Rodents represent an estimated 40% of mammalian species and are important components of small mammal assemblages globally. In sub-Saharan Africa they are significant contributors to agricultural crop loss alongside being hosts for a number of zoonotic pathogens of importance to local human populations. Trapping of individual rodents, identification to species and assays for the presence of potential zoonotic pathogens produces important data for understanding the potential ecological and human health impact of changing small mammal populations.
This scoping review aimed to synthesise the research conducted in West Africa specifically to; a) identify the aims of rodent research and summarise methodological approaches b) identify the locations, habitats and rodent species that have been studied, c) identify what potential pathogens are tested for and which host species they are found in.
A scoping review of all rodent trapping studies identified through a systematic search of online databases for studies conducted within the United Nations West Africa sub-region was performed.
4,282 records were identified from the initial search with 124 studies published between 1974 and 2021 included in narrative synthesis. Studies were conducted in 14 West African countries. The majority of studies investigated rodent ecology (56%) with the remainder investigating potential zoonotic diseases. Study methodology was comprehensively reported from 28 (23%) studies. Specific trapping habitats were not reported using standardised approaches. The included studies reported 73,164 trapped small mammals, predominantly from the order Rodentia. These individuals were identified to 147 species with Mastomys natalensis, Rattus rattus and Mastomys erythroleucus the most common. 55 studies investigated potential zoonotic pathogens, 7 further studies investigated rodent pathogens. Thirty-two microorganisms were tested for including viruses, bacteria and parasites. Mastomys natalensis was found to be most commonly infected by Lassa mammarenavirus and Bartonella sp, Arvicanthis niloticus was most commonly infected with Borrelia sp and Mus musculus was most commonly infected with Toxoplasma gondii.
This scoping review identified 124 studies, conducted in 14 countries at 1,193 study sites trapping 73,164 small rodents. Current research to describe the rodent populations of West Africa have provided some information about the distribution of species across this heterogenous region and the distribution of potential pathogens within individual species. We make recommendations for subsequent research to enable wider data sharing and re-use to to support meta-analysis of obtained data to address questions on rodent ecology and zoonotic disease risk.
Rodents (Rodentia) are abundant and diverse, representing around 40% of all mammalian species (American Society of Mammologists 2021). Rodents typically demonstrate ‘fast’ life history strategies characterised by early maturation, short generation times, low juvenile and adult survival and high fecundity (Dobson and Oli 2007). Between-species heterogeneity exists, with commensal species and those that are reservoirs of zoonoses having traits consistent with faster life histories (Han et al., 2015). Rodent species compete with humans for food produced in agricultural systems (Fiedler 1988), 5-10% of rodent species are classed as major crop pests. Commensal species thrive in human adapted landscapes nesting in houses and scavenging human food storage and preparation areas. In sub-Saharan Africa, Mastomys spp. and Arvicanthis spp. are two species causing significant burden as agricultural pests (Stenseth et al. 2003). Crop loss within Afro-Malagasy small-holder farming communities is high, with 15% of pre-harvest crop being lost to rodent pest activity (Swanepoel et al. 2017).
Alongside crop loss a further risk to local communities is the transmission of infectious diseases from rodents to humans. Zoonotic diseases (i.e. animal pathogens or microorganisms residing in animals that impact human health) are transmitted via two pathways – a direct and an indirect pathway. The direct transmission pathway involves transmission of pathogens through a rodent biting or scratching a human, or through contamination of food or water sources. The indirect pathway involves rodents acting as amplifying hosts for ectoparasites that themselves are vectors of zoonoses (Meerburg, Singleton, and Kijlstra 2009). Rodent-borne pathogens vary in their scale from those with global distribution, such as the bacteria Borrelia burgdorferi (causing Lyme disease) and Leptospira spp. (casing Leptospirosis) and parasites such as Toxoplasma gondii (causing Toxoplasmosis) and Leishmania spp. (causing visceral and cutaneous Leishmaniasis). However, the majority of rodent-borne pathogens have more limited geographic distributions, viral diseases such as Hantavirus pulmonary syndrome (caused by Hantaviridae) are typically concentrated in the Americas and Lassa fever (caused by Arenaviridae) is limited to West Africa (Mills 1998).
