1Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports
Physical distancing and risk 
of COVID‑19 in small‑scale 
fisheries: a remote sensing 
assessment in coastal Ghana
Isaac Okyere 1,2*, Ernest O. Chuku 1,2, Bernard Ekumah 3, Donatus B. Angnuureng 2,  
Justice K. Boakye‑Appiah4, David J. Mills5,6, Raymond Babanawo7, Noble K. Asare1,2, 
Denis W. Aheto1,2 & Brian Crawford7
The novel coronavirus is predicted to have dire implications on global food systems including fisheries 
value chains due to restrictions imposed on human movements in many countries. In Ghana, food 
production, both agriculture and fisheries, is exempted from restrictions as an essential service. 
The enforcement of COVID‑19 prevention protocols, particularly social distancing, has been widely 
reported in Ghana’s agricultural markets whereas casual observations and media reports on fish 
landing sites suggest no such enforcements are in place. This study aimed to provide sound scientific 
evidence as a basis for informed policy direction and intervention for the artisanal fishing sector in 
these challenging times. We employed an unmanned aerial vehicle in assessing the risk of artisanal 
fishers to the pandemic using physical distancing as a proxy. From analysis of cumulative distribution 
function (G‑function) of the nearest‑neighbour distances, this study underscored crowding at all 
surveyed fish landing beaches, and identified potential “hotspots” for disease transmission. Aerial 
measurements taken at times of peak landing beach activity indicated that the highest proportion of 
people, representing 56%, 48%, 39% and 78% in Elmina, Winneba, Apam and Mumford respectively, 
were located at distances of less than one metre from their nearest neighbour. Risk of crowding was 
independent of the population at the landing beaches, suggesting that all categories of fish landing 
sites along the coast would require equal urgency and measured attention towards preventing and 
mitigating the spread of the disease.
The importance of small-scale fisheries to food and nutrition security and the livelihoods of many people within 
and beyond coastal areas of developing countries cannot be overemphasised. Globally, the small-scale fisheries 
sector is estimated to support 32 million direct  jobs1. Over half of the catch in developing countries is produced 
by the small-scale fisheries sector, and 90 to 95 percent of the small-scale landings are used for local human 
 consumption1. In Ghana, the fisheries sector provided an annual revenue of more than US$ 1 billion in  20152. 
With substantially higher per capita consumption of fish (24 kg) in Ghana compared to the world average (16 kg), 
fish remains the principal source of low cost protein accounting for about 60% of animal protein in the local 
 diet3, and contributes substantially to the supply of essential micro-nutrients4. As a dominant contributor to 
livelihoods in coastal areas, the fisheries sector provides direct and indirect employment for approximately 10% 
of the country’s population of 30 million  people2. A majority of these are small-scale (referred to as artisanal) 
fishers operating from 186 fishing villages and nearly 300 landing beaches along the  coast5. They contribute 
approximately 70% of the country’s marine fish  production2. Typical of labour-intensive artisanal fisheries, fishing 
is characterised by clustering of fishers and close human interactions during harvesting while onboard vessels, 
offloading fish from canoes to landing beaches, at processing sites and in fish markets.
OPEN
1Department of Fisheries and Aquatic Sciences, University of Cape Coast, Cape Coast, Ghana. 2Centre for Coastal 
Management (Africa Centre of Excellence in Coastal Resilience - ACECoR), University of Cape Coast, Cape 
Coast, Ghana. 3Department of Environmental Science, University of Cape Coast, Cape Coast, Ghana. 4Institute 
for Infection and Immunity, St George’s University of London/Hospital, Cranmer Terrace, London SW17 0RE, 
UK. 5WorldFish, Bayan Lepas, Penang, Malaysia. 6ARC Centre of Excellence for Coral Reef Studies, James Cook 
University, Townsville, QLD, Australia. 7Coastal Resources Center, University of Rhode Island, Narragansett, RI, 
USA. *email: iokyere@ucc.edu.gh
2Vol:.(1234567890)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
The end of 2019 saw the emergence of the severe acute respiratory syndrome virus 2 (SARS-CoV-2), the 
causative agent of COVID-19 in humans as reported by the World Health Organisation (WHO)8. By the second 
week of March 2020, the disease had so widely spread with rising incidence that the WHO declared it a pan-
demic  emergency6. By its mode of spread, mainly through aerosols, physical distancing (otherwise called social 
distancing) between individuals remains one of the main public health measures to curb its spread. This has 
so far been the focus of efforts to control the disease and reduce infection  rates7. The WHO therefore recom-
mended a minimum distance of 1 m between individuals at any point in time and especially in public  places8. 
Some countries however recommend approximately 2 m (6 feet) in their local  guidelines9.
The COVID-19 pandemic was projected to have significant impacts on global fisheries systems throughout 
value  chains10,11. Although SARS-CoV-2 is not known to infect or contaminate  fish12, fishing communities are 
considered to be at high risk, serving as potential “hotspots” for rapid spread of the virus due to the migratory and 
huddling behaviour of fishers, and at times poor hygienic practices in these  communities10,11. While implement-
ing stringent restrictions on movement and physical contact to curb the spread of the disease, countries have 
also critically weighed the balance between disease control and nutritional needs, and most have allowed sectors 
within the national food system, including the fishing industry, to continue operating. Ghana, in its implementa-
tion of an emergency response to the COVID-19 pandemic, exempted fishers from a partial lockdown due to 
their essential role in food  supply13. Despite being recognised “hotspots” for virus transmission, no targeted and 
coordinated measures have been instituted to control physical contact and ensure adherence to the protocols on 
social distancing and other preventive measures at fish landing beaches.
