Pollen Types Reveal Floral Diversity in Natural Honeys from Campeche, Mexico

Abstract

The Yucatan Peninsula, located in southern Mexico, is a central honey-producing region with extraordinary biodiversity of melliferous plants. Approximately 900 plant species have been described as being a source of nectar and pollen for bees and other pollinators. They provide ecosystem services that help to keep plant biodiversity high and mitigate the effects of climate change. This study aimed to reveal the diversity of the pollen content in honey through a melissopalynological analysis of 22 honey samples collected in February–August 2021 from the north-central area of Campeche, Mexico. The extraction of pollen from the honey was carried out using standard methods for melissopalynological analysis. The honeys were classified by botanical origin to determine their floral sources and a diverse spectrum of 19 pollen types from 13 families was identified. Only eight were predominant: Milleria quinqueflora, Gymnopodium floribundum, Terminalia buceras, Amaranthus spinosus, Zea mays, Talisia floresii, Guazuma ulmifolia, and Croton icche. Our research shows the high quality of the honey analyzed and highlights the diversity and critical role of local melliferous flora and crops in beekeeping development in southern Mexico. The results in this study are useful for confirming the botanical origins of honey, generating information for designing nature conservation and agroecosystem management strategies, and increasing the knowledge of beekeepers in Campeche, Mexico.

1. Introduction

Honey is the natural sweet substance produced by bees from sweet deposits from plants. It is a complex mixture containing nutrients and bioactive compounds and is a sustainable, innocuous food. Its quality is based on its organoleptic properties, such as aroma, flavor, color, physical and chemical composition, geographical area and climatic condition, botanical origin, antiseptic condition, utility (medicinal), and applications (antibacterial) [1,2,3].

Every resource used must be complemented by restoration or conservation activities to protect the vegetation of the area where honey is produced [4]. Therefore, a study based on the floral origins of honey will provide an overview of the vegetation visited by the bees. Bees may prefer a particular plant species to supply them with nectar, pollen, or both [5,6] from which they produce monofloral or multifloral honey [7,8].

Bees are excellent pollinators [2], thus they are relevant for plant ecosystems and farming. Worldwide, beekeeping is an activity dedicated to bee breeding and care to obtain, collect, and consume bee-generated products [9]. Thanks to beekeeping, Mexico is the ninth largest honey producer worldwide and the thirteenth largest honey exporter. The destination countries are mainly located in Europe and Asia. In 2020, the trade volume was 63,362 tons with total revenue of USD 67.9 million [10]. The Yucatan Peninsula (YP) encompasses the states of Quintana Roo, Yucatán, and Campeche. Campeche ranks first in honey production in Mexico, with the municipalities of Champotón, Campeche, Calkiní, Hopelchén, Hecelchakán, and Tenabo being the most critical honey producers [11].

A total of ~900 species of floristic plants that provide nectar and pollen to bees have been reported in the YP. Within this great diversity of honey plants (161 families and 956 genera), only a select group of species are used for beekeeping in the YP. Approximately 90% of the annual honey production comes from the flowers of Viguera dentata (42%; flowers between December and February) and Gymnopodium floribundum (48%; flowers between March and May). The remaining 10% comprises plants from the Fabaceae family and climbing plants, such as Sapindaceae and Convolvulaceae [5,11,12].

Studies on honeybee flora in the YP have focused on the honey’s botanical origin designation by characterizing the pollen through melissopalynological analyses. These analyses determine the different plant species visited by bees since the melliferous flora (MF) varies depending on the time of year and geographical region [13,14]. Information on the floral origin allows us to understand what type of honey is obtained from the apiaries and to classify it based on the predominance of the pollen types found in it, as it adds value that can positively impact beekeepers regarding the sale price of their honey. In addition, studies on floral origin can help in the development of strategies for preserving native MF, which could be a source of food in times of scarcity, thus highlighting the relevance of the environmental conservation of MF [15,16,17].

In the present study, we characterized 22 types of honey corresponding to the north-central zone of Campeche, Mexico. First, the pollen was identified using melissopalynological techniques through quantitative (absolute pollen count) and qualitative microscopical analyses (MF identification) of each sample. Furthermore, the Shannon–Weaver diversity index was used and the types of floral resources were determined in the samples. The results were analyzed through a principal component analysis (PCA) of qualitative data called PRO-PRINQUAL and a non-metric multidimensional preference principal component analysis (MDPREF). These analyses showed relationships between the identified flora species and the ecosystem in the zone sampled. Moreover, similarities were found among the pollen identified within different ecosystems [18]. In addition, the honey classification was based on criteria to determine whether the honey was monofloral or multifloral. Our study found a clear impact of corn on honey production in Campeche, which could be associated with a loss of biodiversity [12].

The present work shows the relevance of the role of the diverse flora and quality honey produced in Campeche and could help increase the knowledge of beekeepers and assist in the development of strategies for the conservation of native floral resources for bees in the different ecosystems of Campeche, Mexico.

2. Materials and Methods

2.1. Study Area

Campeche state is located in southeast Mexico between parallels 17°49′01″ and 20°51′37″ of north latitude and meridians 89°05′20″ and 92°28′21″ of west longitude, with an area of 56,859 km². Campeche borders in the north and northwest with Yucatán, in the east with Quintana Roo state and Belize, in the south with Tabasco state and Guatemala, and in the west with Tabasco and the Gulf of Mexico. Campeche is divided into 11 municipalities, and honey samples were taken from apiaries located in the north-central area of the state in the municipalities of Calkiní, Holpechén, Tenabo, Hecelchakán, Champotón, and Campeche. The average annual temperature is 26.6 °C and the average annual precipitation is 1272.8 mm.

