Determination of Physicochemical Characteristics, Phytochemical Profile and Antioxidant Activity of Various Clover Honeys

This paper reports on some physicochemical and phytochemical characteristics (i. e. pH, electrical conductivity, colour, moisture content, total phenolic content, sugar profile) and in vitro antioxidant activity of honeys harvested from five legume species, red clover (Trifolium pratense), balansa clover (T. michelianum), Persian clover (T. resupinatum), purple clover (T. purpureum) and sanfoin, also known as holy clover (Onobrychis viciifolia), that were grown in enclosed shade houses to ensure that the honeys’ characteristics are reflective of a truly monofloral honey. Glucose and fructose, determined via High‐Performance Thin‐Layer Chromatography (HPTLC) analysis, were found as the main sugars in all investigated honeys with the ratio of fructose to glucose ranging from 1 : 1.2 to 1 : 1.6. The honeys’ pH values ranged from 3.9 to 4.6 which met Codes Alimentarius (CA) requirements. The moisture content was found to be between 17.6 and 22.2 % which in some cases was slightly higher than CA requirements (≤20 %). The honeys’ colour values, prior and after filtration, were between 825.5–1149.5 mAU and 532.4–824.8 mAU respectively, illustrating golden yellow to deep yellow hues. The total phenolic content (TPC) of the honeys was determined using a modified Folin‐Ciocalteu assay. Their antioxidant activity was captured by the Ferric Reducing‐Antioxidant Power (FRAP) assay as well as HPTLC analysis coupled with 2,2‐diphenyl‐1‐picrylhydrazyl (DPPH) derivatisation. The highest total phenolic content was found in red clover honey (45.4 mg GAE/100 g) whereas purple clover honey showed the highest level of activity in the FRAP assay (7.3 mmol Fe2+/kg). HPTLC‐DPPH analysis of the honeys’ organic extracts demonstrated the presence of various bioactive compounds that contribute to their overall antioxidant activity. This study developed a methodology for producing monofloral clover honeys in a space limited, enclosed production system, which allowed to collate important baseline data for these honeys that can serve as the foundation for their potential future development into commercial honeys, including honeys that can be used for medicinal purposes.


Introduction
Honey is a sweet, sticky natural substance produced by bees (in particular Apis mellifera) mainly from flower nectar. [1][4] The range of bioactivities recorded for different honeys, such as antibacterial, antioxidant, antifungal, antidiabetic and anti-inflammatory effects, are mainly related to the floral source and geographical origin of the honey, as well as its specific phytochemical profile, particularly its phenolic constituents, and physicochemical properties. [3,4]lover honey is a unique type of honey produced from various species of the Trifolium genus, [5] which includes clovers and trefoil, which are low leguminous annual, biannual or perennial herbs, often with trifoliate leaves and dense white, yellow, red or purple flower heads. [5]The word clover is likely derived from the old German word Klaiwaz, which refers to the stickiness of the sap of clover or the honey produced from the plant.Over time, this changed to Klaifre and then in Old English to Clafre from which the modern English word Clover is derived. [6]Clover cultivation started in Europe in the 16 th century, by that time clover honey was already well known.It is, for example, mentioned in Kilian's 1599 Dutch dictionary as Klauern Honigh ('clover honey'), where it is defined as "mel optimum and candidissimum, ex trifolio pratensi", which translates as 'good, very clear honey from purple clover'. [6]lover plants grow in a wide range of soils, rainfall zones and climates and are therefore found all around the world, with the highest diversity in species in the temperate northern hemisphere.There are approximately 230 species of clover within the genus Trifolium, of which less than 20 are cultivated as forage. [7]Many of the clover species are of Mediterranean origin and are adapted to low rainfall environments.A large number of species are also found in both South America and Africa, including at high altitudes on mountains in the tropics. [8,9]Clovers are generally small annual, biennial, or shortlived perennial herbaceous plants.Several species are extensively cultivated as fodder plants as they grow freely, shooting up again even after repeated grazing, and commonly produce abundant crop which is palatable and nutritious for livestock.Next to their use as pastures and forage crops, clovers are also popular for green composting and roadside beautification.They also play an important agricultural role with their ability to fix atmospheric nitrogen into the soil, serving as a sustainable alternative to manufactured mineral nitrogen fertilisers. [10]lover species are also a significant nectar and pollen resource for beekeepers as pollination is required for these mostly self-incompatible species to promote seed set. [11]Clovers present a prolific source of pasture honeys with a strong nectar aroma and light nectar flavor. [8]The most important clover species for honey production to date are white or Dutch clover (Trifolium repens L.), alsike clover (T.hybridum), red clover (T.pratense), and crimson clover (T.incarnatum). [11,12]However, physico-and phytochemical data as well as information on bioactivity for these and other clover honeys are sparse.To address this gap in knowledge, the present study reports on some physicochemical and phytochemical characteristics and antioxidant activity of a range of clover honeys (e. g. red, balansa, Persian and purple clover).While not a Trifolium species, sainfoin (Onobrychis viciifolia Scop.), which is also referred to as holy clover, [7] was also included in the study as it is a deep-rooted perennial legume that has been used as a forage legume for hundreds of years in Europe and Asia.Sainfoin produces copious amounts of nectar and is highly attractive to honey bees (https://plants.usda.gov).A particular feature of this study is that all plant species were grown in enclosures (shade houses), using automatic watering systems and adequate fertilisation and, most importantly, that the honey samples were directly collected from these sites, which limited bees' access to only these nectar sources to produce truly monofloral honeys.This study design allows to showcase six clover honeys, which are currently not commercially available in Australia, but show potential as future sources of clover pasture honeys that warrants further research.

