There is growing demand for pollination services in agricultural production, which contrasts with declines of wild and managed pollinator populations. Macadamia (Macadamia integrifolia) is a mass-flowering crop that depends on pollination services and is increasingly cultivated in South Africa. We studied the crop’s pollination in South African orchards considering variation in landscape context and the spatial arrangement of managed honeybees (Apis mellifera). We conducted pollination experiments and pollinator observations on macadamia trees along a distance gradient from orchard edges that bordered either near-natural or human-modified habitats. In addition, we mapped position and density of honeybee apiaries at orchard-level. Nut set of macadamia trees strongly relied on animal-mediated pollination: pollinator exclusion reduced the initial nut set (3 weeks after pollination) by 80% and the final nut set (15 weeks after pollination) by 54%. Supplemental hand-pollination of otherwise untreated flowers increased initial and final nut set by 66% and 44%, respectively, indicating substantial pollination limitation. The landscape context only weakly affected pollinator visitation to macadamia trees, with reduced visitation closer to orchard edges bordering human-modified habitats. Furthermore, we observed almost no wild pollinator species. Instead, honeybees constituted 99% of all visits, whereby honeybee visitation rates increased with a tree’s connectivity to apiaries. However, neither initial nor final nut was related to visitation rates, and the final nut set was actually reduced where honeybee colony density was high, with a predicted 50% reduction in final nut set between the lowest and highest colony densities. Our study demonstrates a strong pollination limitation in South African macadamia orchards, where managed honeybees fail at delivering the increasing need for pollination services. Indeed, increasing their colony densities may further limit their pollination efficiency. A pollination management that also includes non-Apis managed pollinators and wild pollinators is possibly needed to increase nut set and provide solutions for increasing pollination service demands. In intensive macadamia orchards, this can also necessitate the need for more pollinator-friendly management practices, including habitat restoration and reduced pesticide application.

Thesis (PhD)--Stellenbosch University, 2018. | ENGLISH ABSTRACT: Protea is a key component in the Fynbos Biome of the globally recognised Cape Floristic Region biodiversity hotspot, not only because of its own diversity, but also for its role in the maintenance of numerous other organisms such as birds, insects, fungi and mites. Protea is also internationally widely cultivated for its very showy inflorescences and, therefore, has great monetary value. Some of the organisms associated with these plants are destructive, leading to reduced horticultural and floricultural value. However, they are also involved in intricate associations with Protea species in natural ecosystems, which we still understand very poorly. Mites, for example, have an international reputation to negatively impact crops, but some taxa may be good indicators of sound management practices within cultivated systems. Their role in natural systems is even less well-understood. In this dissertation I explore the role of mites within Protea populations in both natural and cultivated systems, focussing on assemblages from inflorescences, infructescences and soil. Protea inflorescences and infructescences provide a niche for a unique assemblage of mites that have associations with a group of arthropod-associated fungi, the ophiostomatoid fungi. The mites feed on the fungi and carry their spores to new inflorescences as phoretic partners of Protea-pollinating beetles. As it was shown that some of the fungi have a panmictic population genetic structure over as much as 1000 km, it was assumed that organisms other than beetles must be responsible for this extremely long-range dispersal. Here I present the first concrete evidence of the ability of birds to vector spore-carrying mites to new Protea trees. I also provide evidence for a newly discovered mite-fungus mutualism within ornithophilous Protea neriifolia inflorescences between a Glycyphagus sp. mite and various species within the ophiostomatoid genus Sporothrix. New mite-mite commensalisms between the Proctolaelaps vandenbergi flower mite and the Glycyphagus sp. mite was also discovered and documented. In this