The Campeche state presents great ecosystem diversity (for clarity, the ecosystem acronyms indicated below include their meaning in Spanish). Campeche ecosystems comprise low-stature subtropical forests (SBSUB = selva baja subperennifolia), medium-stature tropical forests (SMSUP = selva mediana superennifolia), low-stature tropical forests (SBPER = selva baja perennifolia), medium deciduous forests (SMSU = selva mediana subcaducifolia), low deciduous forests (SBSU = selva baja caducifolia), low inundated tropical forests (SBBA = selva baja inundable), and secondary forests (SSSU = vegetación secundaria) (

Table 1. List of geographical locations, time collections, and ecosystems of honey samples.

Sample	Location	Time Collection	Ecosystem	Coordinates	References
C029	Pich	August 2021	Low deciduous forest	N 19°29′11″/O 90°07′05″	[20]
C031	Calkiní	February 2021	Medium-stature tropical forest	N 20°22′21″/O 90°03′03″	[21]
C033	Hopelchén	April 2021	Low-stature tropical forest	N 19°32′30″/O 89°36′30″	[20]
C039	Calkiní	February 2021	Medium deciduous forest	N 20°22′21″/O 90°03′03″	[21]
C040	Calkiní	February 2021	Medium deciduous forest	N 20°22′21″/O 90°03′03″
C043	Calkiní	February 2021	Medium deciduous forest	N 20°22′21″/O 90°03′03″
C045	Calkiní	February 2021	Medium deciduous forest	N 20°22′21″/O 90°03′03″
C046	Calkiní	February 2021	Low inundated tropical forest	N 19°29′77″/O 90°56′52″
C049	Hopelchén	April 2021	Low-stature tropical forest	N 19°32′30″/O 89°36′30″	[21]
C052	Calkiní	February 2021	Medium deciduous forest	N 20°22′21″/O 90°03′03″	[22]
C054	Campeche	August 2021	Medium deciduous forest	N 19°50′55″/O 90°31′	[22]
C056	Campeche	August 2021	Medium deciduous forest	N 19°50′55″/O 90°31′
C058	Tenabo	March 2021	Low deciduous forest	N 20°04′09″/O 90°22′55″	[23]
C060	Tenabo	March 2021	Low deciduous forest	N 19°50′55″/O 90°31′
C063	Tenabo	March 2021	Low deciduous forest	N 19°50′55″/O 90°31′
C066	Hecelchakán	July 2021	Medium deciduous forest	N 20°10′37″/O 90°08′02″	[24]
C071	Hecelchakán	July 2021	Medium deciduous forest	N 20°22′21″/O 90°03′03″
C075	Canasayab	May 2021	Low inundated tropical forest	N 19°29′77″/O 90°56′52″	[25]
C077	La nueva Esperanza	June 2021	Secondary forest	N 19°19′55″/O 90°63′97″	[23]
C080	Ulumali	February 2021	Medium deciduous forest	N 19°27′25″/O 90°62′33″
CO82	Ah Kim Pech,	March 2021	Medium deciduous forest	N 18°99′88″/O 90°24′44″
CO83	Chilam Balam	March 2021	Medium deciduous forest	N 19°00′30″/O 90°56′52″

). Figure 1 shows a map of Campeche built using QGIS, a free and open source geographic information system software (version 3.18, OSGeo, Zürich, Germany), the available source: https://qgis.org/es/site/forusers/download.html. The vector files obtained from the CONABIO geoportal [19]. Vegetation distribution and the points of sampling are shown in each municipality (Figure 1).

2.2. Honey Sampling

Campeche apiaries with hives of Apis mellifera L. (Hymenoptera: Apidae) from the north-central zone were sampled, obtaining 22 honey samples of approximately 300 mL Table 1 summarizes all relevant information: code samples, geographical location samples, time collections, ecosystems, coordinates, and references that describe the ecosystems reported in the municipalities of Campeche.

2.3. Quantitative Pollen Analysis

The absolute pollen count (APC) was obtained from 10 g of honey diluted in 20 mL of distilled water (not above 40 °C), then suspended and centrifuged at 2500 rpm/min for 10 min. The supernatant was decanted and 5 mL of distilled water was added to each tube. The pellet was homogenized and centrifuged at 2500 rpm/min for 10 min. The pollen was cleaned using the acetolysis method reported by Erdtman [26]. The supernatant was eliminated and 5 mL of water was added, shaken, and centrifuged at 2500 rpm/min. Finally, the pelleted pollen was suspended in 1 mL of distilled water, then 10 µL of suspension containing the pollen was added to each compartment of a Neubauer chamber. The counting of purified pollen was carried out in five quadrants, as reported in [27,28].

To calculate the amount of pollen per gram of honey, the following equation was used:

X = A × 50,000/10

(1)

where:

• X = the number of pollen grains in each gram of honey.
• A = the arithmetic mean of pollen grains in the two compartments of the Neubauer chamber (N1 + N2/2).
• 50,000 = the constant for calculating the sample volume. The final units are pollen numbers × 10⁴ per gram of honey [27,28].

The samples were classified into five groups, as proposed by Louveaux et al. [28]: group I (<20,000 pollen grains); group II (20,000–100,000); group III (100,000–500,000); group IV (500,000–1,000,000); and group V (>1,000,000) [28].