Physicochemical Characteristics
The pH, electrical conductivity, colour and moisture content of the six investigated clover honeys are presented in Table 1.All clover honeys were acidic (pH 3.90-4.30)and within the pH range (3.4 to 6.1) for honey stipulated by the Codex Alimentarius Commission. [13]The electrical conductivity of clover honeys ranged from 0.48 to 1.27 mS/cm, meeting Codex Alimentarius requirements (< 0.8 mS/cm for most honeys and � 0.8 mS/cm for honey produced from honeydew, chestnut blossoms, and mixtures of these honeys). [13]The color of a honey is influenced by its botanical origin, the composition of the nectar, processing conditions as well as storage temperature and time.In this study, Persian clover honey was found to be the darkest and red clover NSE honey the lightest in colour for both the filtered and unfiltered sample.The recorded water content of the clover honeys ranged from 17.61 to 22.21 %.According to the Codex Alimentarius Commission, the acceptable moisture content for honeys (except Heather honey) is less than 20 %.Persian and purple clover honey had a water content of 21.21 and 22.21 % respectively, which is slightly outside this range but falls with the range of Heather honey (not more than 23 %). [13]o date the potential of the six investigated clover species for honey production remains largely unexplored.In Australia, monofloral honeys of these species are currently not commercially available and no physicochemical, phytochemical data nor information on their antioxidant activity is available.To provide a geographical point of reference given this paucity in scientific data, their characteristics are discussed here in comparison with a large set (n = 453) of Western Australian (WA) honeys (unpublished data) that have been investigated in the Honey Research Lab of the Division of Pharmacy, University of Western Australia using the same analytical methodologies as employed in this study.The pH values of all clover honeys this study investigated (Table 1) were slightly more acidic than the majority of the investigated WA honeys, which had a mean pH of 4.56, ranging from 3.86 to 5.69. [14]A slightly higher water content was also detected in the investigated clover honeys compared to the majority of the WA reference honeys, which ranged in moisture content from 14.05 to 26 %, with a mean of 17.47 %. [14][15][16] Phytochemical Characteristics