intriguing system the Glycyphagus sp. mites have a mutualistic association with species in the fungal genus Sporothrix. These small mites are phoretic on the larger P. vandenbergi mites that, in turn, are phoretic on Protea pollinating birds, explaining genetic evidence for the long distance dispersal of the fungi. It is well-known that flower-associated mites such as Proctolaelaps kirmsei are nectar and pollen thieves of hummingbird pollinated plants in America. These mites reduce nectar and pollen rewards for pollinators, which influences pollinator visitation patterns and decreases available pollen for dispersal, thereby negatively influencing seed-set and plant population dynamics. This phenomenon has, however, not been investigated in similar systems in other parts of the world. I, therefore, set out to determine the possible role of P. vandenbergi flower mites, the most abundant flower mite within Protea inflorescences, as pollen and nectar thieves and as secondary pollinators of P. neriifolia. I provide the first evidence that P. vandenbergi feeds on nectar and pollen and that its reproduction is strongly linked to pollen availability. Nectar consumption rates of P. vandenbergi likely have little effect on total nectar availability for pollinators, but they can significantly reduce available pollen in inflorescences and may ultimately negatively influence seed set. This is exacerbated by the fact that I could show that they do not contribute to Protea pollination. There is rising global concern about the negative impact of land transformation on natural ecosystems. With the increase in land transformation for agriculture, natural flora is replaced by intensively managed exotic crops. This has devastating effects on biodiversity and ecosystem services. Ecologically more friendly management systems are thus urgently required. One proposed such system is the production of native plants as crops, as these can provide known niche space for native organisms including beneficial ones, which may reduce required management inputs. Protea is of high ecological significance and economic value as it is harvested for export within both natural and cultivated systems in South Africa. Although mites associated with these plants can be beneficial, they are usually regarded as pests and/or organisms that pose significant phytosanitary risks. I, therefore, investigated the impact of Protea repens cultivation on the mite assemblages associated with inflorescences, infructescences (the crop products where the presence of mites pose agricultural risks) and the rhizosphere (where most of the agriculturally beneficial mite species would reside). I show that this indigenous crop may well be able to maintain a large native mite biodiversity component in all three of these niches. However, essential environmental services such as the maintenance of sound soil ecology may be hindered even with very low management intensity. Results also indicated that current intensive pest management strategies do not effectively control mites associated with inflorescences. Continued improvement of post-harvest pest management practices, as difficult as these are for sensitive and fresh produce, are urgently needed. Less reliance on intensive management systems during the production phases of Protea inflorescences would also help preserve some natural ecological processes, such as the ones discovered and described in this dissertation. | AFRIKAANSE OPSOMMING: Protea is ʼn sleutelkomponent in die Fynbos Bioom van die wêreldwyd erkende Kaapse Floristiese Streek biodiversiteit sentrum, nie net as gevolg van die genus se eie diversiteit nie, maar ook vir sy rol in die behoud van verskeie organismes soos voëls, insekte, fungi en myte. Protea word ook internasionaal wyd gekweek vir hul baie aanskoulike bloeiwyses en is daarom van groot monetêre belang. Sommige van die organismes wat met hierdie plante geassosieer word, is destruktief, wat lei tot verminderde hortologiese en snyblom waarde. Hulle is egter ook betrokke in komplekse assosiasies met Protea spesies in hul natuurlike ekosisteme, wat ons steeds baie swak verstaan. Myte, byvoorbeeld, het ʼn internasionale reputasie daarvoor dat hulle gewasse negatief beïnvloed, maar sommige taksa mag goeie aanduiders wees van gesonde bestuurspraktyke binne gekultiveerde sisteme. Hulle rol in natuurlike sisteme word nog swakker verstaan. In hierdie dissertasie verken ek die rol van myte binne Protea populasies in beide natuurlike en gekultiveerde sisteme, en fokus op groeperings