Urban foraging is a global informal phenomenon that has been investigated in the Global North more than other parts of the world. Characterising the nature of urban foraging in the Global South is imperative given the rapid urbanisation and sustainable development priorities in the region. In this study, we interviewed 80 urban foragers in four cities in the eastern coastal region of South Africa, with an aim to understand the nature of urban foraging in a developing nation context. We asked foragers about their initiation to and motivations for foraging, their logistics, yields, and associated activities, descriptions of their foraging grounds, and if and how they had changed, and what they envisage as an ideal future for foraging. Many foragers started foraging in their childhood, in the company of friends and family, and, as in the Global North, regarded it as a cultural and recreational activity. Foraging was mostly done within a five-kilometre radius of home, on a weekly or fortnightly basis, and very few foragers processed or sold their foraged products. Unlike the Global North, formal green spaces were not foraged in and were perceived to be poorly provisioned in urban planning. Forests and roadsides were equally used by the foragers, and very few had been discouraged from foraging. Most foragers were enthusiastic about the possibility of more people foraging, having designated spaces for foraging, and foraging-based businesses such as processed products and ecotourism. We recommend that policymakers and land managers recognise and encourage foraging as a potentially sustainable use for stewardship of urban green spaces. To this end, we list the main wild edible fruit species used by foragers in the area, which could be planted in public spaces. We also suggest harnessing foragers’ knowledge of useful species and spaces to develop green spaces and foraging-based supply chains. 

Site selection

The study area is the Indian Ocean Coastal Belt (IOCB), a relatively small biome with high human population density and high biodiversity (Mucina and Rutherford 2006). The study area is experiencing rapid changes in land cover and land use (Jewitt et al. 2015), and urban foraging is common in this area (Sardeshpande and Shackleton 2020a). Within the IOCB are 10 urban municipalities, and Durban, a metropolitan municipality (Municipalities of South Africa 2018). The mean population in these municipalities is 492,720 people, with a range of about 120,000 people in KwaMbonambi, and about 3.4 million people in Durban (StatsSA 2016). Each of these municipalities was surveyed (including formal residential areas, informal settlements, and marketplaces) for foragers on at least one day between October and November 2018. Municipalities where one day of surveying yielded more than three respondents were selected for continued surveying until 20 forager interviews were collected. Interviews and municipalities where one day of surveying yielded no more than three respondents were discarded. Thus, a total of 80 forager respondents were interviewed in four cities (mean 20±1.6). Photographs of some of the sites where foraging occurs can be found in Supplementary Materials.

Data collection

Adult individuals were opportunistically approached at their doorstep to enquire about their foraging activities if any. No prior criteria (position in household, age, gender) were used to select respondents. One consenting adult per household who foraged was interviewed individually, about their foraging practices and preferences based on the research questions (Appendix 1). Key questions included their household composition (number of adult, child, employed, and foraging members), their introduction to and company while foraging, the frequency of and time spent foraging, the various items foraged and activities undertaken alongside foraging, descriptions of the foraging grounds, and their normative views on the future of foraging, including local enterprises centred on foraging, with a specific focus on wild edible fruits. A booklet of 51 common wild edible fruit species was used to facilitate identification by respondents with the help of pictures of fruits, flowers, leaves, and morphometrics and local names (Sardeshpande and Shackleton 2020a, Supplementary Materials). The semi-structured interviews were conducted in English by MS, or in isiZulu with the help of a field assistant interpreter, according to the respondents’ preference. All interviews were recorded as audio files with verbal consent from the respondents, and transcribed by MS at a later stage. This study was approved by the Rhodes University ethics committee (Application No. ES17/46). As per the ethics approval, respondents were provided with a printed and verbal brief description of the study, verbal consent was obtained from each respondent, and the data were anonymized. Verbal consent was obtained in keeping with the verbal recording format of data collection, and to overcome barriers related to illiteracy and anonymity.

Data analysis

Data were entered into MS Excel for analyses, namely descriptive statistics, and manual coding based on emergent themes along the five lines of investigation, i.e., the research questions. Coding was done by MS using an inductive approach, and framework and thematic analysis based on the research questions (e.g. Braun and Clarke 2019, Robinson et al. 2019). Data were anonymised, and for the purpose of reporting, respondents were identified by alphanumeric representations of their municipality, neighbourhood, and interview number; for example, the third respondent from the neighbourhood of Msinsi in Mtubatuba would be MTMS3. Kruskal-Wallis anovas were used to test for between-site differences for: distances traveled, frequency of, time spent, and quantity and number of wild edible fruit species collected when, foraging. Mann-Whitney U tests were applied to investigate differences between genders for the same variables. Spearman rank correlations were performed to test for relationships between household composition and the aforementioned variables. Inferential statistics were performed in R 3.2.3 (R Core Team 2015).