Fishing communities in Ghana are typically highly populated with poor sanitation  practices14 and low lit-
eracy  rates5. Given the communal nature of artisanal fishing, anticipating voluntary compliance from fishers of 
COVID-19 preventive measures which they may or may not be aware of is unrealistic. Of serious concern is a 
misconception that fishers are immune to the virus due to their proximity to the  sea15. Extra effort will likely be 
needed to break the disease transmission chain and prevent escalation among local fishers should an outbreak 
occur in fishing communities.
Considering the rising number of COVID-19 cases in Ghana at the time of writing, and especially as cases 
have been recorded in all four coastal regions of the country by the Ghana Health  Service16, it is important to 
rapidly assess the risk among different categories of fishing communities to the pandemic, and identify potential 
high risk areas (hotspots) to address rapid control measures. The outcome of the rapid assessments will assist 
government, development agencies and NGOs in developing targeted interventions to be instituted at high-risk 
landing beaches to reduce the risk to fishers, their families and the community.
To this end, this study aimed to assess the extent of clustering of fishers at different categories of small-scale 
fish landing beaches (low, moderate and highly populated) in order to identify potential “hotspots” for spread 
of the disease. The exigency of the pandemic required a swift survey within fishing communities for expedited 
intervention. Hence, we employed a rapid appraisal methodology using an Unmanned Aerial Vehicle (UAV) 
and Geographic Information System (GIS) tools. The aerial view not only provides the physical opportunity to 
assess distance between subjects with no or minimal intervention, but it does so quickly and at minimal risk to 
all involved, whilst effectively preserving anonymity of subjects.
Materials and methods
Study area and selection of landing beaches. The study was carried out at six landing beaches along 
the coast of the Central Region of Ghana. According to the 2016 Ghana Canoe Frame  Survey5, the coastline 
of the Central Region is dotted with 97 landing beaches; the highest density among the four coastal regions in 
the country (the others are Western—89, Greater Accra—59 and Volta—47). Aside from its numerous landing 
beaches and vibrant artisanal fishery, the study was limited to the Central Region (location of the research team) 
due to COVID-19 lockdowns and other travel restrictions imposed on key regional corridors at the time of the 
 study13.
In the context of this study, “fishers” is used for both genders while “fishermen” and “fish processors” refer 
to men and women respectively. Categorisation and selection of landing beaches were based on the reported 
population of fishermen in the 2016 Ghana Canoe Frame  Survey5. Using the estimated population of a minimum 
of 30 and maximum of 2,200 fishermen in the region, the landing beaches were categorised by terciles into low 
(30–700), moderate (701–1400) and highly (1401–2200) populated. To ensure a balance of spatial distribution 
of study locations across the region, the coastline was bisected into East and West zones (Fig. 1). One each of 
low, moderate and highly populated landing beaches were selected in each zone based on, firstly, location with 
the highest number of fishermen within the category, and secondly for the low population beaches, and finally 
based on the knowledge of the researchers on the existence of active fishing at the location. Using these criteria, 
the landing beaches selected were Winneba ‘Ayipei’ (high), Apam ‘Main’ (moderate) and Mumford ‘Main’ (low) 
in the East zone, and Elmina ‘Main’ (high), Biriwa ‘Abaka Ekyir’ (moderate) and Cape Coast ‘Abrofo Mpoano’ 
(low) in the West zone.
UAV deployment and capturing of aerial imagery. An unmanned aerial vehicle, UAV–DJI Phantom 
4, equipped with propeller guards (to prevent direct contact with any obstacle including flying animals) was 
deployed at the six landing beaches to capture aerial photos on different days from 10th April to 23rd April 
2020. Drone deployment was carried out between 08:00 and 09:00 GMT since fish trade at the landing beaches 
usually spans the hours of 08:00 to 16:00 GMT with most vibrancy in the mornings. The flight missions were 
pre-planned and executed using DroneDeploy v.2.66.2 on an iPad mini 4. Several pre-flight tests informed the 
appropriate flight altitude of 60 m based on easy location of humans in the images, obscured identities in the 
3Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
overhead images, and safe height to avoid collision with any obstacles. The resolution of images captured was 
1.8 cm. UAV flights were coordinated by pilots certified by the Ghana Civil Aviation Authority.
Point data (human locations) extraction and measurement of nearest‑neighbour dis‑
tances. The aerial photos obtained were uploaded into Agisoft Metashape Professional v.1.5.3.8407 and 
processed into seamless orthomosaic images, and identifiable features related to artisanal fishing at the various 
landing beaches marked (Fig. 2). The locations of people in orthomosaic images were manually extracted as 
point data in ESRI ArcMap v.10.3 using the editor tool. From the point data, the distance from each point to the 
nearest other point, that is the nearest-neighbour distance (NND), was measured for all individuals present in 
each of the six landing beaches in this study.
The median distances were compared to the WHO and Center for Disease Control (CDC) standards on 
physical (social) distancing. The term social distancing is used more frequently in this article due to its usage 
in common parlance. The application of UAV in a way that subjects are not identifiable and does not link infor-
mation collected to the subjects qualifies the study for exempt human subject research (HSR)17, thus ethical 
clearance was issued by the Institutional Review Board (IRB) of the University of Cape Coast for the study. It is 
noteworthy, however, that aerial images from the UAV flights were unable to detect human presence at sheltered 
areas of the landing beaches, and as such there is likely an underestimation of proximity particularly in areas 
where fish trade occurs under sheds. Minor errors are also anticipated through the multiple capture of mov-
ing objects (persons) at margins of images. In addition to the imaging, hand washing facilities that have been 
placed at the landing beaches were manually counted and their geo-locations recorded with handheld Garmin 
GPS (Garmin GPSMAP 64s).