Total Phenolic Content
The total phenolic content (TPC) of the investigated honeys is presented in Table 2. Several studies have linked the color of honey to its phenolic content and its antioxidant capacity with darker colours commonly corresponding to higher TPC and antioxidant activity [17][18][19] This correlation was, however, not noticed in this study, which might be due to the presence of antioxidant compounds such as simple phenolics that do not contribute to the honey's color.NSE red clover honey, which was found to be lightest in colour, displayed the highest phenolic content (45.4 mg GAE/100 g), followed by NFE red clover, sainfoin, Persian clover, purple clover and balansa clover honey.
In the TPC assay, the recorded values for the investigated clover honeys ranged from 17.21 to 45.41 mg GAE/100 g, with red clover (NSE) honey recording the highest level of phenolic constituents (45.41 GAE/100 g).Multiple pair-wise comparisons by one-way ANOVA found a significant difference (p < 0.0001) in TPC among all the honeys except for purple and balansa clover for which a less significant difference (p = 0.0383) was found.The TPC of some WA honeys was previously reported to range between 18.90 and 75.56 mg GAE/100 g, [20] however these values were derived using the traditional Folin-Ciocalteu assay carried out at pH 10.8, which has been found to result in a significant overestimation of the TPC of honey due to the interference of reducing sugars which were also noted to react with the reagent. [21]The assay used in this study and also for the large number of other WA honeys (n = 453) investigated in our lab employed a modified Folin-Ciocalteu protocol in which the concentration of the sodium carbonate solution was optimised in such a way (pH 7.9) that sugar interference was muted.When compared to other WA honeys using these assay conditions, the clover honeys investigated in this study were found to have medium-range TPC values.For Jarrah honey, for example, which is a highly sought after honey endemic to WA that is renowned for its antibacterial and antioxidant activity, [21] a TPC of 56.03 was determined, WA Manuka honey had a TPC of 54.00 and a multifloral honey of 35.10 GAE/100 g.

Quantification of Selected Sugars by HPTLC
In the HPTLC-based analysis, glucose presented as a green band with a Rf value of 0.30 and fructose as an orange band with a Rf value of 0.14 (Figure 1).Fructose (F) and glucose (G) contents were easily quantifiable in all honeys, whereas sucrose and maltose, if potentially present, were below the limits of detection and quantification.All clover honeys contained more fructose (F) than glucose (G), as can also be seen in their respective F : G ratio.
As in other honeys produced by European honey bees, the main sugars present in the studied six clover honeys were fructose and glucose.Their F : G ratios ranged from 1.21 to 1.60 (Table 3) which is within the range of the large number of WA honeys studied (their F : G ratio ranged from 0.9 to 1.84).The multiple pair-wise comparisons by one-way ANOVA for fructose and glucose content in clover honeys in shown in Table 3. Significant differences in both fructose and glucose content   Samples are named as A-F.a-s Different superscript letters in the column denote significant differences (ANOVA.p < 0 .05).
were found for few of the investigated pairs (Table 3).Across the globe wider ranges for F : G ratios have been found, with honeys from some botanical species (e. g.Leptospermum) featuring more glucose than fructose (F : G 0.4 and 1.6). [16,22]oneys with a large F : G ratio, like the clover honeys investigated here, might be of interest to some consumer segments that need to restrict their glucose intake as some studies have reported honey to have hypoglycaemic effects.
[25] Moreover, there is evidence that fructose tends to lower blood glucose in animal models [26,27] which is postulated to stem from its stimulation of glucokinase in hepatocytes, which in turn influences the hepatic uptake and storage of glucose as glycogen.Glucose on the other hand, which is also present in honey, enhances the absorption of fructose and promotes its hepatic actions through its enhanced delivery to the liver. [28,29]tioxidant Activity