2.4. Qualitative Pollen Analysis

To identify the frequency of pollen grains, we used 10 g of honey diluted in 20 mL of distilled water, not above 40 °C, then resuspended and divided it into two tubes of 50 mL of an equivalent volume (10 mL each) and centrifuged it for 10 min at 2500 rpm/min to draw off the supernatant liquid. The pellet was transferred to new tubes, centrifugated for 5 min, and the liquid phase was extracted. The tubes were placed upside down on filter paper to remove the supernatant until the sediment was as dry as possible. An acetolysis mixture (1 mL of sulfuric acid plus 9 mL of acetic anhydride) was carefully added to the pellet and resuspended. The samples were incubated in a 70 °C water bath for 10 min and then centrifuged for 5 min [26]. The supernadant was carefully removed and 5 mL of water was added, shaken vigorously, and centrifuged again (5 min). Finally, the supernatant was decanted and resuspended in 1 mL of a 1:1 glycerin–water solution. From the resuspended pellet, 10 µL was taken and added to a 24 × 24 mm slide and a coverslip was placed on top. Analysis was carried out with an optical microscope at 400× or 1000× magnification. For each sample, approximately 200 pollen grains were counted [28]. The analyzed pollen content samples were prepared on a slide and photographed under the Carl Zeiss microscope model Primo Star (S/N 3153000475) using Zeiss Zen blue edition software, 2.5, (Carl Zeiss, Jena, Germany) (Figure 2A–S). The identification of the pollen types was based on that reported by Alfaro-Bates et al. [29] and Ramos-Díaz, which was a catalog of the main pollen types found in honey from the Yucatan Peninsula [30].

The percentage frequency of the pollen taxa in all samples was calculated. The pollen type was allocated to one of four frequency classes: (I) predominant pollen (>45%); (II) secondary pollen types (16–45%); (III) important minor pollen (3–15%); and (IV) minor pollen (<3%). The honey samples were characterized as monofloral if they contained a predominant nectariferous pollen type (pollen of polleniferous plants was excluded). Otherwise, they were considered multifloral [28,29,31].

In addition, an ecological parameter, the Shannon-Weaver diversity index, was used to calculate the pollen diversity in each sample according to the following equation:

$$H=-{\displaystyle \sum }_i^{n}PilnPi$$

(2)

where H is the Shannon–Weaver diversity index, Pi represents the proportion of each pollen type i encountered in the sample, and ln denotes the natural logarithm. In the case of the Shannon-Weaver diversity index, the calculation was performed using a student’s t-test analysis by comparing the diversity index of the monofloral and multifloral honey sample groups [31].

2.5. Statistical Analyses

A multivariate analysis of data obtained from 22 honey samples was carried out using Statistical Analysis System (SAS) software. For this purpose, we used the number of plants obtained from the identification and classification of MF species per pollen sample, which were organized into percentages, represented by the sum of the different plants identified in the pollen samples.

First, a specialized form of principal components analysis (PCA) called multidimensional preference (MDPREF) analysis was conducted to explore the relationship between the locations and plant species present in the honey samples. The MDPREF analysis was carried out using the PRINQUAL procedure [18,32]. This analysis uses the transformation of linear and nonlinear variables with the alternate least-squares method, thus promoting the determination of the correlation or matrix covariance of the transformed variables. The MDPREF analysis is the result of a matrix with columns corresponding to ecosystems and rows corresponding to the pollen of the plant species identified (Table S1). The percentages of pollen with the identified pollen types in the different ecosystems, up to the species found in only one ecosystem. The percentages are the preferences of the ecosystem type for each pollen type [18]. The results are a graph of the two principal components (components 1 and 2), representing the ecosystem preferences for certain species in Campeche. The rates are presented as vectors emanating from the origin and pointing toward their most preferred priority.

3. Results

3.1. Pollen Qualification

Our research shows the absolute pollen count presented in units of 10⁴ pollen/g of honey (Figure 3). The samples were classified into five groups (see the Materials and Methods section): for group I, samples C060 and C082 (Figure 3); for group II, samples C029, C031, C033, C040, C043, C046, C045, C052, C058, C063, C066, C071, and C083 (Figure 3); for group III, samples C039, C040, C049, C054, C056, C077, and C080; and for groups IV and V, we did not find honey based on the absolute pollen count samples (Figure 3).

3.2. Flora Identification

We analyzed the pollen occurrence and its relationship with the species identified in the different ecosystems in the aforementioned Campeche municipalities. The pollen profiles in the 22 samples of honey were analyzed using melissopalynology and qualitative data was obtained. Floral identification analyses identified 19 plant genera corresponding to the Combretaceae, Poaceae, Malvaceae, Polygonaceae, Asteraceae, Amaranthaceae, Sapindaceae, Euphorbiaceae, Bignoniaceae, Fabaceae, Convolvulaceae, and Sapindaceae families.

In particular, our results show the percentages of the pollen types occurring in the honey samples botanically related to the 19 species identified within the seven different ecosystems analyzed: Albizia lebbeck (Alb), Albizia tomentosa (Albt), Amaranthus spinosus (Ama), Cassia fistula (Cas), Coccoloba reflexiflora (Coc), Croton icche (Cro), Guazuma ulmifolia (Gua), Gymnopodium floribundum (Gym), Milleria quinqueflora (Mil), Neomillspaughia emarginata (Neo), Operculina pinnatifida (Ope), Pithecellobium lanceolatum (Pit), Pseudobombax ellipticum (Pse), Senegalia gaumeri (Sen), Tabebuia rosea (Tab), Talisia floresii (Tal), Terminalia buceras (Buc), Thouinia paucidentata (Tho), and Zea mays (Zea) (Figure 4,

Table 2. Percentage of pollen occurrence per species in different ecosystems in Campeche, Mexico (abbreviation of scientific names (ABB) and their meaning in Spanish). Vegetation types: low-stature subtropical forests (SBSUB = selva baja subperennifolia, in Spanish); medium-stature tropical forests (SMSUP = selva mediana superennifolia); low-stature tropical forests (SBPER = selva baja perennifolia); medium deciduous forests (SMSU = selva mediana subcaducifolia); low deciduous forests (SBSU = selva baja caducifolia); low inundated tropical forests (SBBA = selva baja inundable); and secondary forests (SSSU = selva secundaria).