Focus group discussions. The Chief Fishermen of all selected landing beaches were informed of the study 
through telephone calls, and their consent sought for the participation of their community in AUV surveys 
and associated stakeholder interviews. To engage fishers and industry participants in the study to understand 
their knowledge of COVID-19, perception of risks, and opinions on plausible solutions, we initially planned to 
conduct key informant interviews by telephone. These were, however, unsuccessful, and the strategy was revised 
Figure 1.  Study area showing locations of the six fish landing beaches selected for the study.
4Vol:.(1234567890)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
Figure 2.  Orthomosaic images of the (A) Elmina ‘Main’, (B) Winneba ‘Ayipei’, (C) Biriwa ‘Abaka Ekyir’, (D) 
Apam ‘Main’, (E) Cape Coast ‘Abrofo Mpoano’, and (F) Mumford ‘Main’ landing beaches with artisanal fishing-
related features marked.
5Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
to conduct brief focus group discussions (FGDs) in three communities. Two FGDs of not more than three key 
informants were conducted in each of the communities, ensuring strict adherence to the WHO’s protocols on 
social distancing, hand washing and wearing of face masks. For each landing beach, one FGD targeted fishermen 
(constituted of Chief Fisherman, a canoe owner and a fisherman) and another, the fish processors (the queen 
fish processor and a fish processor) to achieve disaggregation by gender of participants. These groups are the key 
stakeholders of the artisanal fisheries production value chain in Ghana. The discussions were recorded for sum-
mative analyses on the issues, and the results were later validated with the participants.
Data analyses. Density and distance-based methods of point pattern analysis  techniques18 were used to 
analyse the location of people (point data) in determining the hotspots of observation at the landing beaches 
based on quadrat density and Kernel density. All imagery analyses were carried out in the R-programme with 
the “sf ”, “raster”, “sp”, “spatstat” and “maptools” packages. Quadrats of 2 m2 grids were employed, assuming 2 m 
as appropriate COVID-19 social distancing protocol. Further, homogeneity of locations of people at the landing 
beaches was tested with a non-parametric Chi-squared test of Complete Spatial Randomness (CRS) with the 
assumption that point process was Poisson Process (See Supplementary Table S1). Nearest-neighbour distance 
(NND) and the cumulative distribution function (G-function) of the NND, i.e. G(r), were used to compute the 
shortest distance from each person’s location to the nearest other person. The significance of any departures 
from Complete Spatial Randomness (either clustering or regularity) was evaluated using simulated “confidence 
envelopes”.
The means (± S.E.) of  Log10(NND) for the landing beaches were compared using one-way ANOVA in Minitab 
19.1. Independent measurements of NNDs were treated as repeated responses (response) for each landing beach 
(factor) and landing beach category (factor). Specific differences between landing beaches and categories were 
identified using Tukey HSD as post-hoc test. The data transformation to  Log10(x) was applied following normal-
ity (Kolmogorov–Smirnov) and homogeneity (Cochran’s) tests. Median was, however, reported as the central 
tendency for NND due to the large proportion of outliers and skewness in the data (see Fig. 6). The level of 
significance and confidence level for all inferential statistics in the study were set to 0.05 and 95% respectively.
Analyses of information from the focus group discussions were guided by an inductive content analysis 
 approach19. Responses were organized into main sections similar to the sections in the instrument, and general 
themes that emerged were developed after which sub-themes were created and assigned codes. Views and expe-
riences of the fishers on the COVID-19 subject were identified under sub-themes to aid comparison, and final 
analyses were summarised and tabulated into a matrix of thematic issues and summary responses.
Ethics declarations. Ethical approval for this study was given by the Institutional Review Board of the Uni-
versity of Cape Coast, all procedures followed the required ethical (Belmont) guidelines, and informed consent 
was obtained from all participants.
Results
Human densities at the landing beaches and potential clustering hotspots. From the UAV 
imagery, a total of 2,890 individuals were recorded at the Elmina ‘Main’ landing beach, 1,632 at Winneba ‘Ayipei’, 
1,053 at Apam ‘Main’, 236 at Cape Coast ‘Abrofo Mpoano’ and 209 at Mumford ‘Main’, with Biriwa ‘Abaka Ekyir’ 
having the lowest of 79 people. Quadrat density analyses of the point data (Fig. 3) showed the occurrence of 
humans across the length of all six landing beaches with noticeable aggregation at certain areas of each land-
ing beach. In these clusters, concentration of 3 to 4 persons within 2 m2 grids as indication of crowding were 
prominently observed at Elmina and Mumford, considerably at Winneba, Apam and Cape Coast, and sparingly 
at Biriwa.
Results of the cumulative distribution function (G-function—see supplementary Fig. S1) further buttresses 
the occurrence of clustering at the beaches; clustering occurs when the observed G(r) curve is above the theoreti-
cal G(r) curve [i.e. G(r)obs > G(r)theo]. Hotspots in the Kernel density plot (Fig. 4) shows the specific locations of 
clustering at each landing beach, which are the potential hotspots for fast spread of COVID-19 in the communi-
ties in case of an outbreak. Generally, the clusters occurred at the main fish offloading sites at the landing beaches 
and imperceptibly at the canoe berths and net mending areas. At Elmina, two potential hotspots were observed, 
the first and most prominent being the main fish landing quay which incorporates a fish market, and the second 
occurring at the transhipped fish (locally known as “Saiko” fishing) offloading site. Similar dual hotspots were 
observed at Apam where the offloading site for the purse seine “watsa” fishers had even more intense clustering 
than the main offloading site. Like the Elmina fish landing quay, Winneba ‘Ayipei’ landing beach which also 
has a fish market integrated into the beach had the hotspot of clustering occurring within the market enclave.