FRAP Assay
Table 4 displays the average FRAP antioxidant activity of the investigated honeys, expressed as mmol Fe 2 + equivalent/kg.Among the six investigated honeys, purple clover showed the highest FRAP antioxidant activity followed by holy clover, red clover (NSE), Persian clover, red clover (NFE) and balansa clover.Multiple pair-wise comparisons by one-way ANOVA (Table 4) showed a significant difference in FRAP activity for about half the investigated pairs.Next to its many topical applications for wound healing [30][31][32][33][34][35][36] honey is also commonly used in cold and cough medicines and home remedies due to its health-promoting properties, including its antioxidant and antibacterial qualities and generally soothing effect on sore throats. [30]Despite these potential health benefits, limited clinical studies have to date been carried out using clover honeys, which focus in particular on its wound healing properties.In one 3-month study in which clover honey was used as a dressing for 30 different diabetic foot ulcers, 43 % of wounds healed completely, and another 43 % were significantly reduced in size and bacteria count. [37][33][34][35][36] When compared with FRAP data derived for the large set of WA honeys, which had a mean FRAP value of 6.49 with a range of 3.47 to 11.66 mmol Fe 2 + / kg, the clover honeys investigated in this study can, again, be considered to have average antioxidant activity.Interestingly, although not particularly high in TPC, in the FRAP assay purple clover honey displayed the highest antioxidant activity (7.30 mmol Fe 2 + /kg) amongst the investigated clover honeys.Many previous studies, in contrast, reported a close correlation between performance in the FRAP assay and TPC. [20]It therefore appears that purple clover honey might have antioxidant constituents that are not phenolic in nature, so this constitutes an interesting finding which should open a future line of inquiry.

HPTLC-DPPH Assay
The fingerprints of honey extract samples after derivatisation with DPPH reagent are shown in Figure 2 where individual antioxidant bands are reflected by a yellow colour.The level of antioxidant activity of the various clover honeys, expressed semi-quantitatively based on their % DPPH radical scavenging activity, is shown in Tables 5 and 6.
In the HPTLC-DPPH assay, the solvent system used in the development of the HPTLC plate is an important factor to separate individual band.After developing the HPTLC plates in two different mobile phases of varying polarities (Mobile Phases A and B), the plates were derivatised with DPPH reagent which allowed to specifically capture antioxidant bands.Their DPPH radical scavenging activity was then calculated based on the band's respective colour by converting its RGB value (generated automatically by the HPTLC software) into a corresponding hue value. [15]The obtained hue value was used to express the level of scavenging activity in relation to that of gallic acid, which is the standard reference compound to express total antioxidant activity in the FRAP assay (Table 5 and Table 6).
In the HPTLC-DPPH bioautographic analysis, more antioxidant bands were found using Mobile Phase B compared to  Mobile Phase A, suggesting that antioxidant compounds of a more polar nature were present in the investigated honeys.For example, in case of NFE clover honey, three unidentified bands were found with Rf values of 0.49, 0.70, 0.73 using Mobile Phase A while six unidentified bands were found at Rf values of 0.07, 0.20, 0.29, 0.50, 0.62, 0.70 using Mobile Phase B. While the findings of this study establish a basic understanding of the nature and strength of the antioxidant activity of the investigated clover honeys, future research is necessary to shed more light on the chemical identity of the compounds that contribute to the honeys' antioxidant bioactivity and their relative concentration in the respective honey matrix.
Next to generating new data on a range of clover honeys that have not yet been investigated for their physicochemical and phytochemical characteristics and their antioxidant activities, this study also introduced a new approach to harvesting monofloral clover honeys produced under standardised growing conditions.This production pathway of monofloral honeys of annual and perennial clovers makes it much easier to establish key physico-and phytochemical as well as bioactivity data compared to honeys harvested from natural areas where co-flowering species might introduce variability.Methods were also developed for maintaining the health of honey bees on pasture legume species in small enclosed shade houses.The baseline data generated taking this approach can form the impetus for more in-depth studies, particularly with respect to the identification and quantification of non-sugar compounds that might contribute to the honeys' bioactivities.This was demonstrated in this study for the analysed purple clover honey, where more in-depth investigations into the presence of non-phenolic antioxidant constituents were found to be warranted.