These economic and public health burdens have led to the monitoring of rodent distributions being performed across multiple scientific disciplines, including (but not limited to) conservation science, development studies and infectious disease epidemiology. The distribution of rodent populations can be effectively monitored through rodent trapping, with traps deployed in habitats expected to contain rodent species of interest.
Further, there is increasing interest in the understanding the associations between environmental, human activity and rodent populations with land use change. A recent review of studies included in the PREDICTS database (Hudson et al. 2014) identified that land use change from primary (i.e. intact forests) to managed and urban systems is associated with a change in rodent assemblages towards species that are more likely to be hosts of zoonoses (Gibb et al. 2020). Mastomys natalensis populations (an important pest and zoonotic reservoir species) in Tanzania, East Africa are sensitive to seasonal meteorological cycles, in years with below average rainfall and short rainy seasons population density is observed to be lower (Makundi, Massawe, and Mulungu 2007) with reduced pest activity. Future climate change projections include longer and wetter West African monsoon seasons which could lead to increased zoonotic disease host rodent populations (Akinsanola and Zhou 2019). As West Africa contains several known rodent zoonotic reservoirs (e.g. Mastomys natalensis and Arvicanthus niloticus) and several potential reservoirs of zoonotic diseases (e.g. Lemniscomys striatus), this scoping review will be limited to considering the evidence produced from West Africa.
To date there has been no comprehensive overview of rodent trapping studies conducted in West Africa. The aim of this scoping review is therefore to systematically map the research conducted in this area, to identify currently used methods, gaps in current knowledge and to synthesise the knowledge on rodent distributions throughout West Africa obtained from included trapping studies.
We specifically aim to address the following research questions:
A scoping review method was adopted to identify and map the available information on rodent assemblages, abundance and diversity across West Africa. This is a recommended approach when examining how research is conducted within a topic area (Munn et al. 2018). The review followed the PRISMA extension for Scoping Review guidance (Tricco et al. 2018). The protocol for this review was not pre-registered.
The following terms as keywords were searched for in OVID Medline, Web of Science (Core collection and Zoological Record), JSTOR, BioOne, African Journals Online, Global Health and the pre-print servers BioRxiv and EcoEvoRxiv:
Other resources including the UN Official Documents System, Open Grey, AGRIS FAO and Google Scholar were searched using combinations of the following terms:
Additional articles were identified through references within included articles and reports known to the study team. No time constraints were placed on searches. Searches were completed on 2021-03-01. The exploded search terms and links to the respective web portals are provided in supplementary material.
One reviewer screened titles, abstracts and full texts against the inclusion and exclusion criteria. A random subset of each of these (10%) were reviewed by a second reviewer.
Data from eligible studies were extracted using a standardised data abstraction tool designed for this study on a subset of eligible studies and subsequently refined. Data were extracted by a single reviewer. A second reviewer verified a random subset (10%) of included studies. A data extraction form was maintained on a Google sheets document. Data extracted included i) study identifiers; ii) study aims; iii) trapping methodology; iv) geolocation data; v) method of speciation; vi) trapping locations and dates; vii) trapped species; viii) number of trap-nights and ix) microorganisms of interest. The data extraction forms and data dictionaries used during the analysis are reproduced in the supplementary material.
Explicitly stated study aims were extracted, for those with no specified aims the aims were inferred from the introduction or conclusions of the manuscript or report. The aims of studies were categorised to a higher order grouping, namely, rodent ecology research or studies on the risk of zoonoses.
