Table 1.
Characterization of food consumption studies assessing biodiverse foods.
| Study number | Article data | Research setting and number of participants | Research design | Objective | Biodiverse foods assessed | Methods applied to assess biodiversity | Indicators | Main results | Main limitations reported | Quality |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Blundo-Canto et al. (19) | Ucayali, Peru, n = 53 households in 4 rural communities | Observational, longitudinal | Test the hypothesis that the expansion of commercial crops is associated with deforestation, reduced agrobiodiversity, and food access changes | All foods, including bushmeat and species of local plants | Use of one 24-h recall in 2000 and another in 2015. Besides 24-h recall in 2015, the analysis combined in-depth interviews and focus groups with key informants in order to evaluate consumption changes | Household dietary diversity score (HDDS) | Regarding food consumption, the authors observed a reduction in the household food diversity score (HDDS) of 1.3 food groups when comparing data from 2000 and 2015. In addition, the variety of food items consumed also decreased from 69 food items mentioned in 2000 to 35 in 2015 | Researchers used respondents' recall and perception to build their conclusions instead of empirical quantification | Strong |
| 2 | Boedecker et al. (20) | Benin, n = 120 households, non-pregnant and non-lactating women (>18 y), rural community | Observational, cross-sectional | To evaluate the contribution of wild food plants within the diets of women living in the buffer zone around Lama forest, southern Benin | Wild edible plants | The authors evaluated the consumption of wild plants with two non-consecutive 24-h recall. They conducted the taxonomic identification of some species in the field. The species they were unable to identify were collected (dry specimens and photos) and taken to the herbarium | Women Dietary Diversity Score (WDDS) and Nutrient Adequacy Ratio (NAR, considering EAR) | The contribution of WEPs to total dietary intake was low due to infrequent use and small portion sizes. The highest nutrient contributions of WEPs measured were for copper (13.9%) and iron (4.6%). Women's dietary diversity was significantly higher among WEP consumers than non-consumers, mainly due to higher consumption of dark green leafy vegetables | Some plants cited by participants were not available due to seasonality, while others were inaccessible due to legal restrictions in the forest | Strong |
| 3 | Broegaard et al. (21) | Laos, n = 33 households in 3 rural communities | Observational, cross-sectional | Examine the role of agricultural and forest landscapes in providing wild foods in farming communities in northern Laos | Wild foods characterized as sources of protein in the local context, e.g., rice, rat, and fish | Use of field collection diaries in four agricultural cycles (i.e., slashing and burning, planting, weeding, and harvesting) to ensure variation in different seasons. Research assistants identified all hunted animals to the taxonomic group rather than at the species level. Researchers conducted interviews and participant observation, but there is no mention of classic dietary assessment methods | Proportion of protein from wild food in relation to the per capita value of 50 g of protein per day | Increased land use pressure alters the cultivation landscape and, consequently, the quality of the diet, with adverse effects on protein intake. Wild plants contributed much more to a diversified diet in villages with traditional crop farming systems and much less in villages dominated by commercial crop production | Participants may have (intentionally) failed to report or display all collected products, especially the illegal ones. Authors report that it was difficult to estimate protein contributions from wild foods due to lack of composition data | Moderate |
| 4 | Chyne et al. (22) | Meghalaya, India, n = 160 households in 15 rural communities comprising n = 510 (women and children <5 y) | Observational, cross-sectional | Examine whether the prevalence of malnutrition and chronic disease among the Khasis of North East India is associated with available food biodiversity | All food sources | The authors applied a 24-h recall to assess consumption. To evaluate the knowledge about food plants (cultivated or wild), they conducted free listing and focus groups. However, researchers did not identify the taxonomy of the species consumed; they just recorded local names | Average consumption of nutrients compared to Recommended Dietary Allowances (RDA) and Species consumed by food groups (number without scientific identification) | Nutrient intake was below recommended levels. The prevalence of anemia in children aged 1–5 was 68%, and vitamin A deficiency was 59%, and among women, the same number ranged between 83 and 48%, respectively. Through focus groups, villages reported using 372 food species, most of which were wild (225). According to the authors, malnutrition is unacceptably high among the Khasis, considering the richness of food biodiversity | Due to the lack of nutrient data for many local foods, the numbers of nutrient intake are just estimations. Furthermore, the authors did not perform taxonomic identifications for more than two-thirds of the food species | Strong |