Conclusions
This study introduced a novel approach for producing truly monofloral honeys harvested under standardised conditions, which allowed to establish reliable physico-and phytochemical characteristics and bioactivity data for six clover honeys from five different botanical species as a foundation for more indepth follow up studies.Although the demand for specialist pharmaceutical honey products is gaining momentum worldwide, currently, only limited published information is available on these and other clover honeys in terms of their key chemical characteristics and antioxidant bioactivities.The potential future utilisation of clover-derived honeys as therapeutic agents presents an intriguing opportunity for the beekeeping industry in Australia.The study revealed that it is possible to sustain a small bee colony within an enclosed shade house containing a diverse range of annual and perennial clover species.This innovative method can now enable the production of pure monofloral honeys, which can be utilised, for example, for medicinal testing purposes.This study therefore makes impor-

Experimental Section Plant Growth
Six pasture cultivars from five legume species (both annual and biannual) were established in a glasshouse from seed (June 2022).Two red clover cultivars (referred to here as NSE and NFE), which were previously developed for phytomedicinal applications, were included.Three seeds of each cultivar were sown into Jiffy peat pots (size 522 Jiffy products Ltd, Norway).Each population was sown in a seedling tray containing 30 peat pots and covered lightly with commercial potting mix.Three days after sowing, appropriate rhizobium inoculant was applied by watering can, by adding 20 mL of inoculum to 10 L of water.Peat pots were watered daily.Prior to transplanting, seedlings were thinned randomly to one per peat pot.In August 2022, seedlings of all cultivars/selections were transplanted at 30 cm spacing into weed stop matting (woven polypropylene) in their allocated plots in the enclosed shade houses (10 m×4 m) at the University of Western Australia Shenton Park Field Station (Figure 3, Table 7).The purpose of the enclosed shade house plots was to collect monofloral honeys from a hive of bees foraging in each respective enclosure.The shade houses were adjacent to each other and were similar in soil type.Irrigation was provided through 19 mm premium black polypipe (Holman) using micro jets (Pope half-circle Veri-flow).Seedlings after transplanting and established plants were fertilised as required with soluble fertiliser (Thrive® All-Purpose) using a watering can.

Nucleus Beehives
Each enclosed shade house had a nucleus beehive placed when half of all plants were in flower.The design of the hives included a marked dry-drawn frame, foundation (x2), brood (x2) and a food frame (Supplementary Figure ).The nucleus beehives were checked regularly by an apiarist.To ensure bee health, water was made freely available.

Honey Collection
The nucleus bee hives were checked regularly and frames with capped honey were transported to the laboratory and placed in 4 °C storage for 3 days.The honeys were then manually removed from the frame (i.e. without an extractor) by first excising the honey comb using a clean knife followed by collection of the honeys with a plastic pipette (Supplementary Figure ).The collected honeys were stored in plastic containers at 4 °C and protected from light until further analysis.

Statistical Method
Experiments were completed in triplicates, and the results were calculated by one-way analysis of variance (ANOVA) followed by Tukey's significant (TukeyHSD) test using GraphPad Prism 9.4.1 (GraphPad Software, San Diego, CA, USA).A p-value of less than 0.05 was considered statistically significant.

Physicochemical Characteristics pH
The pH of the honeys was measured at room temperature using a calibrated pH Meter (Eutech PC 2700-Eutech Instruments, Vernon Hills, IL, USA) following the method described by Hossain et al. [13] Briefly, a honey solution was prepared by adding 7.5 mL of carbon dioxide-free water to 1 g of honey and the pH determined.Results are expressed as mean � SD of three independent samples.

Electrical Conductivity
A solution of 20 % (w/v) clover honey was prepared in deionized water.The electrical conductivity of the prepared solution was measured at 22 °C using an Electrical Conductometer (Eutech PC 2700-Eutech Instruments, Vernon Hills, Illinois, US) and expressed as milliSiemens per centimeter (mS/cm). [38]

Colour
The colour of the honey samples was determined following a method outlined by Hossain et al. [11] Firstly, honey samples were dissolved in deionised water (50 % w/v) followed by measuring the optical density at 450 nm and 720 nm using a UV-Vis Spectrophotometer (Cary 60, Agilent Technologies, Santa Clara, CA, USA).To obtain the respective colour value, the difference in optical density between the two measurements was expressed in milli-absorbance units (mAU).Colour values were determined in triplicates for all honeys, both before and after filtration of samples using a 0.7 μm syringe filter (Merck KGaA, Darmstadt, Germany).Results are expressed as mean � SD.