Taxa	ABB	SMSUP	SBBA	SBPER	SMSU	SBSU	SSSU	SBSUB
Albizia lebbeck (L.) Benth.	Alb	<10%	<10%
Albizia tomentosa (Micheli) Standl.	Albt		<10%			<10%
Amaranthus spinosus L.	Ama	<20%
Cassia fistula L.	Cas	<10%			<10%
Coccoloba reflexiflora Standl.	Coc			<10%
Croton icche Lundell	Cro						>60%
Guazuma ulmifolia Lam.	Gua	<10%			>40%	<10%	<10%
Gymnopodium floribundum Rolfe	Gym	<10%				>20%
Milleria quinqueflora L.	Mil	<10%	<20%		<10%	<20%	<10%	>60%
Neomillspaughia emarginata (H. Gross) S.F Blake	Neo	<10%
Operculina pinnatifida (Kunth) O’ Donell	Ope		<10%
Pithecellobium lanceolatum (Willd.) Benth.	Pit	<10%					<10%
Pseudobombax ellipticum (Kunth) Dugand	Pse	<10%				<10%
Senegalia gaumeri (SF Blake) Britton and Rose	Sen	<10%
Tabebuia rosea (Bertol.) DC.	Tab	<10%		<10%
Talisia floresii Standl.	Tal	<10%			<10%
Terminalia buceras (L.) C. Wright	Buc	~20%	~50%	>60%	>10%	~30%	~30%	-
Thouinia paucidentata Radlk.	Tho		<10%
Zea mays L.	Zea	>20%	~20%	>20%	>10%	<10%		>30%

).

3.3. Relationship between Flora Identified through Pollen Types and Ecosystem Sampling

To further evaluate the relationship between the original floral species and the ecosystems sampled, we analyzed the percentages of the ecosystem preferences for each pollen type (Figure 5). Thus, we produced a biplot representation of the ecosystem occurrence by the identified species, grouped into seven main components (ecosystems) (Figure 5A). The components show the eight values of the selected components, where the highest number is associated with greater information capacity. Figure 5C illustrates a proportion scale representing the pollen percentage and presents the main components on the X-axis, showing a 64.02% presence for component 1 and a 21.17% presence for component 2.

The added line represents an accumulated explanation for each component, explaining the 100% information variability (Figure 5A,B).

Multidimensional analysis of the main components revealed the relationship between the identified pollen types and the sampling municipalities. The sum of components 1 and 2 explained 85.19% of the information variability. We also observed that the Terminalia buceras pollen type was preferentially associated with the SBPER, SBSU, SBBA, and SSSU ecosystems (Figure 5C). The Zea mays and Milleria quinqueflora species were related in the upper part of the biplot and were mainly associated with the SBSUB ecosystem. In addition, the Albizia lebbeck, Guazuma ulmifolia, and Gymnopodium floribundum species were associated with the SBSUB ecosystem (Figure 5C). Notably, the Thouinia paucidentata, Tabebuia rosea, Amaranthus spinosus, Neomillspaughia emarginata, Coccoloba reflexiflora, Senegalia gaumeri, Albizia tomentosa, Pithecellobium lanceolatum, Operculina pinnatifida, Talisia floresii, Cassia fistula, and Croton icche pollens were found grouped behind the vectors in the PCA analysis, suggesting that the species could be preserved in the seven ecosystems sampled (Figure 5C).

3.4. Honey Classification

Furthermore, for the mono- or multifloral honey classification, we discovered that Milleria quinqueflora was present in sample C029 (67%); Gymnopodium floribundum was present in C031 (48%), C060 (72%), and C080 (85%); Terminalia buceras was found in samples C033 (49%), C045 (75%), C049 (80%), C063 (58%), C075 (60%), and C083 (80%); Amaranthus spinosus appeared in C039 (77%) and C043 (80%); Talisia floresii was present in C052 (70%); Guazuma ulmifolia was found in C054 (89%), C056 (86%), and C066 (84%); Zea mays appeared in C082 (50%); and Croton icche was discovered in C077 (64%) (Figure 6).

The C046 sample presented Milleria quinqueflora (31%) and Terminalia buceras (43%) pollen; C040 showed Milleria quinqueflora (42%), Zea mays (40%), Tabebuia rosea (10%), and Albizia lebbeck (8%); C071 displayed Zea mays (55%), Terminalia buceras (35%), and Guazuma ulmifolia (4%); and C058 presented Milleria quinqueflora (20%), Terminalia buceras (39%), and Zea mays (1%). These samples showed an absence of a predominant pollen type, classifying them as multifloral honeys (Figure 6).

To determine the grade of the diversity present in the honey samples analyzed, we calculated the Shannon–Weaver diversity index. For the group of monofloral samples found, the index ranged from 0.27 to 1.13 (

Table 3. Pollen analytical data of honey samples from north-central Campeche, Mexico, showing samples, absolute pollen counts (APC)/1 g honey, classifications by APC (as described by Louveaux et al.), Shannon–Weaver diversity indexes, nature of honeys, and floral sources—the type of floral source is indicated with an “x” for each type: nectariferous (N) or polleniferous (P). The student’s t-test shows the statistical significance of the Shannon–Weaver diversity index comparing the monofloral and multifloral group samples (* p < 0.05).