| 5 | Conti et al. (15) | Arusha, Tanzania, n = 141 households in urban and peri-urban areas, women of reproductive age (15–4 9y) | Observational, cross-sectional | Explore aspects that contribute to micronutrient adequacy in women of reproductive age in Tanzania, focusing on indigenous vegetable consumption and other sociodemographic factors | Indigenous vegetables, i.e., neglected or underutilized plants | Use of a 24-h recall applied by trained interviewers, supported by a photo guide to assess portion sizes. Researchers used a list of edible resources elaborated in a previous study to identify the vegetables mentioned in the 24-h recall | Minimum dietary diversity index for women (MDD-W) | Results showed that sixteen percent of the women consumed at least one serving of indigenous vegetables per day. Indigenous vegetable consumption was positively associated with micronutrient adequacy | In statistical terms, the women interviewed in the survey did not represent the Arusha region and Tanzania, considering demographic, social, educational, and economic variables. In addition, researchers did not assess rural areas and the data gathering just covered the dry period | Strong |
| 6 | Da Silva and Begossi (23) | Amazon, Brazil, n = 114 households in the urban area of Barcelos and in 3 rural communities | Observational, cross-sectional with a longitudinal component to assess seasonality in food consumption | Describe the food consumption of riverine populations, comparing the composition, origin, diversification, and seasonal variations in diets of different communities | Animals: fish, birds, and mammals | Researchers performed data collection both in dry and rainy seasons, with the support of the 24-h recall and participant observation. The research team identified the taxonomy of fish samples by comparison in collections of research institutes, birds and mammals, by using field guides; and, finally, experts reviewed mammals identification | Shannon-Wiener Diversity Index (H′) and Species Richness | About 80 species of consumed fish were collected and identified in the studied populations. Fish account for 70% of the protein in the main meal, being unusual the consumption of beef. Families in the rural setting, and with low-income, had lower species richness in their diets. Industrialized and imported animal protein products (e.g., sausage, beef, frozen chicken, dairy products, powdered milk) account for 134 items consumed among urban households vs. 13 items among rural households. The consumption of imported items increased species richness, both in urban and rural areas | There is no report | Moderate |
| 7 | Fungo et al. (24) | Cameroon, n = 279 households in 12 rural communities, women | Observational, cross-sectional | Determine the contribution of forest foods to diets and estimate their association with household food insecurity | Forest plants | Use of two 24-h recall with an interval of 1 week. Researchers conducted focus groups in order to build a list of forest plants consumed by the population. Plant specimens were photographed, collected, and taken to the herbarium to be identified | Household dietary diversity score (HDDS), Food variety score (FVS), Forest food consumption score (FFCS), Household food insecurity access scale (HFIAS), and contribution to nutrient intake (NAR/EAR) | Researchers identified 47 forest plants. Of these plants, 17 were consumed by 98% of respondents over a week. Forest foods contributed approximately half of women's total daily energy intake in the case of women. The most significant contributions were for intakes of vitamin A (93%), Na (100%), Fe (85%), Zn (88%), and Ca (89%). Despite the high biodiversity, most families (83%) suffered from high food insecurity. Results shown that forest foods play an essential role in ensuring food security in these forest-dependent communities | There is no report | Strong |
| 8 | Ghosh-Jerath et al. (25) | Gumla, Jharkhand, India, n = 143 households in 4 rural tribal communities, women of reproductive age | Observational, cross-sectional with a longitudinal component to assess seasonality in food consumption | Assess the availability and consumption of indigenous food plants and the nutritional status of women from the Oraon tribal community in Jharkhand, India | Indigenous food plants, i.e., those obtained locally through cultivation or collection. | Use of 24-h recall on two consecutive days in winter, summer, and monsoon season. In winter, researchers also applied the FFQ. FFQ items, which included conventional and indigenous foods, were identified in focus group discussions. The authors mention in their acknowledgments that they developed the taxonomic classification of species samples | Nutrient adequacy ratio (NAR, RDA) and Household food security score | Food security and consumption of indigenous foods were low. Although the community reported