Moisture Content
Moisture content was determined as described by Hossain et al. [39] Briefly, 100 mg honey samples were placed on the sample well of a Refractometer (HI96801, Hanna Instruments, Smithfield, RI, USA), readings were taken in triplicate and the result expressed as a percentage (w/w). [15]ytochemical Characteristics

Total Phenolic Content (TPC)
The total phenolic content (TPC) of the various honey samples was determined by a modified Folin-Ciocalteu (FC) assay described by Lawag et al. [21] Briefly, 1 mL of FC reagent was added to 200 μL of the respective honey solution (20 % w/v in deionised water) / gallic acid standards (30-90 μg/mL).The mixtures were allowed to react for 5 min at room temperature followed by the addition of 800 μL of Na 2 CO 3 solution (0.75 % w/v) to create a basic environment (pH 7.9) to mute potential sugar interference.The resulting mixtures were incubated without light exposure for 2 h before their absorbance was measured at 760 nm (Carry 60 Bio UV-Vis Spectrophotometer).

Quantification of Selected Sugars by High Performance Thin Layer Chromatography (HPTLC)
The quantification of selected sugars (i.e., glucose, fructose, sucrose, maltose) in the honeys was conducted following a method described by Islam et al. [39]

Standards and Reagent Preparation
To prepare the glucose, fructose, sucrose and maltose standard solutions, 25 mg of the respective sugars were dissolved in 100 mL of 50 % aqueous methanol.To prepare the derivatisation reagent, 2 g of diphenylamine and 2 mL of aniline were dissolved in 80 mL of methanol, 10 mL of phosphoric acid (85 %) were added, and the solution was made up to 100 mL using methanol.

Sample Preparation
Hundred milligrams of each honey were dissolved in 80 mL of 50 % aqueous methanol by sonication, then made up to 100 mL with 50 % aqueous methanol.

Chromatographic Method
The standard solutions and samples were applied as 8 mm bands at 8 mm from the lower edge of the HPTLC plate (glass plates 20×10 cm, silica gel 60 F 254 ) at a rate of 50 nLs À 1 using a semiautomated HPTLC application device (Linomat 5, CAMAG, Muttenz, Switzerland).To prepare the glucose, fructose, sucrose, and maltose standard curves 1 μL, 2 μL, 3 μL, 4 μL, and 5 μL of the respective standard solutions were applied.For the analysis of sugars in clover honey, 3 μL of the sample solution was applied.
The chromatographic separation was performed in a saturated (33 % relative humidity) automated development chamber (ADC2, CAMAG, Muttenz, Switzerland).The development chamber was saturated for 60 min, the plates were pre-conditioned with the mobile phase for 5 min, and automatically developed to a distance of 85 mm at 25 °C and dried for 5 min.
After initial documentation of the chromatographic results, each plate was derivatised with 2 mL of aniline-diphenylaminephosphoric acid reagent using a TLC derivatiser (CAMAG Derivatiser, Muttenz, Switzerland).The derivatised plate was heated for 10 min at 115 °C using a TLC Plate Heater III (CAMAG, Muttenz, Switzerland).The plate was then cooled to room temperature and analysed with the HPTLC imaging device under transmission white (T white) light.The chromatographic images were digitally processed and analysed using a specialised HPTLC software (visionCATS, CAMAG, Muttenz, Switzerland) which was also used to control the individual instrumentation modules.

FRAP (Ferric Reducing-Antioxidant Power) Assay
The FRAP assay was performed according to the protocol described by Lawag et al. [20]

Sample and Reagent Preparation
Clover honeys were prepared in deionised water at a concentration of 20 % (w/v).For the FRAP reagent, 10 mM TPTZ (prepared in 40 mM HCl), 20 mM FeCl 3 (aqueous) and 300 mM acetate buffer (pH 3.6) were mixed in a ratio of 1 : 1 : 10 (v/v/v).Ferrous sulphate (FeSO 4 .7H 2 O) standards (200 μM to 1200 μM) were used to construct the standard curve for quantification of the recorded antioxidant activity.

