Loading...
Thumbnail Image
Publication

Nectar variation in Aotearoa New Zealand tree species

Abstract
In this thesis, I investigated whether and how the nectar of New Zealand flowering trees varies within and among species and across regions. For this, I quantitatively assessed interspecific variation in floral (size, weight, shape, and colour) and nectar traits (volume, concentration, and chemical composition) across 50 woody species in New Zealand. Sampled species included, but were not limited to, those associated with honey production and those susceptible to Myrtle Rust, such as the Myrtaceae Leptospermum scoparium (mānuka), Kunzea ericoides (kānuka), Lophomyrtus spp. (rama rama), Metrosideros spp. (pōhutukawa, rāta), and Knightia excelsa (rewarewa). I also examined the impact of climate on nectar traits using a subset of eight common native tree species sampled from five coastal regions. Additionally, the study investigated the Removal-Enhanced Nectar Replenishment (RENR) response of Vitex lucens Kirk (pūriri, Lamiaceae), an endemic bird-pollinated tree, to frequent two-hourly nectar removal. Sampling comprised approximately 10,500 flowers from 436 trees across all six climate regions of New Zealand, covering both main islands (35-45° S / 170-177° E), and measuring the diameter at breast height (DBH) of these trees. Most flowers (10,000 flowers from 428 trees) were analysed for the interspecific study (Chapter 4), with a subset of these (4,276 flowers from 164 trees) used in the inter-regional climate study (Chapter 3), and 120 pūriri flowers from 8 trees examined in the nectar removal study (Chapter 2). For all studies, nectar was extracted using micropipettes; flower size was measured with digital callipers; flower and nectar weight were determined using a scale; flower colour by applying picture analysis software; nectar concentration was assessed with a refractometer; and nectar chemical composition was analysed using High-Performance Liquid Chromatography (HPLC) and Liquid Chromatography-Mass Spectrometry (LC-MS). Data were analysed using a comprehensive suite of statistical techniques, including linear regression, Spearman rank and Pearson’s correlations, and generalised additive models (GAMs) and generalised additive mixed models (GAMMs), to identify correlations, detect multicollinearity, and elucidate multivariate trends. Additionally, a range of tests, including t-tests, Wilcoxon rank-sum tests, Tukey's Honestly Significant Difference (HSD) tests, Dunn's tests (with Bonferroni correction), and Kruskal-Wallis tests, as well as Analysis of Variance (ANOVA), were employed to determine covariances and assess statistical significance among groups. Lastly, I tested traits for phylogenetic signals using Pagel’s lambda. Flower traits ranged from 3–879 mg in fresh weight and 2–67 mm in size, and they secreted 1–82 μL of nectar containing 0.01–54% solubles. I identified 62 distinct nectar components, comprising 25 sugars and 37 non-carbohydrate compounds. On average, nectar solubles consisted of 97% carbohydrates and 3% non-carbohydrates, including six alkaloids, sixteen amino acids, eleven phenolics, and four vitamers. Only 8% of the 50 tested species produced sucrose-rich nectar, defined as nectar in which sucrose comprised more than 50% of total solubles. Species exhibited unique nectar sugar profiles, containing between 4 and 25 sugar types. In the RENR study (Chapter 2), V. lucens flowers exhibited a neutral response to frequent nectar removal: total nectar volumes from flowers sampled five times at two-hour intervals were similar to those from control flowers sampled once after ten hours. Interestingly, the replenishment patterns of frequently sampled flowers followed a diurnal rhythm that correlated with changes in Vapour Pressure Deficit (VPD). The inter-regional study (Chapter 3) highlighted that nectar and floral trait variation among regions was significant for all but one species, underscoring the highly species-specific nature of climate-trait relationships. Climate factors each accounted for approximately 30-80% of the regional variation in plant traits for most species. The analysis of a substantial interspecific dataset (Chapter 4) revealed positive correlations between nectar volume, concentration, and alkaloid content with floral traits such as size, weight, shape (especially those more difficult to access, such as tube- and flag-shaped flowers), and colour (notably yellow, orange, or red). In contrast, levels of hexasaccharides were negatively correlated with these floral characteristics, with higher concentrations found in smaller, white, green, or purple flowers of shapes that are more accessible (e.g. dish-shaped). Hence, some nectar traits aligned with the species’ pollination syndrome. Moreover, significant phylogenetic signals were observed; for example, nectar from Fabid and Campanulid taxa had higher sucrose concentrations than other clades, whereas nectar from Myrtales and Lamiales was characterised by higher glucose content than other clades. I found that nectar variation is highly complex and species-specific, shaped by phylogenetic relationships, climate and water status. These findings offer new insight into the factors driving nectar composition across species. The results can support large-scale estimates of nectar availability, with applications in honey production and habitat conservation. In addition, the identified nectar profiles provide a foundation for comparison with honey chemistry to develop new authenticity markers for high-value honeys. For future studies, I propose conducting simultaneous nectar removal experiments across multiple species under controlled environmental conditions to assess whether phylogenetic relationships influence species-specific responses to nectar removal. Particular attention should be given to the type of nectary vascularisation and secretion mechanism, as well as to potential effects of flower colour and shape on nectar production concerning heat management. In field-based RENR studies, I recommend that nectar sampling be carried out by multiple researchers simultaneously, as replenishment is highly time-sensitive. Additionally, sampling should be paired with pollinator observations, and the analysis should incorporate information on the diurnal activity patterns of key pollinators. Similarly, studies on inter-regional differences in nectar composition should account for local pollinator communities and investigate additional potential drivers of variation, including the species' nectary type, vascularisation, secretion mechanism, plant water status, soil nutrient availability, and phylogenetic relationships. To build a detailed picture of regional nectar variation, I recommend using large sample sizes per species per region, with a focus on closely related species known for their high nectar production. Further research into interspecific nectar variation should explore whether the observed phylogenetic patterns persist across broader taxonomic and ecological contexts, and whether species-specific nectar profiles are consistently shaped by nectary structure and function.
Type
Thesis
Type of thesis
Series
Citation
Date
2024
Publisher
The University of Waikato
Rights
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.