 N en 50 P (T3), 100 N en 100 P (T4), en ’n finale behandeling van 200 N en 100 P (T5). Die kunsmismengsels en vlakke van bemesting is gebaseer op vorige bemestingseksperimente in Suid- en Wes- Kaapland. Opmetings om die groeireaksie te bepaal, is op 0, 6, 12, 18 en 24 maande ná behandeling geneem. Die primêre doelwit van die studie was om groeiplek-spesifieke kunsmistoedienings te formuleer vir denne-opstande in die Kaapse plantasiegebiede. Die navorsing is opgedeel in vier subprojekte. Die eerste projek het ondersoek of grondwatertekort (bereken deur die Thornthwaite (1948; 1953) metodiek) ’n beter skatting is van grondwaterbeskikbaarheid relatief tot die algemene (en toenemend komplekse) skattings wat in die bosbou- en landboubedrywe gebruik word. Tweedens is die SNAP model (Soil Nitrogen Availability Predictor) gebruik om die N-mineralisasievermoë van elke veldproef te bereken. Dié waarde is gebruik om vas te stel of die N-mineralisasievermoë die groeireaksie 24 maande ná behandeling affekteer. Die derde subprojek was ’n uitvloeisel van die tweede – N- en P-mineralisasietempo’s is in gekontroleerde laboratoriumtoestande bereken deur inkubasies van grondmonsters vanaf elke proef. Die bogenoemde mineralisasietempo’s, insluitend die basale mineralisasietempo van die SNAP-model, is met behulp van regressiewe tegnieke vergelyk met die groeireaksie van elke proef tot en met 24 maande ná behandeling. Die finale subprojek het ondersoek of N- en P-inhoud in die kroondak beduidend beïnvloed word kunsmistoedienings in die Tsitsikamma en terselfdertyd of die toedienings enige voedingstofgebrekke oor die studietypderk kon ophef. Bevindings: Die studie het vasgestel dat die Thornthwaite-grondwater-tekortmetodiek ’n verbeterde en akkurate skatting is van plant beskikbare waterinhoud relatief tot die gemiddelde reënval, asook die (FAO) indekse van ariditeit en vog-groeiseisoen. Die metodiek het minder datavereistes en skets ’n verbeterde algehele beeld van grondwaterbeskikbaarheid tydens seisoenale en uiterste klimaatsveranderinge. Die metodiek het die potensiaal om groeiplekke te identifiseer waar hoër en laer vlakke van grondwater beskikbaar is en boomgroei kan beïnvloed. Die gronde van die Tsitsikamma het hoër vlakke van beskikbare grondwater relatief tot verskeie plantasiegebiede in Suid- en Wes-Kaapland. Die groeireaksies 24 maande ná bemesting is nie beduidend beïnvloed deur die jaarlikse N-mineralisasievermoë wat deur die SNAP-model voorspel is nie, alhoewel die model ’n jaarlikse mineralisasie van 149 en 238 kg N ha-1 jr-1 (aanvanklike 𝑁𝐻4+-konsentrasie uitgesluit en ingesluit) voorspel het in proef A en dié proef ook die swakste groeireaksie tot gevoegde bemesting oor die eksperimentele tydperk getoon het. Daarbenewens het proef A die hoogste pH gehad en ’n beduidende interaksie is waargeneem tussen die N-mineralisasievermoë en die grond-pH (p = 0.040). Die bevinding het bewys dat die N-mineralisasievermoë van ’n grond ’n beduidende faktor is wat ingereken kan word met die beplanning van jaarlikse bemestingsveldtogte. Beduidende Pearson-korrelasies is waargeneem tussen die totale N-inhoud, anaerobiese N en aërobiese P-mineralisasietempo’s, en die groeireaksie 24 maande ná bemesting. Proewe met ’n hoër N-inhoud het laer groeireaksies op N-toedienings getoon in die teenwoordigheid van P. Daarbenewens is die teenoorgestelde waargeneem vir toenemende P-toedienings: proewe met hoër totale N-inhoud het beter groeireaksies getoon met toenemende P-toedienings. Proewe met hoër anaerobiese N-mineralisasietempo’s het swakker groeireaksies op verhoogde N- toedienings in die teenwoordigheid van P getoon. In dié studie was die anaërobiese skattings van N beter as die aërobiese skattings, alhoewel die voorspelde basale N-mineralisasietempo van die SNAP-model (wat van die aërobiese N-skatting gebruik maak) ook beduidend gekorreleer het met die groeireaksies op bemesting. Die kroondak N-inhoud is matig beduidend beïnvloed deur die kunsmisbehandelings (p = 0.059). Die N-inhoud het beduidend toegeneem volgens die toenemende hoeveelhede kunsmis wat in die studie gebruik is (T0 tot T5). Beduidende verskille in kroondak P-inhoud is ná 24 maande (p = 0.014) tussen die behandelings waargeneem. Die hoogste kroondak P-inhoud is waargeneem vir die hoogste P-behandeling (T2) in die afwesigheid van N. Dit het voorgekom asof blaarvoedingstatus gedryf word deur groeiplek-spesifieke edafiese en topografiese toestande en in ’n mate deur die hoër N- en P-kunsmisbehandelings. Hierdie bevinding, saam met die groeireaksies wat in al die proewe waargeneem is, dui daarop dat verbeterde groeireaksies in die Tsitsikamma behaal kan word met matige N- en P-toedienings. Dit is noodsaaklik om verdere eksperimente uit te voer op bekalking en spoorelemente in die streek, omdat suur grond en sub-optimale vlakke van spoorelemente in die naalde geassosieer is met die swakste opstandgroei in die reeks proewe.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/104850
This item appears in the following collections: