This dataset provides nectar production values for 178 native plant species found on Grootbos Nature Reserve, South Africa.
Floral nectar availability varies seasonally, creating potential hunger gaps or deficits for nectar-feeding species. Invasive flowering plants may supplement nectar supplies in introduced ranges and stabilise availability over time, but this effect remains poorly quantified. We quantify landscape-scale nectar provision across three vegetation types in South African Fynbos, modelling how habitat diversity and three invasive Eucalyptus species influence temporal nectar stability for pollinators.
Fortnightly surveys at 24 sites recorded flowering phenology and nectar production for native plant species and three invasive Eucalyptus species. Adding ten Eucalyptus trees per hectare of natural vegetation increased temporal nectar stability by 11–26% and total nectar availability by up to 94%. Eucalyptus camaldulensis flowered during otherwise nectar-poor periods, reducing seasonal ‘hunger gaps’ for pollinators. Across all vegetation types, a small subset of native species contributed the majority of nectar resources.

Mathematical simulations further showed that access to multiple vegetation types significantly enhanced nectar stability, highlighting the importance of habitat heterogeneity for pollinators. Together, these findings provide a framework for estimating nectar provision at landscape scales and demonstrate how native plant diversity, alongside carefully managed non-native species, can help sustain pollinators in changing environments.

Data collection was carried out at the Grootbos Nature Reserve, Walker Bay, South Africa. The reserve is c. 26km² in size and largely made up of three distinct Fynbos vegetation types: Agulhus Dune Strandveld (ADS), Agulhus Limestone Fynbos (ALF), and Overberg Sandstone Fynbos (OSF).

Twenty-four 100m long transects were surveyed fortnightly from January to November 2023, eight in each of the three vegetation types. Floral diversity and abundance at the transects were recorded during surveys by placing a 2x2m quadrat every 10m along each transect, and recording the abundance and diversity of inflorescences.

For each flowering species, we aimed to collect twenty nectar samples from multiple plants in at least two separate locations, as per Baude et al. 2016 (Baude et al. 2016). In short, pollinators were excluded from each sampled flower for 24 hours by enclosing flowers within mesh bags, after which nectar was collected using glass microcapillary tubes. The volume of nectar within the tube was measured using calipers, and then nectar sugar concentration was measured using a hand-held refractometer. The volume and concentration of each sample were then used to determine the total amount in µg of sugar per flower, using the formula used by Corbett (2003) (Corbet 2003); d = 0.0037921C + 0.0000178C² + 0.9988603, where C is percent sugar concentration. In addition, to assess how nectar production correlates with inflorescence volume, the width and height of inflorescences were measured for 54 of the 159 species recorded at the transects, using the same methods as Baude et al. 2016 (Baude et al. 2016).
A different nectar collection method was used for species in the genus Protea, as their flowers typically produce large volumes of nectar not easily measured using microcapillary tubes. Nectar sugar production of Protea inflorescences were measured using a ‘field centrifuge’ (Armstrong and Paton 1990; Nottebrock et al. 2017), in which flower heads are removed from the plant, secured in plastic collecting bags, tied to a rope and spun by hand for 40 rotations. The centrifugal force displaces the nectar from the flower so that it can then be extracted from the plastic bag using a syringe to measure the volume, and the sugar concentration measured using a refractometer, as above.
In total the nectar production of 159 species was measured in the field; a total of 3,085 samples. The mean, standard deviation and variance of nectar sugar produced per day for each species were calculated, per inflorescence and per one square kilometre of Fynbos vegetation, based on the mean abundance of each species per 2x2m quadrat, across all transect surveys where the species was recorded. A linear regression was used to test the effect of plant family on nectar production. Regression equations for each family were created to estimate the nectar production of a species based on its inflorescence volume, for four major plant families: Proteaceae, Geraniaceae, Iridacece and Asteraceae, an approach used by Baude et al. 2016 (Baude et al. 2016). Where regressions were not available due to small numbers of species recorded, nectar values from species within the same genus and of closest inflorescence volume were used for missing species (a total of 14 species). Together the field-collected and predicted measurements provided nectar production data for 178 species.