Ultrapure water from a Rephile Ultrapure water system (Boston, MA, USA) was used to prepare the infusions, and as solvent for the analysis. Acetonitrile and formic acid used for analysis were obtained from Sigma-Aldrich (Johannesburg, RSA). The reference standards, namely caffeic acid, 4-chlorogenic acid (4-CGA), chlorogenic acid (CGA), 1,5-dicaffeoylquinic acid (1,5-DCQA), 3,5-DCQA, 4,5-DCQA, 3,4-DCQA, ferulic acid, luteolin, neo-chlorogenic acid (neo-CGA), quercetin, scopolin, and scopoletin that were used for the comparative analysis, were purchased from Chengdu Alfa Biotechnology (Chengdu, China). All reference standards had a purity grade of > 98%.

The four A. afra plants that were investigated in this study were located in the Botanical Garden of North-West University (NWU), Potchefstroom, South Africa (GPS coordinates, − 26.6823° S, 27.0950° E). Three of the plants were growing in a maintained section of the garden, whilst the fourth plant grew in an unmaintained (wild) section of the garden. Moreover, plant 1 was growing in full shade whilst plants 2–4 were growing in a semi-sun environment. Samples of approximately 10 cm of the top part of a twig from each of the four plants were collected monthly to yield 12 batches per plant over a 1-year period. After collection, the material was immediately placed in an oven at 40 °C for 6 days, whereafter the dried leaves were stripped from the twigs and stored until analysis. Herbarium specimens have been deposited at the A.P. Goossens herbarium at NWU for positive identification (specimen numbers: PUC0015502, PUC0015503, PUC0015504, PUC0015505).

Table 1 provides the date, time of day, and climatic conditions on each day of collection. More exact and expanded climatic data for collection days can be assessed from a weather archive (World Weather Online 2023).

Infusions were prepared by adding exactly 1 g of plant material to 5 ml boiling ultrapure water, which were left to infuse for 10 min whilst covered. The infusions were then vortexed for 30 s to ensure proper mixing. Of each infusion, 1 ml was filtered through a 0.45-µm syringe filter into UPLC vials for analysis.

Standard solutions of the thirteen purchased reference standard compounds (caffeic acid, 4-CGA, CGA, 1,5-DCQA, 3,5-DCQA, 4,5-DCQA, 3,4-DCQA, ferulic acid, luteolin, neo-CGA, quercetin, scopolin, and scopoletin) were prepared by accurately weighing 1 mg of each compound into individual 2-ml amber UPLC glass vials. The compounds were then dissolved with 1 ml of methanol to yield a 1000 μg/ml concentration. A volume of 100 μl of each reference standard solution was mixed together in another vial to yield a stock mixed solution with a final concentration of 76 μg/ml of each marker.

The samples were analysed on a Shimadzu i-Nexera ultra-high performance liquid chromatography (UPLC) system equipped with a quaternary pump, an auto sampler, and a photodiode array detector. The system was fitted with an Agilent Poroshell 120 EC-C18, 3 × 150 mm, 2.7-µm column with the column oven set to 35 °C. The solvent system consisted of water + 0.1% formic acid (A) and acetonitrile (ACN) + 0.1% formic acid (B) and a step-wise gradient system was employed: 0 min, 10% B; 5 min, 10% B; 6 min, 20% B; 10 min, 20% B; 11 min, 85% B; 16 min, 85% B; 16.10 min, 100% B; 17 min, 100% B, 17.10 min, 10% B, 20 min 10% B. UV detection was carried out at 254 nm and UV spectra were collected between 190 and 800 nm. The flow rate was 0.49 ml/min and 10 µl of each sample was injected. After every 12 samples, a blank injection was conducted in order to test for and prevent any carry over between samples.

The method used to determine the enzyme inhibition was adapted from Kazeem et al. (2013). α-Glucosidase from Saccharomyces cerevisiae was used to determine the α-glucosidase inhibition of selected phytocompounds present in the A. afra infusions. A p-nitrophenyl glucopyranoside (pNPG) solution was prepared in 20 mM phosphate buffer (pH 6.9). A volume of 100 µl α-glucosidase (1 U/ml) solution was preincubated with 50 µl of the reference standards (1 mg/ml) for 10 min at 37 °C in a transparent flat-bottom 96-well plate (Sarstedt AG, Sevelen, Switzerland). A volume of 50 µl of 3 mM pNPG solution in phosphate buffer (pH 6.9) was added to the α-glucosidase/reference standard solution as the substrate to start the reaction. The reaction mixture was incubated at 37 °C for 20 min. The activity was determined by measuring the yellow-coloured paranitrophenol released from the pNPG at 405 nm using a Spectramax™ Paradigm plate reader. All experiments were conducted in triplicate and averages were obtained. Distilled water and acarbose were used in the place of the reference standards as negative and positive controls, respectively. Once the absorbance was measured, the % inhibition of each sample was calculated using the following formula:

$$\mathrm=\left(\frac-}}}\right)\times\;100$$