vanuit bloeiwyses, saadkeëls en die grond. Protea bloeiwyses en saadkeëls bied ʼn nis vir ʼn unieke versameling myte wat assosiasies het met ʼn groep fungi wat weer met geleedpotiges geassosieer word, naamlik die ophiostomatoide fungi. Die myte voed op die fungi en dra hul spore na nuwe bloeiwyses as foretiese maats van Protea-bestuiwende kewers. Aangesien dit getoon is dat sommige fungi ʼn panmiktiese populasie genetiese struktuur oor meer as 1000 km het, is dit aangeneem dat ander organismes as kewers verantwoordelik moes wees vir hierdie geweldige langafstand verspreiding. Hier bied ek die eerste konkrete bewyse van die vermoë van voëls om as vektore van spoordraende myte na nuwe Protea bome op te tree. Ek verskaf ook bewyse vir ʼn nuut ontdekte myt-fungus mutualisme binne voëlbestuifde Protea neriifolia bloeiwyses tussen ‘n Glycyphagus sp. myt en verskeie Sporothrix spp. fungi. Nuwe myt-myt kommensialismes tussen die Proctolaelaps vandenbergi blommyte en die Glycyphagus sp. myte is ook ontdek en gedokumenteer. In hierdie interessante sisteem het die Glycyphagus sp. myte ʼn mutualistiese assosiasie met die Sporothrix spp. fungi. Hierdie klein myte is foreties op die groter P. vandenbergi myte wat op hulle beurt weer foreties is op Protea-bestuiwende voëls, wat die genetiese bewyse van langafstand vervoer van die fungi verduidelik. Dis is goed bekend dat blomgeassosieerde myte soos Proctolaelaps kirmsei nektar en stuifmeel diewe van kolibrie bestuifde plante in Amerika is. Hierdie myte verminder nektar en stuifmeel belonings vir bestuiwers, wat bestuiwer besoekpatrone beïnvloed en die hoeveelheid beskikbare stuifmeel en nektar verminder. Dit beïnvloed saad-vorming en plant populasiedinamika negatief. Hierdie fenomeen is egter nog nooit in eenderse sisteme in ander dele van die wêreld ondersoek nie. Ek het daarom ten doel gehad om die moontlike rol van P. vandenbergi blommyte, die mees volop blommyt binne Protea bloeiwyses, as stuifmeel en nektardiewe en as sekondêre bestuiwers van P. neriifolia te ondersoek. Ek verskaf die eerste bewyse dat P. vandenbergi op nektar en stuifmeel voed en dat sy reproduksie sterk gekoppel is aan die beskikbaarheid van stuifmeel. Tempo van nektarinname het waarskynlik min effek op die totale beskikbaarheid vir bestuiwers, maar hulle kan die hoeveelheid beskikbare stuifmeel in die bloeiwyse beduidend verminder, en mag so uiteindelik saadvorming negatief beïnvloed. Dit word vererger deur die feit dat ek kon wys dat hulle nie bydra tot Protea bestuiwing nie. Daar is toenemende globale kommer oor die negatiewe impak van landtransformasie op natuurlike ekosisteme. Met die toename in landtransformasie vir landbou, word natuurlike flora verplaas deur intensief beheerde uitheemse gewasse. Dit het verwoestende effekte op biodiversiteit en ekosisteem dienste. Ekologies vriendeliker bestuursisteme word dus dringend benodig. Een voorgestelde sodanige sisteem is die produksie van natuurlike plante as gewasse, aangesien hulle natuurlike nisspasies vir inheemse organinsmes, insluitend voordeliges, kan bied, wat die bestuursinsette wat benodig word mag verminder. Protea is van groot ekologiese en ekonomiese belang aangesien dit geoes word vir uitvoere in beide natuurlike en aangeplante sisteme in Suid Afrika. Alhoewel myte wat met hierdie plante geassosieer word voordelig kan wees, word hulle gewoonlik as peste en/of organisme wat ʼn fitosanitêre risiko dra beskou. Ek het daarom die impak van Protea repens aanplanting op myt samestellings wat met bloeiwyses, saadkeëls (die gewasprodukte waar die aanwesigheid van myte landboukundige gevare inhou) en die risosfeer (waar meeste van die landboukundig voordelige myte aangetref word) geassosieer word ondersoek. Ek wys dat hierdie inheemse gewas wel in staat mag wees om ʼn groot natuurlike myt biodiversiteit in al drie hierdie nisse te onderhou. Essensiële omgewingsdienste soos die voorsiening van gesonde grond-ekologie mag egter verhinder word deur self lae bestuursintensiteit. Resultate het ook aangetoon dat die huidige intensiewe pesbestrydings strategieë nie myte wat met bloeiwyses geassosieer word doeltreffend bestry nie. Volgehoue verbetering van na-oes pesbeheer praktyke word dringend benodig, ongeag hoe moeilik hulle toepassing is vir sensitiewe en vars produkte. ʼn Verminderde afhanklikheid van hierdie intensiewe bestuursisteme tydens die produksie fases van Protea bloeiwyses sal ook help om sommige natuurlike ekosisteem prosesse te bewaar, soos dié wat in hierdie dissertasie ontdek en beskryf is. | Doctoral