knowing various indigenous foods (244 food items), the study revealed that the regular consumption of these items was insufficient. Higher intake of essential micronutrients, calcium, and iron, was observed among those who consumed indigenous foods. About 40% of women had degrees of chronic energy deficiency | Researchers calculate energy requirements by considering moderate physical activity levels for women. However, they believe that some women could have higher physical activity levels because activities such as cutting firewood, working in brickyards, herding cattle are common in these communities. These underestimations of energy may explain the higher prevalence of chronic energy deficiency observed, despite a caloric intake close to 80% of recommended levels. In addition, 24-h recall of two consecutive days was conducted in only one-third of the study sample | Strong |
| 9 | Ghosh-Jerath et al. (26) | Gumla, Jharkhand, India, n = 151 households in 4 rural tribal communities, women of reproductive age | Observational, cross-sectional with a longitudinal component to assess seasonality in food consumption | Explore the contribution of indigenous foods to nutritional status and nutrient intake | Indigenous food plants, i.e., those obtained locally through cultivation or collection | Use of 24-h recall on two consecutive days during the rainy season reproduced in winter and summer; FFQ was applied in the rainy season. Researchers conducted qualitative surveys (e.g., key informant interviews and focus group discussions) to capture the variety of foods consumed by the community. The researchers classified these species by their taxonomy | Nutrient adequacy ratio (NAR, RDA) and Household food security scale | The communities reported knowing a wide variety of indigenous food plants even though they did not consume them. Women consumed adequate energy and protein, but micronutrient intake was inadequate. For example, consumption of Ca, Fe, vitamin B2, folate, and vitamin B12 was insufficient in more than half of the participants. Women who consumed indigenous foods during the survey period had significantly higher intakes of Ca and Fe when compared to those who did not consume | The 24-h recall for two consecutive days may overestimate the prevalence of inadequacy of micronutrients in the population. Researchers did not analyze the composition of three foods items | Strong |
| 10 | Ghosh-Jerath et al. (27) | Gumla, Jharkhand, India, n = 204 households in 18 rural tribal communities, women of reproductive age | Observational, cross-sectional with a longitudinal component to assess seasonality in food consumption | Explore the association between production and access to agroforestry foods and consumption of indigenous plants with nutrient adequacy in the Sauria Paharia community | Indigenous food plants, i.e., those obtained locally through cultivation or collection | Use of 24-h recall on two consecutive days during the rainy season, repeated in winter and summer; FFQ was applied in the rainy season. Researchers conducted qualitative surveys (e.g., key informant interviews and focus group discussions) to capture the variety of foods consumed by the community. The researchers classified these species by their taxonomy | Food Accessed Diversity Index (FADI) and Median minimum dietary diversity score for women (MDD-W) | Access to agroforestry diversity was low (low FADI), despite the extensive knowledge of people about local plants. Women with the highest dietary diversity score (MDD-W) had higher intakes of energy, protein, fat, iron, calcium, zinc, B vitamins, vitamin A, and vitamin C. In addition, consumers of indigenous plants had higher intakes of calcium and vitamin A | Researchers calculated nutrient intake using software that provides the nutritional value of raw Indian foods. The FFQ was too long (300 items), which could lead to reporting and recall bias | Strong |
| 11 | Golden et al. (28) | Madagascar, n = 719 individuals (0–73 y), 152 households, in 2 rural communities | Observational, longitudinal | To characterize the consumption patterns of the Malagasy, living in remote rainforest areas in northeastern Madagascar | All food sources, including bushmeat and native plants | Weighed Food Record of three meals a day during nine consecutive months. Considering the information provided in the paper, researchers did not classify species by their taxonomy | Mean Adequacy Ratio (MAR, EAR), Household dietary diversity score (HDDS), Food consumption score (FCS), and Minimum dietary diversity for women (MDD-W) | Although the HDDS and FCS reflect the high diversity of the diet, the MDD-W indicator suggests poor micronutrient adequacy. For example, the median individual consumed <50% of their average requirement (EAR) for Ca and vitamins A, B12, D, and E; and <100% of their EAR for energy, riboflavin, folate, and Na | Lack of adequate nutrient composition data for many locally endemic foods, driving the authors to use established proxies that may not accurately represent the nutrient content of local foods | Moderate |