Sample	APC/1 g Honey	Class (Louveaux et al. [28])	Shannon–Weaver Diversity Index	Nature of Honey	Floral Source: Nectariferous (N) or Polleniferous (P)
Monofloral					N	P
C029	43,500	II	0.27	Milleria quinqueflora (67%)	x	x
				Zea mays (33%)		x
C031	50,500	II	0.36	Neomillspaughia emarginata (45%)	x	x
				Gymnopodium floribundum (48%)	x	x
				Amaranthus spinosus (4%)	x
				Albizia lebbeck (3%)	x	x
C033	52,500	II	0.83	Termilaria buceras (49%)	x	x
				Zea mays (47%)		x
				Coccoloba reflexiflora (4%)	x	x
C039	121,500	III	0.67	Amaranthus spinosus (77%)	x
				Zea mays (19%)		x
				Cassia fistula (2%)	x	x
				Milleria quinqueflora (2%)	x	x
C043	46,000	II	0.28	Amaranthus spinosus (80%)	x
				Zea mays (10%)		x
				Tabebuia rosea (10%)	x	x
C045	69,000	II	0.92 *	Termilaria buceras (75%)	x	x
				Zea mays (4%)		x
				Albizia lebbeck (5%)	x	x
				Gymnopodium floribundum (1%)	x	x
C049	239,000		0.50	Termilaria buceras (80%)	x	x
				Tabebuia rosea (20%)	x	x
C052	44,500	II	0.94	Talisia floresii (70%)	x	x
				Pseudobombax ellipticum (12%)	x	x
				Zea mays (10%)		x
				Albizia lebbeck (8%)	x	x
C054	436,000	III	0.40	Guazuma ulmifolia (89%)	x	x
				Thouinia paucidentata (9%)	x	x
				Talisia floresii (2%)	x	x
C056	189,000	III	0.46	Guazuma ulmifolia (86%)	x	x
				Termilaria buceras (12%)	x	x
C060	18,500	I	0.83	Gymnopodium floribundum (72%)	x	x
				Milleria quinqueflora (19%)	x	x
				Termilaria buceras (4%)	x	x
				Albizia lebbeck (3%)	x	x
				Albizia tomentosa (1%)	x	x
C063	32,000	II	1.13	Termilaria buceras (58%)	x	x
				Guazuma ulmifolia (19%)	x	x
				Milleria quinqueflora (13%)	x	x
				Zea mays (10%)		x
C066	63,000	II	0.44	Guazuma ulmifolia (84%)	x	x
				Zea mays (16%)		x
C075	17,000	I	0.90	Termilaria buceras (60%)	x	x
				Zea mays (30%)		x
				Thouinia paucidentata (10%)	x	x
C077	191,900	III	0.82	Croton icche (64%)	x	x
				Termilaria buceras (32%)	x	x
				Milleria quinqueflora (2%)	x	x
				Pithecellobium lanceolatum (1%)	x	x
C080	155,500	III	0.55	Gymnopodium floribundum (85%)	x	x
				Guazuma ulmifolia (10%)	x	x
				Zea mays (3%)		x
				Milleria quinqueflora (2%)	x	x
C082	16,500	I	0.83	Zea mays (50%)		x
				Termilaria buceras (46%)	x	x
				Milleria quinqueflora (4%)	x	x
C083	47,000	II	0.64	Zea mays (50%)		x
				Termilaria buceras (46%)	x	x
				Milleria quinqueflora (4%)	x	x
Multifloral
C040	157,000	III	1.16 *	Milleria quinqueflora (42%)	x	x
				Zea mays (40%)		x
				Tabebuia rosea (10%)	x	x
				Albizia lebbeck (8%)	x	x
C046	69,500	II	0.97 *	Termilaria buceras (43%)	x	x
				Milleria quinqueflora (31%)	x	x
				Zea mays (11%)		x
C058	60,500	II	1.10 *	Pseudobombax ellipticum (40%)	x	x
				Termilaria buceras (39%)	x	x
				Milleria quinqueflora (20%)	x	x
C071	47,500	II	0.99 *	Zea mays (55%)		x
				Termilaria buceras (35%)	x	x
				Pithecellobium lanceolatum (6%)	x	x
				Guazuma ulmifolia (4%)	x	x

), and for the samples classified in the multifloral group, we calculated an index of 0.97 to 1.10, which shows the differences between the groups. This was validated using a student’s t-test and a statistical significance was found (* p < 0.05) in a comparison of the diversity index of the monofloral and multifloral samples (Table 3).

Finally, Table 3 presents the information obtained in the study, which includes the Shannon–Weaver diversity index, the flora identification, and percent frequency for each sample, as well as the type of floral source: nectariferous (N) or polleniferous (P) or both. The data show the relationship between the absolute pollen counts (APC) and the Louveaux et al. [28] classifications. The types of pollen were allocated to three of five groups: group I (<20,000 pollen grains); group II (20,000–100,000); and group III (100,000–500,000) [28], which were grouped into monofloral or multifloral honeys based on the presence of a predominant pollen (>45%). This coincides with the diversity index, showing more monofloral samples close to zero than multifloral samples, which was corroborated by the student’s t-test (Table 3).

4. Discussion

Mexico is a honey-producing country. Its geographical location allows the development of honey varieties within different ecosystems. In particular, the YP is an area with a large concentration of melliferous flora, favoring beekeeping development. Campeche is part of this area located in southeast Mexico and is recognized for producing high-quality honey, which is exported to other countries [4]. From an environmental point of view, bees provide ecological services, such as pollination, and promote the biodiversity of Mexican ecosystems. Thus, bees and other pollinators need major resources for food and materials to build hives. Melliferous flora supplies resin, pollen, and nectar to bees, who provide pollination services to promote flower blossoming and fruit formation. The latter is a source of food for humans and animals, thus contributing to the equilibrium of ecosystems [11].