Physical distancing at the landing beaches. In all, 2,890, 1,632, 79, 1,053, 236 and 209 independent 
measurements of closest distance between persons (NNDs), in metres, were obtained at Elmina ‘Main’, Winneba 
‘Ayipei’, Biriwa ‘Abaka Ekyir’, Apam ‘Main’, Cape Coast ‘Abrofo Mpoano’, and Mumford ‘Main’, respectively. Over 
70% of the distances between persons recorded at each landing beach were less than 2 m from the nearest-neigh-
bour (Fig. 5). Of a serious concern is the finding that the largest proportion of persons at the Elmina (56.06%), 
Winneba (48.22%), Apam (38.46%) and Mumford (77.51%) landing beaches were at distances less than the 
WHO recommended minimum of 1 m from their nearest-neighbour.
Further, box plots of NND observations at the various landing beaches (Fig. 6) indicate median NNDs of 
less than 1.4 m for Winneba, Biriwa, Apam and Cape Coast whereas median NNDs below 1.0 m were recorded 
at Elmina and Mumford. Essentially this underscores the undermining of social distancing as a COVID-19 
prevention protocol at the fish landing beaches. The situation was accentuated at Mumford, which had the most 
6Vol:.(1234567890)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
huddled individuals (median NND = 0.65 m). Although there was a significant difference (F = 52.56; P = 0.000) in 
NND among the categories of landing beaches, there was no sequential order from lowly, through moderately to 
highly populated categories of landing beaches in this study. Lowly and highly populated landing beaches were 
statistically similar (F = 46.79; P = 0.000) in NND. One-way ANOVA of  Log10(NND) showed significant differ-
ences in crowding among the six landing beaches (F = 46.79; P = 0.000). Notably, Biriwa, Apam and Cape Coast 
Figure 3.  Quadrat density plots of humans at the six surveyed beaches (legend represents number of humans 
occurring in 2 m grid).
7Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
had similar crowding situations comparing means of  Log10 (NND) simultaneously (P > 0.05). The situation at 
Winneba was also similar to Biriwa (P = 0.000).
Summary of outcomes from the focus group discussions. Feedback from the FGDs (Table  1) 
showed that fishers were aware of COVID-19, with some sensitisation being carried in the fishing villages by 
Municipal Assemblies and community opinion leaders through community radio stations and other informa-
tion systems. Measures such as hand-washing facilities had been provided at the landing beaches to improve 
Figure 4.  Kernel density for point pattern showing landing beach-specific hotspots for potential spread of 
infectious disease. Scale represents Kernel Function; independent for each landing beach and not for comparing 
landing beaches).
8Vol:.(1234567890)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
hygiene and mitigate spread of the disease. During the field visits, seven hand-washing stations were counted 
at Elmina, four at Winneba, two each at Cape Coast and Apam, and one each at Biriwa and Mumford. While 
the fish processors were of the view that fishers were not cooperative in observing social distancing and other 
measures, the fishermen indicated that fishers were cooperative. To the fishermen, the most difficult period to 
maintain appropriate distance from others was during offloading fish and hauling their canoes to berth while for 
the women, their difficulty was during purchases of offloaded fish and selling at the landing beach. The pandemic 
had negatively impacted the fishermen through restrictions on their movements to neighbouring West African 
countries to fish, and they had additionally reduced the crew size in the canoes to decrease crowding, thereby 
reducing the labour force, catch and income which had in turn affected sales and income of the women. The fish-
ers mentioned logistical and financial support was needed to help prevent and mitigate the spread and impact of 
COVID-19 at the landing beaches, suggesting the need for engaging law enforcement agencies to ensure compli-
ance with social distancing at the landing beaches.
Figure 5.  Nearest-neighbour distance at the six landing beaches (dotted vertical lines represent WHO reference 
of 1 m and CDC reference of 2 m minimum physical distance).
9Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
Discussion
From the onset of COVID-19, the pandemic has been projected to take a significant toll on global fisheries, with 
predictions of potentially negative consequences for the livelihoods and incomes of the multitudes of small-
scale fishers in developing  countries10,11. Although it is seemingly early to meaningfully appreciate the effects 
of the disease on fisheries, at the time of this study, reports had started trickling in on its impacts on small scale 
 fisheries20. Changing consumer demands, market access and border restrictions among others, are predicted to 
be the major drivers of the impact of COVID-19 on  fisheries10,11. Aside from these, there is heightened apprehen-
sion that fishing communities are at high risk of COVID-19 due to the migratory and clustering behaviour of 
fishers, making the fishing communities potential “hotspots” for rapid spread of the  disease11.
Figure 6.  Distribution of nearest-neighbour distances (NND) at the various fish landing beaches (n = number 
of observations/distances). Alphabets represent significant differences (one-way ANOVA) in mean ± S.E. of 
 Log10(NND) among the landing beaches; Groups that do not share a letter are significantly different.
Table 1.  Summary of FGDs conducted at the landing beaches.