| 12 | Jones (29) | Malawi, n = 3,000 households of agricultural families, national | Observational, longitudinal | Determine the association of cultivated species richness with the diversity and quality of family diets in Malawi and assess the hypothetical mechanisms for this association through livelihood and market-oriented pathways | Cultivated plants, including market and subsistence crops | 7-day recall. Considering the information provided in the paper, researchers did not classify species by their taxonomy | Crop species richness (CSR) and Diet diversity score (DDS) | Agricultural biodiversity was a key determinant of the diversity and quality of the diets of farming families in Malawi. Neither the proportion of the harvest sold nor the distance to the nearest population center changed the relationship between CSR and DDS. Families with higher CSR were more commercially oriented | The DDS calculation was based on 7-day recall data at the household level. Given the extended recall period and data aggregation at the household level, the DDS was high for most households. Although the DDS was correlated with energy and nutrient intake, a more discerning indicator based on individual 24-h recall data would likely serve as a better metric. Second, food composition tables may be limited in their ability to accurately quantify the nutrient composition of food items from distinct agroecological contexts | Strong |
| 13 | Jones et al. (30) | Andes, Peru, n = 600 households in 48 different communities, women of reproductive age | Observational, cross-sectional | To determine the association of agricultural biodiversity within agricultural properties with the diversity and quality of the diet among women of reproductive age in Peru, evaluating the effect of the market on this association | Cultivated plants and animal husbandry (poultry, goats, and sheep) | Use of a 24-h recall, with a repeat in a sample of 100 women. Considering the information provided in the paper, researchers did not classify species by their taxonomy | Dietary species richness (DSR), Diet diversity score (DDS), Minimum dietary diversity for women (MDD-W), Probability of adequacy (PA), and Crop species richness (CSR) | Agricultural biodiversity within farms was associated with moderately more diversified and more micronutrient-rich diets among Peruvian women. Agricultural market orientation did not mediate these associations | With an observational design, the ability to draw a causal inference from the observed associations is limited. The recall period (i.e., 24 h) for the independent variable (i.e., agricultural diversity in the prior agricultural season) did not fully align, which may lead to underestimating the association between farm biodiversity and diet outcomes. There are also limitations in the data from the food composition table available, which made it impossible to calculate the probable intakes of vitamins B6 and B12 | Strong |
| 14 | Kennedy et al. (31) | Saturia, Bangladesh, n = 313 households in 10 rural communities | Observational, cross-sectional | Describe the consumption of plant genetic diversity | Cultivars of food plants (rice, pulses, eggplant, potatoes, and bananas) | Researchers used various qualitative research techniques to adapt indicators to the local context, such as interviews with key informants, free listing, market research, and participant observation. They also adapted the 24-h recall to include indicators of genetic diversity and applied it to women in the households | Plant genetic diversity (PGD) of cultivars and germplasms | Using these two indicators in the context of using 24-h recall was an initial attempt to classify crop diversity. Women interviewed were able to identify many of the cultivars consumed. This degree of agricultural knowledge would likely decrease in urban settings or areas where agriculture is not the primary occupation. The authors also found that girls are a fundamental source of knowledge about green leafy plant species, as they are primarily responsible for collecting these vegetables. In-depth qualitative research before the dietary survey is essential for developing food codes in order to achieve a reasonable level of precision when applying the indicators | There is no report | Moderate |
| 15 | Lachat et al. (4) | Diversos países: Benin, Cameroon, Congo, Ecuador, Kenya, Sri Lanka, Vietnam, n = 6,226 (2,188 women e 4,038 children), rural communities | Observational, cross-sectional | Test different indicators for diets assessment in order to recommend a cross-cutting indicator capable of measuring biodiversity in the human diet, guiding interventions for human, and environmental health simultaneously | All food sources | Secondary data from the application of one 24-h recall (n = 6,226) between 2009 and 2015 | Dietary species richness (DSR), Diet diversity score (DDS), Simpson Diversity Index (D), Functional diversity (FD), and Mean adequacy ratio (MAR) | People consumed 234 different species, of which <30% were eaten in more than one country. Nine species were consumed in all countries and provided, on average, 61% of the total energy consumption and a significant contribution of micronutrients in the rainy season. Comparing indicators, the DSR was the one that best predicted diet quality. For each additional species consumed, the adequacy of dietary nutrients increased by about 3%. Therefore, the authors recommended DSR as the EST indicator to assess food biodiversity in diets | They used a single 24-h recall per subject. This method does not allow accounting for within-person variability and estimation of usual dietary intake. During the analysis, there was a lack of nutrient composition data for some foods, species, and varieties consumed | Moderate |