In the first instance, we analyzed the absolute pollen count determinant: the quantity ranged from 18,500 to 436,000. Our results coincide with the report by Song et al. [31], which described the honey as classified into groups I, II, and III, suggesting that the honey was of high quality. Following the Louveaux et al. [28] classifications, these results indicate that the local flora could be a high-quality honey source [28,31].

In Mexico, there are reports of 23,314 vascular plants conforming to 73 orders, 2854 genera, and 297 families, including 149 gymnosperms and 22,126 angiosperms. Other regions of Mexico, such as Veracruz, show 271 families, 1956 genera, and 8497 species [12]. The YP contains 2329 taxons, 161 families, and 956 genera as native and wild plants. Despite the great biodiversity present in the YP, only a select group of plants are used by bees, mainly Viguiera dentata and Gymnopodium floribundum [33]. The literature confirms that the flora families present in the ecosystems of Campeche state include Combretaceae, Poaceae, Malvaceae, Polygonaceae, Asteraceae, Amaranthaceae, Sapindaceae, Euphorbiaceae, Polygonaceae, Bignoniaceae, Fabaceae, Convolvulaceae, and Sapindaceae. Our floral identification study agrees with the species described as MF in Campeche, México [33].

Furthermore, this study also analyzed the relationship between the floral origin of the honey and the seven different ecosystems sampled (Table 1). The geographical location and the geological evolution of the landscapes in Campeche favor the biodiversity of ecosystems at the regional and local levels [34]. In terrestrial environments, the high forests stand out. Representative trees of this type of vegetation are chicle (Manilkara zapota), mahogany (Swetenia macrophylla), pukte’ (Termilaria buceras), and ramón (Brosimum alicastrum). For the medium deciduous and subdeciduous forests, the vegetation reported comprises pich (Enterolobium cyclocarpun), chechem negro (Metopium brownei), ceiba (Ceiba pentandra), and ya’axnik (Vitex gaumeri). The evergreen and subevergreen lowland forests present the characteristic trees of these forests, which are the red (Haematoxylum campechianum), chooch kitam (Hyperbaena winzerlingii), boob chi’ich’ (Coccoloba cozumelensis), sak cheechem (Cameraria latifolia), and satj’iitsa (Neomillspaughia emarginata). The subdeciduous forests contain trees such as the chak ch’ooy (Cochlospermum vitifolium), chak kiis (Gyrocarpus americana), silil (Diospyrus cuneata), pixoy (Guazuma ulmifolia), and chak kuy che’ (Pseudobombax ellipticum) [29]. These ecosystems preserve plant and organism communities and extend from the south to the north-central zones of Campeche state [29,35].

In the central region of the Champotón and Hopelchén municipalities, low-stature tropical and subtropical forests are present in the north, and there are low deciduous and subdeciduous forests gradually extending from the north of Champotón municipality towards Yucatán [34,35]. Regarding plant species biodiversity, more than 145 identified plant families were reported, with Fabaceae, Poaceae, Orchidaceae, Asteraceae, Euphorbiaceae, and Bromeliaceae being the most relevant. Other families described in the YP were the Annonaceae, Boraginaceae, Cactaceae, Cyperaceae, Heliconaceae, Icanaceae, Orchidaceae, and Olygonaceae families. In addition, the species that were reported as being nectariferous or polleniferous or both as floral sources in the YP, Viguiera dentata, Bursera simaruba, Thouinia paucidentata, Mimosa pudica, Termilaria buceras, Sabal yapa, and Mimosa pigra, were classified as monofloral honey sources. Other species described—Lysiloma latisiliquum, Convolvulaceae, and Trema micrantha—were reported as being multifloral, and Gymnopodium floribundum, Piscidia piscipula, Acacia gaumeri, Croton fragilis, Trixis inula, Eugenia sp., Metopium brownie, and Talasia oliviformis have been associated with being a source of monofloral/multifloral honeys [29,31].

Interestingly, the species Termilaria buceras, Zea mays, and Guazuma ulmifolia were related to the SBSU and SBSUB ecosystems. In the case of Termilaria buceras, our results are consistent with the reported growth in low- and medium-stature tropical forests and low-inundated tropical forests [33,34]. Guazuma ulmifolia was reported to grow in low- and medium subdeciduous flooded forests and in secondary vegetation. Therefore, this explains and confirms its frequency in the C054, C056, C066, C075, and C039 samples [35,36].

These results coincide with previous reports describing that the honey production from Campeche apiaries comes mainly from two species: 42% from Viguiera dentata Spreng var. dentata (Asteraceae) and 48% from Gymnopodium floribundum (Polygonaceae), whereas 10% comes from legumes (Fabaceae) and vines (Sapindaceae and Convolvulaceae) [8,29]. Thouinia paucidentata, Tabebuia rosea, Amaranthus spinosus, Neomillspaughia emarginata, Coccoloba reflexiflora, Senegalia gaumeri, Albizia tomentosa, Pithecellobium lanceolatum, Operculina pinnatifida, Talisia floresii, Cassia fistula, and Croton icche were the identified species located in the YP and grown in the ecosystems situated in the north-central zone of Campeche [30,37]. In addition, these species are important alternative food sources visited by bees during food crises to obtain nectar and pollen [29].