Thematic issue
Summary of response
Fishermen Fish processors
Awareness of COVID-19 Aware through radio and TV
Mode of sensitization in the community Community radio and information centres, posters Community radio and information van
Measures in place to mitigate spread Provided hand-washing facilities and advising on social distancing
Provided hand washing facilities and advising on social 
distancing and personal hygiene
Cooperation of fishers to mitigation measures Very cooperative and adhering to the measures Not too cooperative
Most difficult period to maintain appropriate distance from 
others
Offloading fish and hauling their canoes to berth after 
landing Buying offloaded fish and selling fish at the beach
Impact of COVID-19 on fishing activities
Restrictions especially to other countries for fishing, 
low fish catch due to reduction in fishing crew, reduced 
incomes
Reduction in sales due to low fish catch, reduced income
Roles of Fisher Leadership in mitigating measures Provided some of the hand-washing facilities, have been sensitising fishers Have been sensitizing fishers about the disease
Role of Fisheries Commission and Municipal Assemblies in 
helping prevent COVID-19
Municipal Assemblies carry out sensitization, have pro-
vided some of the hand-washing facilities and conducted 
mass testing at some landing beaches awaiting results. 
Marine police of the Fisheries enforcement unit also car-
ried out some education
Support needed to enhance safety of fishers Financial support, more hand-washing facilities, hand sanitizers, face masks, and food supply
Attitudinal change by fishers, provision of sanitizers, face 
masks, food supply and financial support
How to effectively adhere to social distancing during fish-
ing activities
Properly organise the process of buying fish at the beaches, 
intensify sensitization and use law enforcement agencies
Ensure compliance through use of law enforcement agen-
cies
10
Vol:.(1234567890)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
Considering that the disease spreads mainly through aerosolisation during person-to-person  contact9, it is 
critical to ensure that fishers adhere to the protocols on preventive measures while carrying out their activities. 
Central to this study is the adherence to the protocol on social distancing, prescribed by WHO as a minimum 
of 1 m8 and the CDC as minimum of 6 ft (≈ 1.8 m)9 physical distance required to be kept from others in avoid-
ance of contracting the virus.
Results of this study have shown that huddling and social distancing remain a major challenge at fish landing 
beaches in the Central Region of Ghana in the face of COVID-19. Clustering of fishers at densities up to four 
people within a 2 m2 area at most of the landing beaches, and over 70% of the people across landing beaches 
occurring at less than 2 m from their nearest-neighbours importantly underscores the need to devote attention 
to the crowding situation at the beaches. Another worrying situation is the occurrence of a substantial proportion 
of people at Elmina (≈ 56%), Winneba (≈ 48%), Apam (≈ 39%) and Mumford (≈ 78%) at a distance of less than 
1 m from their nearest-neighbours which provides ample evidence of inability of fishers to voluntarily adhere 
to both WHO and CDC social distancing protocols despite being aware of the disease and the continuous sen-
sitisation. Concerningly, these findings contradict the views of the fishermen that fishers are very cooperative in 
observing the mitigation measures, but supports the opinion of the women fish processors that fishers are not 
too cooperative with the measures. Heat maps from the Kernel density plot further point to the fish offloading 
sites and fish trading areas at the landing beaches as the hotspots for clustering, hence, the potential areas for 
escalating the spread of the disease. This corroborates the feedback from the fishermen that the most difficult 
period to maintain appropriate distance from others was during offloading and hauling of their canoes, and in 
the case of the fish processors, when purchasing offloaded fish and trading at the landing beach. Specific meas-
ures targeting the various activities at the landing beaches will therefore be required to mitigate the spread of 
the disease as articulated in latter sections of the discussion.
Hypothetically, it was expected that the highly populated landing beaches documented in the canoe frame 
 survey5, may have higher incidence of crowding and therefore increased risk of people contracting the COVID-19 
disease through huddling. However, the absence of a correlation between huddling and landing beach population 
categories (i.e. low, moderate and highly populated) reinforces that it is the nature of the activities engaged in, 
rather than population size, that drives risk. For instance, Biriwa ‘Abaka Ekyir’ which fell within the moderately 
populated category with 246  fishermen5, had relatively low activity compared to the “lowly populated” Cape 
Coast ‘Abrofo Mpoano’ and Mumford ‘Main’. The contrast between the intensity of activities observed at these 
landing beaches at the time of field data collection and the expected intensity based on available  data5, could 
mainly be resulting from low fish catch in the lean fishing season [i.e. November to  May3]. During this fishing 
season, fishers migrate from rural landing sites to urban fish landing beaches to access cosmopolitan fish markets 
such as that of Cape Coast and Elmina. In addition, Mumford ‘Main’, a low population category landing site, 
which in this study had about one-fourteenth the number of people observed at Elmina ‘Main’ (highest number 
of people in this study), was estimated to be the most crowded amongst the six (median NND = 0.65 m). The 
mode of business transaction at the landing site is a major contributing factor here. Mumford ‘Main’ is a sandy 
beach landing site with no sheds, forcing fish traders (including middle-men) and processors to congregate in 
proximity in anticipation of fish offloads and direct sales from canoes. This brings to the fore the need for more 
industry specific analysis of worker functions and work environment context in making decisions on interven-
tions in preventing the spread of COVID-19 and potentially other similar pandemics.
Our research suggests that fish landing sites across the coast, irrespective of size and population density, 
would require similar urgency and attention towards the prevention of the spread of highly infectious diseases 
like COVID-19. In this regard, such rapid, remote, and real-time assessments as used in this study could be a 
transformative tool to facilitate immediate informed response. The undesirable clustering at the landing beaches 
in the Central Region is akin to the situation at most of the nearly 300 coastal landing beaches in the country. 