| 16 | M'Kaibi et al. (32) | Kenya, n = 525 households in 2 rural communities, children (24–59 m) | Observational, cross-sectional with a longitudinal component to assess seasonality in food consumption | Assess the effects of agricultural biodiversity and seasonal rains on the adequacy of diet and food security of families of preschool children | Whole diet, focusing on animals (hunted and farm-raised) and other food items obtained in natural habitats through hunting or gathering | The authors conducted a pre-dietary survey assessment to map local agrobiodiversity, including interviews with key local informants (elders), and focus group discussions. They applied four 24-h recall per person, two non-consecutive in the dry season and two in the rainy season | Household food insecurity access scale (HFIAS), Nutrient adequacy ratio (NAR), and Mean adequacy ratio (MAR) | Food intake was low, with most households not meeting the RDA for many nutrients. However, intake of energy, protein, Fe, Zn, Ca, and folate significantly improved in the rainy season. Agricultural biodiversity was positively related to all NARs and MAR, indicating a significant positive relationship between household agricultural biodiversity and children nutrition | The two areas studied were not so similar in terms of their agricultural and physical resources, despite being physically very close. The authors evaluated only foods grown or obtained in nature when assessing agricultural biodiversity, excluding foods purchased in stores and markets | Strong |
| 17 | Mathewos et al. (33) | Yayo Reserve, Ethiopia, n = 96 households in 4 rural communities | Observational, cross-sectional | Assess the contribution of fruit and vegetables grown in backyards to the livelihood of families and conservation of biodiversity | Fruits and vegetables produced in domestic kitchen gardens | The authors produced the biodiversity inventory through the assessment of 48 home gardens, considering plant diversity, frequency, and density. Researchers classified plants by their taxonomy. To assess the perceived contribution of fruits and vegetables to the livelihoods of family farmers, the researchers conducted semi-structured questionnaires, focus group discussions, and direct observation. The research did not use recognized dietary assessment methods | Shannon-Wiener Diversity Index (H′) | The size of the home gardens had a positive association with species richness. The authors demonstrated that fruits and vegetables in home gardens contributed considerably to the diets of families. The result revealed the direct and indirect contribution of fruits and vegetables in home gardens to biodiversity conservation. The direct contribution is that the species of fruits and vegetables grown in their backyards contributes to the increase of plant diversity. The indirect role is thar fruits and vegetables in the home gardens reduced the use pressure of forest products | There is no report | Moderate |
| 18 | Ntwenya et al. (34) | Kilosa District, Morogoro Region, Tanzania, n = 307 households in 3 rural communities | Observational, cross-sectional with a longitudinal component to assess seasonality in food consumption | Describe the foods available and consumed in the Kilosa district of Tanzania | All food sources | The researchers listed edible resources available before conducting the dietary survey by consulting members of local communities and by conducting market surveys. They classified plants by their taxonomy with the help of a botanist. Then, they performed a dietary assessment by using a 24-h recall in both stations. Different parts, forms, or stages of ripeness of the same food reported were counted separately | Food biodiversity score (FBS) | A total of 183 edible food items were reported by households, with more reports in the rainy season (n = 82) than in the harvest season (n = 64). The average number of foods consumed per day during the rainy season was significantly higher than in the harvest season. About 50% of families mentioned that their family members had a low acceptance of wild edible foods | There is no report | Moderate |