Interestingly, our results coincide with the Castillo-Cázares et al. [37] report that described the botanical composition of honey from the YP using molecular methods such as qPCR (quantitative polymerase chain reaction) [37]. This research determined that the main species of beekeeping were Viguiera dentata, Gymonopodium floribundum, Piscidia piscipula, Acacia angustissima, and Mimosa bahamensis. The frequency of V. dentata (tajonal), G. floribundum (t’sit’silche’), and M. bahamensis (sak káatzim) was mainly present in the analyzed samples [37].

Granados-Argüello et al. [38] described that the genuses Bursera spp. (Combretaceae) and Tillandsia spp. (Bromeliaceae) were reported with a high proportion in honey of Chiapas, Tabasco, Oaxaca, Veracruz, and the YP. In addition, our results of the minor secondary pollen types identified as Operculina pinnatifida (Convolvulaceae), Zea mays (Poaceae), and Croton icche (Euphorbiaceae), coincide with a report that the secondary vegetation families of Convolvulaceae, Poaceae, and Euphorbiaceae were found in the honeys of Veracruz, Mexico [38].

In particular, Zea mays was found in some samples analyzed. Thus, this is the first report that describes the presence of predominant pollen crops in honey. It should be noted that Campeche has a farming area of 218,671 ha, 80% of which is dedicated to producing white corn. Remarkably, from 2007 to 2011, the northern municipalities of Calkiní, Hecelchakán, Tenabo, and Champotón produced 12,640, 42,176, 15,782, and 16,218 tons of corn, respectively [39]. Currently, Campeche has 14,574 ha of corn, yielding 31,719 tons [40], which could explain the presence of Zea mays pollen in the C033, C040, C071, and C082 samples that corresponded to Hopelchén, Calkiní, Hecelchakán, and Champotón. This finding may demonstrate that changing the soil used in the forests has a negative environmental impact, causing the fragmentation of habitats and diminishing soil fertility. Loss of biodiversity contributes to a reduction in environmental services; thus, implementing sustainable food production techniques is needed [41]. In Mexico, the Poaceae family has been found in honey from Chiapas, Veracruz, Tabasco, and Valle de Mexico. Zea mays is classified as a polleniferous species that contributes a great quantity of pollen to honey as a source of nutrients and food for bees [42,43].

The association between the pollen and ecosystem data was confirmed by PCA (PROC PRINQUAL and MDPREF), which associated the identified pollen with the distinct ecosystems of the municipalities of Campeche. PCA showed that pollen was identified at <1% due to the great diversity of MF in Campeche. Pollens were grouped behind the vectors, which suggests a negative association. Alternatively, the flora was perhaps conserved in the different ecosystems across the sampled area [18]. Jacinto-Pimienta et al. [18] described the PROC PRINQUAL and MDPREF analyses applied to Apis mellifera L. honey samples. The study consisted of 38 samples from six municipalities corresponding to five sub-regions in Tabasco state (Mexico). They reported the presence of 22 multifloral, 9 monofloral, and 7 bifloral honeys. Remarkably, the authors found that component 1 explained 38.59% and component 2 showed 26.45% of variability, which showed a high MF conservation in the sampled apiaries regarding the zone and floral origin. The PROC PRINQUAL technique allowed for the discovery of a pollen preference in the sampling zone through PCA of qualitative or quantitative data or both. The MDPREF analysis was based on a matrix where the rows (pollens) and columns (municipalities) built a components graphic that relates the pollen samples to the municipality preferences (Table S1) [18].

In another study, Homrani et al. [14] reported the study of 62 samples of honey taken in Algeria from different biogeographical areas (semiarid, sub-humid, and arid zones). In contrast, the PCA analysis showed statistical significance in the floral resources in ecosystems by the effect of the geographical area. This means that significant floral variability produces different types of honey [14]. In summary, our results are in agreement with other studies that show that the type of ecosystem in a specific geographical area determines the flora and, thus, the type of honey produced.

A clear example of the impact of the flora and the environment on honey composition in a geographical region was described by Feás et al. [44], who analyzed 31 samples of honey in the Entre-Douro and Minho regions of Portugal. They showed that five samples were listed as Eucalyptus honey, one as citrus honey, and twenty-five as multifloral honey [44].

Finally, our honey classification results were based on the percentage frequency of the taxa in all of the honey; monofloral honey was considered if the honey contained a predominant pollen type, otherwise, it was considered multifloral, as reported in other publications [8,14,28,31,36,44]. Our results showed a major abundance of monofloral honey (n = 18) compared with multifloral honey (n = 4), suggesting that north and central Campeche during the February–August period present high concentrations of floral resources for bees [8,11,45]. In this work, the classification was based on the percentage of pollen type present in the samples. However, this does not mean that the bees only visited these resources to create the honey. Honey is produced from the nectar of plants, from the secretions of the living parts of plant parts, or from the excretions of plant-sucking insects. Some of the components (carbohydrates, water, enzymes, traces of organic acids, amino acids, pigments, and wax) are added by bees, some are derived from plants, and some are due to the maturation of the honey. Some analytical methods are suitable for indicating the botanical and geographical origins based on the analysis of specific components or multi-component analysis, for example, aroma compounds, flavonoids patterns, profiles of amino acids, oligosaccharides, and trace elements, among others [28,46,47].

Villanueva-Gutierrez et al. [8] described monofloral honeys in the YP with >45% of the predominant pollen, highlighting the presence of the Viguiera dentata, Bursera simaruba, Piscidia piscipula, Eugenia sp., Pimenta dioica, Melothria pendula, Gymnanthes lucida, Phoradendron quadrangulare, Gymnopodium floribundum, Rubiaceae, Thouinia paucidentata, Thouinia sp., and Pouteria mammosa species. Our report is in agreement with this study describing the existence of many monofloral honeys in the YP [8].