Unfortunately, stringent measures instituted by the Metropolitan, Municipal and District Assemblies (MMDAs) 
to enforce appropriate physical distancing focused on agricultural markets has been at the neglect of the fish 
landing beaches. So stringent have these measures been, that non-compliant agricultural markets have been shut 
 down21. Contrary to the agricultural markets, no fisher was spotted wearing a face mask at the landing beaches 
during the field visit. As noted from the field and focus group discussions, the leadership of the fishers acquired a 
small number of hand-washing facilities to improve hygiene at the beaches, and the respective MMDAs had also 
provided additional hygienic facilities coupled with occasional sensitisation, including carrying out mass testing 
at areas such as Winneba. Beyond these localised efforts, the Ministry of Fisheries and Aquaculture Development 
distributed a significant number of hand-washing facilities and face masks to most fishing communities along the 
coast approximately one month after our field survey. However, a recent assessment of all 300 landing beaches 
revealed that the hand-washing facilities were functional at less than ten landing beaches, and about 96% of the 
landing sites had only a few or none of the people wearing face  masks24. Given the communal nature of artisanal 
fishing, low literacy rate of  fishers5 and the illusion that the proximity of fishers to saline water renders them 
immune to COVID-1915, it is perhaps unrealistic to expect fishers to voluntarily comply with social distancing 
with the level of support currently offered.
Based on observation of direct behaviours and attitudes expressed in FGDs, our pilot study demonstrated the 
inadequacy of current provisions to reduce risk of COVID-19 transmission in the focus sites. A targeted preven-
tion strategy is importantly needed for coastal communities as three of the four coastal regions are currently 
among the top five regions with the highest COVID-19 cases out of the sixteen regions in Ghana. These coastal 
regions are the Greater Accra Region (24,826), Western Region (2974) and Central Region (1931), which together 
represent 62% of the total confirmed cases of 47,690 as of Sunday, October 25,  202016. A two-pronged approach 
would be suitable for reducing the risk of fishers; first by implementing targeted landing beach-wide interven-
tions and secondly, workplace or activity-specific measures. At the landing beach level, deploying strategies 
specific to each fishing community within which the landing beach is located will be helpful. Vacant land spaces 
11
Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
adjacent to landing beaches (e.g. in Elmina) could be repurposed to provide additional space for fish trading 
activities in order to minimise crowding in the main offloading sites and keep fish mongers and buyers spread 
out. Where there are no vacant spaces, regulating the number of canoes to fish in a day, staggering landing times, 
running a shift system for the fishers, and designating entry and exit points could be considered. These will limit 
the number of people and personal contact at the beaches. Though this could deepen the economic impact on 
the fishers, it has generally been successful in agricultural markets. Economic and other livelihood support for 
vulnerable fishers affected by such measures could be considered to ameliorate the impacts on their households.
The workplace or activity-specific measures should target offloading of fish, hauling of canoes and nets, pur-
chasing and marketing, as well as processing of fish. In carrying out labour-intensive activities such as manual 
hauling of canoes and offloading of fish, which make it practically difficult for fishers to maintain appropriate 
physical distances from others, wearing of face masks by the fishers should be adopted as primary prevention 
protocol as prescribed by the WHO. This is particularly important as the process of hauling canoes and nets by 
artisanal fishers in Ghana are characterised by the culture of chanting work-songs accompanied by instructional 
shouts through which aerosols could be transmitted and possibly spread the disease if present. The wearing of 
face masks should also be extended to the fish purchasing, marketing and processing activities as they involve 
yelling for customers and active bargaining.
Considering that regular human and material contacts are inevitable during fishing activities due to the lack 
of physical distancing demonstrated in this study, frequent washing and disinfection of hands and all fishing 
equipment are imperative. From the FDGs, there were indications of some provisions of hand washing facilities 
at the landing beaches. Nonetheless, some of the facilities were observed without water, soap and other needed 
consumables during the field visit. Adding to the existing hand washing facilities and importantly ensuring 
frequent supply of these consumables including hand sanitizers are critical to enhancing COVID-19 hygienic 
practices at the landing beaches. Although very little fish in the artisanal fish value chain goes through cold stor-
age, cold-chain transportation of food including fish is reportedly an important way of spreading the  virus25. 
Therefore, instituting modalities for disinfecting frozen fish at imported fish outlets and the transhipped fish 
(Saiko) landing sites is also very critical as both the transhipped fish from industrial trawlers and imported fish 
end up mainly with fish processors along the coast.
For a rather lasting intervention, the approach of behavioural change communications could be used as a 
strategy to achieve voluntary compliance to physical distancing by the fishers. By disabusing the fisherfolk of their 
misconception about the disease, presenting customised COVID-19-related information relevant to their trade, 
and supporting with the specific workable practices discussed in this paper, fishers could understand better the 
need for social distancing and practice same. Direct feedback to communities and the local MMDAs from rapid 
appraisal such as that conducted here may be instrumental in gaining recognition that current practices are not 
adequate and reinforcing the areas of high risk. A coordinated effort of MMDAs and the Fisheries Commission is 
crucial in addressing the situation at the landing beaches, and funds for the intervention could be collaboratively 
drawn from the District Assemblies Common Fund, the Fisheries Development Fund and the Government of 
Ghana’s COVID-19 intervention fund. Even though studies have suggested that SARS-CoV-2 thrives better and 
possibly spreads faster in the mid-to-higher latitudes than in the lower latitudes due to optimal temperature and 
humidity for its spread in these climatic  zones26 which may account for the lower number of cases recorded in 
Ghana and other parts of Africa, it is important not to relent on ensuring adherence to the preventive protocols 
in the face of rising number of cases in the country.