Working Procedure
Twenty microlitre of the honey solutions or the ferrous sulphate standards were mixed with 180 μL of FRAP reagent in a 96-well microplate (Greiner Bio-One 96-well Microplate Flat Bottom), and after 30 min incubation at 37 °C, the absorbance of the solution was measured at 620 nm (BMG Labtech POLARstar Optima Microplate Reader).The recorded FRAP antioxidant activity was expressed as mmol Fe 2 + equivalent (FE)/kg fresh weight of the honey.

High-Performance Thin-Layer Chromatography (HPTLC) Coupled with 2,2-Diphenyl-1-Picrylhydrazyl (DPPH) Derivatisation
The HPTLC-DPPH antioxidant activity of the investigated honeys was determined semi-quantitatively using the method described by Lawag et al. [40] with minor amendments.The employed HPTLC-DPPH analysis enabled the visualisation of individual antioxidant components present in the honeys' organic extracts and thus reflects the antioxidant band activity of their non-sugar fractions.

Standard Solution and Reagent Preparations
A gallic acid standard solution (20 μg/mL) and 4,5,7-trihydroxyflavanone reference solution (0.5 mg/mL) in methanol were prepared.Two mobile phases were used to provide adequate band separation between the range of extracted compounds, namely mobile phase A consisting of dichloromethane: ethyl acetate: acetic acid in a ratio of 12: 1: 2 (v/v/v), and mobile phase B, prepared by mixing ethyl acetate: methanol: formic acid: glacial acetic acid in a ratio of 9 : 1 : 1 : 1 (v/v/v/v).The DPPH derivatisation solution was prepared in an amber glass bottle by dissolving 40 mg DPPH in 10 mL of a mixture of 50 % methanol and 50 % ethanol.Being light and temperature sensitive, the prepared derivatisation solution was stored at 4 °C protected from light.

Sample Preparation
One gram of each honey was mixed with 2 mL of deionised water.The resulting solutions were each extracted three times with 5 mL of a mixture of dichloromethane and acetonitrile (50 : 50 v/v).The combined organic extracts were dried at room temperature and any trace of water was removed with anhydrous MgSO 4 .The extracts were kept at 4 °C until further analysis.For HPTLC analysis, the samples were reconstituted in 100 μL methanol.

HPTLC Analysis
In this assay, chromatographically separated compounds that possess antioxidant activity would quench the DPPH radical either by electron transfer or hydrogen atom transfer through radical attack, which is observed as a discoloration at 517 nm due to the formation of 2,2-diphenyl-1-hydrazine or a substituted analogue hydrazine. [41]Seven microliters of each honey extract were used for the analysis alongside the gallic acid standards in varying volumes (2-5 μL).After development, the plates were derivatised (yellow nozzle, spray level 1) with 3 mL of 0.4 % DPPH solution. [16]Plate images were obtained in transmittance mode at white light after 2 h.Peak profiles at 517 nm were also generated and from these the Rf values of the respective peaks derived.Each band's colour in the form of RGB values was determined and then converted into corresponding hue values. [40]Gallic acid was used as positive control; its quenching activity resulted in a maximum hue value of 40°(yellow colour).The DPPH radical scavenging activity (% DPPH RSA) of a band of interest was then calculated as follows: where: H°P!40°-Hue values (°) of unreacted DPPH on the plate, ΔH°B-hue values (°) of the bands up to 40°.The % DPPH RSA of bands was then classified into four groups: 0.0 % (no activity), 1.0-33.3% (low activity), 33.4-66.6 % (medium activity) and 66.7-100.0% (high activity).

Figure 3 .
Figure 3. Clover plants grown in enclosed shade houses (a) seedlings in glasshouse, (b) transplanting of seedlings into weed stop matting, (c) flowering of NSE red clover, (d) flowering of NFE red Clover, (e) flowering of sainfoin clover, (f) Checking of nucleus hives.

Table 5 .
Percent DPPH RSA antioxidant activity of individual bands (Mobile Phase A).

Table 6 .
Percent DPPH RSA antioxidant activity of individual bands (Mobile Phase B).

Table 7 .
Clover honey samples including botanical origin.