Alfaro-Bates et al. [29] described how a beekeeper assigns the local name of the species whose flowering dominates the landscape to the honey. This is related to the entry of nectar into the colonies, but sometimes the botanical origin does not correspond to the said species. Honey characteristics in the YP are the result of mixtures of multiple nectar flower sources, such as Bucida-Diphysa-Poaceae, M. pudica-Heliocarpus-Cecropia, Chrysophyllum-Bucida-Conocarpus, Diphysa-Conocarpus-Bursera, Piscidia-Croton-Bursera, Bursera-Celtis-Talisia, and Croton-G. floribundum-Bucida. A highlight is the family Convolvulacea, identified by its Mayan names, xtabentun (Turbina corymbosa), tso’ots kàab (Merremiaaegyptia), ak’il xíiw (Jacquemontia pentantha), or tsalam (Lysilomalatisiliquum) [5,19,29,36,38].

However, Convolvulaceae pollen is typically under-represented, being in the category of minority or trace (<3%) and is rarely considered. Another critical aspect in the classification of honey is the type of floral resource: nectariferous or nectariferous-polleniferous [5,8,12,29]. In the samples, the identified species described as nectariferous-polleniferous are Termilaria buceras, Guazuma ulmifolia, Gymnopodium floribundum, Milleria quinqueflora, Talisia floresii, Croton icche, Pseudobombax ellipticum, Neomillspaughia emarginata, Tabebuia rosea, Operculina pinnatifida, Albizia tomentosa, Thouinia paucidentata, Pithecellobium lanceolatum, Senegalia gaumeri, Cassia fistula, and Coccoloba reflexiflora. Amaranthus spinosus is described as nectariferous and Zea mays is classified as a polleniferous plant (Table 3 and Table S2).

A report in Mexico by Villanueva-Gutierrez [48] described the families in the YP that contributed to the pollen species presence in honeys as Poaceae, Asteraceae, Boraginaceae, Euphorbiaceae, Fabaceae, Sapindaceae, Convolvulaceae, Myrtaceae, Tiliaceae, and Sapotaceae, and highlighted that the Fabaceae and Asteraceae families were very important in the diet of the bees [48].

This highlights the importance of studying the botanical origin in honeys using melissopalynological techniques [14,28,46,47]. In contrast to the commercially valued monofloral honeys and from an ecological point of view, multifloral honeys should be more valued as they are richer in different pollens and nectars. Multifloral honeys demonstrate that the bees visited different plants, thus promoting biodiversity conservation.

In this work, the honey samples were classified as monofloral or multifloral according to mellisopalynological analysis. The samples C040 and C071 were classified as multifloral, even when we found the predominant pollen Zea mays because the nectar is the base that constitutes the honey; instead, pollen is essential to the diet of the bees [8]. The bees frequently collect a wide variety of pollen types but generally concentrate on a few species. Nevertheless, monofloral or multifloral honey classifications are still considered for classifying floral resources as nectar, pollen, or both [8].

The Shannon–Weaver diversity index shows the high diversity of pollen types in the four multifloral honeys with a range of 0.93 (sample C052) to 1.16 (sample C040). The values in the samples indicate the rich nectar and pollen sources in the Calkiní and Tenabo municipalities, whereas comparatively, the monofloral honeys have lower diversity index values ranging from 0.27 (sample C029) to 1.12 (simple C063), which suggests that the diversity of pollen types changed in the honey samples collected from different localities in February–August 2021. These results agree with those in a study by Song et al. on 19 honey samples from Shanxi, China, which had high Shannon-Weaver index values of 1.79–2.21 for the multifloral samples, suggesting rich nectar and pollen sources [31]. Furthermore, Shukla et al. [48] described high Shannon-Weaver diversity index values in honey samples from India, indicating high plant heterogeneity foraged by the bees [49]. Our diversity index calculations show a statistical significance between the monofloral and multifloral samples, indicating that diversity is prominent in the multifloral honeys [31].

Our study provides new insights into honey pollen composition from north-central Campeche, Mexico. In summary, we described the monofloral and multifloral honey classifications of the samples using the percentage frequency, absolute pollen count, and diversity index. The differences between the monofloral and multifloral honeys were complemented by the identified nectariferous or polleniferous floral resource, suggesting that during the sampling period (February–August), the diversity of the floral resources available for bees is associated with the different ecosystems in the study zone.

The present analysis shows the potential for beekeeping activities in Campeche, which has high floral diversity that honeybees exploit to produce honey. Eighteen different monofloral honey types were produced due to the abundant native plant resources from low and medium deciduous forests. The information obtained in this work can contribute to improving beekeeping performance as beekeepers need to maintain local native plants and reforest disturbed sites using native plants that are now known to be honeybee plants.

The determinations of the botanical origins of the samples suggested a particular flora and complemented the classification practice based on beekeeper knowledge. These aspects must be considered and communicated to beekeepers to assist in caring for their hives. This information could also be used to develop conservation strategies for maintaining plants and reforesting disturbed sites using native plants that are now known to be melliferous, thus promoting ecosystem conservation. Another aspect that should be studied in depth is the impact of crops on the production of honey since the bees were found to pollinate corn. This is of interest as this crop could be analyzed and validated as a marker of damage to the forests or jungles and could help in understanding the effect of the crops in the production of honey, as well as the environmental factors in Campeche.

5. Conclusions

The results of this paper revealed the great variability in the local honey production of the Campeche beekeeping industry and the potential of the different honey types that could be obtained due to native melliferous flora and traditional crop production. Our research highlights the critical role of the diversity of local honeybee plants and traditional crops for beekeeping in southern Mexico. This information could be useful for designing strategies for nature conservation and agroecosystem management in the production of honey in Campeche, Mexico.