Lastly, the safe utilisation of UAV in this work, which qualifies for exempt UAV-Human Subject  Research17,22, 
advances drone applications in fisheries  research23 through the application of UAV in assessing clustering related 
to the fish trade in a developing country towards curtailing the spread of infectious diseases. Among others, the 
demonstration in this study could serve as a reference for rapid behavioural prognostics in fisheries, especially 
during diseases of the nature of COVID-19.
Conclusion
Unmanned Aerial Vehicle (UAV) technology was successfully used to rapidly and efficiently assess the risk to 
artisanal fishers from the pandemic using social distancing as a proxy: an approach that can be easily and quickly 
scaled up for broader application. The findings of the study have demonstrated that clustering and social distanc-
ing remains a challenge at fish landing beaches in the Central Region of Ghana in the face of growing cases of 
COVID-19 in the country. A significant level of clustering occurred within distances of less than 1 m, indicating 
the inability of fishers to voluntarily adhere to both WHO and CDC social distancing protocols irrespective of 
their awareness of the disease. Enforcement at these beaches was also of great concern as no designated measures 
had been put in place to ensure supervised compliance. The risk of artisanal fishers to the COVID-19 disease 
from clustering was independent of the size of landing beach. Therefore, an urgent concerted effort across 
relevant health and fisheries governance institutions is required to institute radical prevention and mitigation 
measures. This effort should be focused primarily on implementing targeted landing beach-wide interventions 
on one hand and workplace or activity-specific measures on the other hand. Specifically, the judicious use of 
land spaces adjacent to landing beaches could present an opportunity for interspersing fish trading activities to 
reduce clustering. It is also recommended that fishers be compelled to adhere to wearing face masks during fish 
offloading, canoe hauling, and fish trading, when it is impractical to observe social distancing. It is noteworthy 
that this assessment was conducted during the lean fishing season and crowding situation could worsen in the 
main or bumper fishing season (July – September) posing an even greater risk should the disease persist. These 
findings provide a point of entry for government agencies, donors and NGOs in identifying potential support 
that could be provided to the landing beaches to reduce the risk to fishers, their families and community to the 
novel coronavirus disease.
12
Vol:.(1234567890)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
Data availability
The datasets generated and used during this study are available on the Mendeley’s free data repository platform 
(https ://data.mende ley.com/datas ets/2s6x2 5xsrd /1) with the https ://doi.org/10.17632 /2s6x2 5xsrd .1
Received: 1 July 2020; Accepted: 14 December 2020
References
 1. The World Bank. Hidden harvest: the global contribution of capture fisheries. 66469-GLB. Washington, D.C. http://docum ents.
world bank.org/curat ed/en/51570 14681 52718 292/pdf/66469 0ESW0 P1210 120Hi ddenH arves t0web .pdf (2012).
 2. Ministry of Fisheries and Aquaculture Development (MoFAD). Fisheries management plan of Ghana: a national policy for the 
management of the marine fisheries sector 2015–2019. Ghana. https ://mofad .gov.gh/wp-conte nt/uploa ds/2016/07/FISHE RIES-
MANAG EMENT -PLAN-OF-GHANA .pdf (2015).
 3. FAO Fisheries & Aquaculture. Fishery and aquaculture country profiles—the republic of Ghana. http://www.fao.org/fishe ry/facp/
GHA/en (2016).
 4. Hicks, C. C. et al. Harnessing global fisheries to tackle micronutrient deficiencies. Nature 574(7776), 95–98. https ://doi.org/10.1038/
s4158 6-019-1592-6 (2019).
 5. Dovlo, E., Amador, K. & Nkrumah B. Report on the 2016 Ghana marine canoe frame survey. Fisheries Scientific Survey Division 
(FSSD), Fisheries Commission (FC)-MoFAD, Republic of Ghana. Information Report No. 36 (2016).
 6. Lai, C.-C., Shih, T.-P., Ko, W.C., Tang, H.J. & Hsueh, P.R. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and 
coronavirus disease-2019 (COVID-19): the epidemic and the challenges. Int. J. Antimicrob. Agents 105924 (2020).
 7. Dietz, L. et al. 2019 novel coronavirus (COVID-19) pandemic: built environment considerations to reduce transmission. mSystems 
https ://doi.org/10.1128/mSyst ems.00245 -20 (2020).
 8. The World Health Organisation. Coronavirus disease (COVID-19) advice for the public https ://www.who.int/emerg encie s/disea 
ses/novel -coron aviru s-2019/advic e-for-publi c (2020).
 9. Center for Disease Control. Corona virus disease 2019 (COVID-19): how COVID-19 spreads. https ://www.cdc.gov/coron aviru 
s/2019-ncov/preve nt-getti ng-sick/how-covid -sprea ds.html (2020).
 10. Bennett, N. J. et al. The COVID-19 pandemic, small-scale fisheries and coastal fishing communities. Coast. Manag. https ://doi.
org/10.1080/08920 753.2020.17669 37 (2020).
 11. Food and Agriculture Organisation. How is COVID-19 affecting the fisheries and aquaculture food systems https ://doi.org/10.4060/
ca863 7en (2020).
 12. Bondad-Reantaso, M. G. et al. Viewpoint: SARS-CoV-2 (the cause of COVID-19 in humans) is not known to infect aquatic food 
animals nor contaminate their products. Asian Fish. Sci. 33, 74–78. https ://doi.org/10.33997 /j.afs.2020.33.1.009 (2020).