| 19 | Penafiel et al. (14) | Andes, Guasaganda, Ecuador, n = 178 households in 10 rural communities, indigenous women | Observational, cross-sectional | Assess dietary diversity and nutrient contribution of traditional foods | Traditional plants, i.e., grown locally or wild | Use of two 24-h recall with an interval of 14 days between applications during the rainy season. Afterward, wild species were collected with the help of a local guide and identified by their taxonomy. Researchers also used a FFQ was used to estimate the frequency of consumption | Nutrient adequacy ratio (NAR/EAR), Mean adequacy ratio (MAR), Dietary species richness (DSR), Traditional food diversity score (TFDS) and Minimum dietary diversity for women (MDD-W) | Researchers found a positive association between the consumption of traditional foods and the adequate intake of macro and micronutrients. The average diet had a MAR of 0.78. The consumption of traditional foods contributed 38.6% of the total energy intake. Higher consumption of local species was associated with an increase in the median MAR of macronutrients of 3.3% and micronutrients of 5.2% | The research team did not reach all households during one of the research stages due to torrential rain that limited access to some families. Due to legal issues, some animal samples were not collected | Strong |
| 20 | Remans et al. (35) | Malawi, Kenya, and Uganda (Sub-Saharan Africa), n = 170 households, in 3 rural communities | Observational, cross-sectional | Explore how functional nutritional diversity indicators can provide insights into the nutrient diversity of agricultural systems | Food plants | Use of a 24-h recall. Researchers classified species by their taxonomy, comparing samples with studies of the local flora | Functional diversity (FD): total, macronutrients, minerals, and vitamins.; Household food security access scale (HFIAS); Months of inadequate household food provisioning (MIHFP); and Household diet diversity scores (HHDDS) | Researchers identified a total of 77 species of edible plants. The application of FD allowed the identification of key species that add nutrient diversity to the system. The analysis has shown that adding or removing individual species can radically alter nutritional diversity. The authors advocate the use of this indicator as a tool capable of relating inputs from agriculture, human nutrition, and ecology | Researchers did not collect data on the quantities of food plants produced or the uniformity of these species. In addition, composition data did not include foods below the species level | Moderate |
| 21 | Termote et al. (12) | Congo, n = 492 women, rural (129) and urban areas (363) from different cities | Observational, cross-sectional | Describe the contribution of wild plants to local diets | Wild food plants | In previous ethnobotany research, researchers collected and identified species. They assess food consumption by using two non-consecutive 24-h recall | Nutrient adequacy ratio (NAR, by RDA) | The results showed that wild plants are insufficiently consumed in a biodiversity hotspot and precarious food security. The most significant contribution came from Dacryodes edulis, contributing 4.8% of the total energy intake. Considering the nutrient composition of the various wild plants available in the region and known to indigenous peoples, the potential for increasing food security is vast. Researchers argue that ethnobiological research must find ways to separate knowledge from the consumption of plant species | There was no information on the nutritional composition of these foods. Democratic Republic of Congo at the time did not have a food composition table | Strong |
| 22 | Wertheim-Heck and Raneri (36) | Hanoi, Vietnam, n = 389 households, women of reproductive age, urban area | Observational, mixed methods | Test whether the diversified retail offer contributes to more diversified and nutritionally balanced diets | All food sources | The study began by mapping retail food outlets within the study area. Next, the household consumption survey used the 24-h recall, with a sub-sample (n = 60) of non-consecutive application, considering species and varieties. Considering the information provided in the paper, researchers did not classify species by their taxonomy. In the second phase of the study, the authors conducted interviews and shopping trips; and finally, developed a documentary film with three participants detailing food purchase and consumption practices at home | Diet diversity score (DDS), Minimum diet diversity (MDD), Mean adequacy ratio (MAR), and Nutrient adequacy ratio (NAR) | The study found that supermarkets and convenience stores offer a higher percentage and a more extensive range of ultra-processed products than traditional open-air markets. Furthermore, the authors state that traditional establishments (e.g., wet markets, street markets, and market stalls) were essential for maintaining minimum dietary adequacy for poor people living in cities. However, in statistical terms, the diet quality in the different strata was not influenced by the geographic proximity of formal points of sale. On average, the MAR was just 0.54, which means that the women consumed just over half of their daily nutrient requirements | A limitation of the study design is the large response load on participants, generated through several follow-up surveys. Another limitation was the repetition of the 24-h recall with an interval of 1 year | Moderate |