 13. Executive Instrument (E.I.) 65. Act 1012—Imposition of Restrictions Coronavirus Disease (COVID-19) Pandemic (No.2) Instru-
ment (EI 65), 2020). Gazetted March 30, 2020. https ://verfa ssung sblog .de/wp-conte nt/uploa ds/2020/05/E.I.-65.pdf (2020).
 14. Okyere, I., Chuku, E.O., Dzantor, S., Adade, R. & Ahenkorah, V. Assessment of the State of Fisheries Governance in Ghana. EU 
funded Far Ban Bo Project. https ://doi.org/10.13140 /RG.2.2.18441 .26721 (2020). 
 15. Kasapa FM Online. Sea water does not protect you from COVID-19—fishermen cautioned. https ://kasap afmon line.com/2020/04/
sea-water -does-not-prote ct-you-from-covid -19-fishe rmen-cauti oned (2020).
 16. Ghana Health Service. Regional distribution of COVID-19 cases. Ghana Health Service. https ://www.ghana healt hserv ice.org/
covid 19 (2020).
 17. Resnik, D. B. & Elliott, K. C. Using drones to study human beings: ethical and regulatory issues. Sci. Eng. Ethics 25(3), 707–718. 
https ://doi.org/10.1007/s1194 8-018-0032-6 (2019).
 18. Baddeley, A., Rubak, E. & Turner, R. Spatial Point Patterns: Methodology and Applications with R (CRC Press, Boca Raton, 2015).
 19. Lincoln, Y. S. & Denzin, N. K. (eds) Turning Points in Qualitative Research: Tying Knots in a Handkerchief Vol. 2 (Rowman Altamira, 
Lanham, 2003).
 20. Jomitol, J., Payne, A. J., Sakirun, S. & Bural, M. O. The Impacts of COVID-19 to Small Scale Fisheries in Tun Mustapha Park, Sabah, 
Malaysia; What Do We Know So Far? https ://doi.org/10.20944 /prepr ints2 02005 .0287.v1 (2020).
 21. Ghanaweb. Coronavirus: Kumasi central market shut down hours after opening. https ://mobil e.ghana web.com/Ghana HomeP age/
regio nal/Coron aviru s-Kumas i-Centr al-Marke t-shut-down-hours -after -openi ng-92903 8 (2020).
 22. He, S. Ethics of drone use for human research. https ://condu ctsci ence.com/ethic s-of-drone -use-for-human -resea rch (2019).
 23. Harris, J. M., Nelson, J. A., Rieucau, G. & Broussard, W. P. III. Use of drones in fishery science. Trans. Am. Fish. Soc. 148(4), 687–697. 
https ://doi.org/10.1002/tafs.10168 (2019).
 24. Okyere, I., Chuku, E. O., Angnuureng, D. B., Asare, N. K. & Aheto, D. W. COVID-19: mapping and baseline survey of fish landing 
beaches, processing sites and markets. The USAID/Ghana Sustainable Fisheries Management Project (SFMP). Narragansett, RI: 
Coastal Resources Center, Graduate School of Oceanography, University of Rhode Island, and Centre for Coastal Management, 
University of Cape Coast, Cape Coast, Ghana. GH2014_COV119_UCC. 21 pp (2020).
 25. Han, J., Zhang, X., He, S. & Jia, P. Can the coronavirus disease be transmitted from food? A review of evidence, risks, policies and 
knowledge gaps. Environ. Chem. Lett. https ://doi.org/10.1007/s1031 1-020-01101 -x (2020).
 26. Huang, Z. et al. Optimal temperature zone for the dispersal of COVID-19. Sci. Total Environ. https ://doi.org/10.1016/j.scito 
tenv.2020.13948 7 (2020).
Acknowledgements
The authors are grateful to the Department of Fisheries and Aquatic Sciences and the Centre for Coastal Man-
agement (Africa Centre of Excellence in Coastal Resilience—ACECoR), University of Cape Coast for providing 
the needed logistics for the research. We are also thankful to the USAID-Ghana Sustainable Fisheries Manage-
ment Project (SFMP) for providing partial funding support for the work (cooperative agreement number AID-
641-A-15-00001). Input from DJM was supported by the CGIAR Research program on Fish Agri-food Systems 
(FISH), funded by contributions from the CGIAR Trust Fund. UAV imagery was possible through the support 
of the BEACHMON Project funded by the National Geographic Society (Grant No. CP-107T-17).
Author contributions
I.O. and E.O.C. designed the work, acquired data-sets, wrote the manuscript and coordinated the research; B.E. 
and D.B.A. designed the work, acquired and analysed the remote sensing and GIS data-sets; J.K.B.-A. designed 
the work and wrote the health related sections of the manuscript; R.B., D.J.M., N.K.A., D.W.A. and B.C. designed 
the work, wrote different sections of the manuscript, reviewed the manuscript and supervised the research.
13
Vol.:(0123456789)
Scientific Reports |        (2020) 10:22407  | https://doi.org/10.1038/s41598-020-79898-4
www.nature.com/scientificreports/
Competing interests 
The authors declare no competing interests.
Additional information
Supplementary Information The online version contains supplementary material available at https ://doi.
org/10.1038/s4159 8-020-79898 -4.
Correspondence and requests for materials should be addressed to I.O.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and 
institutional affiliations.
Open Access  This article is licensed under a Creative Commons Attribution 4.0 International 
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or 
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the 
Creative Commons licence, and indicate if changes were made. The images or other third party material in this 
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the 
material. If material is not included in the article’s Creative Commons licence and your intended use is not 
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from 
the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.
© The Author(s) 2020