GPS locations were extracted for the most precise location presented i.e. trap, trap-line, study-site or study region. Coordinates were extracted in the format reported and converted to decimal degrees. Where no GPS location was reported coordinates were matched to study locations using the National Geospatial Intelligence Agency NGA GEOnet Names Server (National Geospatial-Intelligence Agency n.d.) based on study site name/region and comparison to maps presented in the manuscript (when available).
The brand, if reported, or the description of the structure of the rodent trap was obtained from each study. For studies using multiple devices all types were recorded. The method employed when setting the traps was also extracted, studies that reported setting a transect line were classified as such, for studies reporting using a line of traps but not through a specified habitat the term trap line was used.
The trapping effort in trap-nights (the setting of a single trap for a single night) was recorded. If this was reported at the habitat level or the level at which the study reported rodent captures it was classified as a complete recording of trapping effort. For studies that reported trapping effort at different levels to the capture level the recording of trap-nights was classified as incomplete. For studies reporting number of nights trapping occurred for or for where the number of traps set was stated but the number of nights not recorded they were classified as incomplete.
The habitat classification scheme a study used was recorded. For studies not using standardised recording of habitat types the explicit description from the study was associated with the trap. For studies reporting multiple habitat types for a single trap, trap-line or trapping grid a higher order classification of habitat type was recorded.
Species classifications may vary over time as morphological and genetic information emerges, to handle potential species classification synonyms and reclassification the reported names of trapped species was mapped to the Global Biodiversity Information Facility (Global Biodiversity Information Facility n.d.) identification code. For species with updated taxonomies the reported species identity was converted to the accepted species name for all subsequent analyses. Where a subspecies is reported by study authors the species complex it belongs to will be used for subsequent analysis.
The presence and absence of trapped individuals and genus/species will be extracted. For studies reporting on all trapped individuals (i.e. not those only reporting on the presence of specific species of interest) the absence of a reported capture of a species recorded elsewhere in the study will be explicitly recorded as an absence at the study location.
In studies investigating rodents for potential zoonoses all pathogens tested for were extracted. The number of rodents tested and the number of positive or negative samples were extracted alongside the type of assay used to detect the pathogens (e.g. Polymerase Chain Reaction (PCR), Enzyme Linked ImmunoSorbent Assay (ELISA) or viral culture). Where possible pathogens were identified to the species level, however, where an assay only allows for attribution to a family of viruses or bacteria the higher order grouping was used (i.e. PCR using a non-specific arenavirus primer).
No formal risk of bias assessment tool was used.
Descriptive analysis of included studies was conducted in R (v4.0.2) (R Core Team 2020) within the RStudio IDE (RStudio Team 2020). The analysis code and packages used for this analysis are available on GitHub repository.
The locations of traps, trap-sites and studies were assigned to level 2 administrative areas in West Africa using shapefiles maintained by GADM (Database of Global Administrative Areas n.d.).
A total of 4,282 records were identified, with 124 studies included in narrative synthesis (see Figure 1). A summary table of the included studies and the citations of the study or report are presented in supplementary table 1. The earliest studies identified were from 1974 with the majority (61%) published since 2010. Most of the included studies were classified as rodent ecology research (56%) with the remainder of studies on zoonotic diseases (44%).
Table 1 summarises the included studies by publication year, and methodology with grouping by study aim.
|Characteristic||Overall, N = 1241||Ecology, N = 701||Zoonoses risk, N = 541|
|Benin||7 (5.6%)||3 (4.3%)||4 (7.4%)|
|Burkina Faso||1 (0.8%)||0 (0%)||1 (1.9%)|
|Cabo Verde||1 (0.8%)||0 (0%)||1 (1.9%)|
|Côte D'Ivoire||5 (4.0%)||4 (5.7%)||1 (1.9%)|
|Ghana||18 (15%)||13 (19%)||5 (9.3%)|
|Guinea||14 (11%)||7 (10%)||7 (13%)|
|Guinea Bissau||1 (0.8%)||1 (1.4%)||0 (0%)|
|Liberia||3 (2.4%)||3 (4.3%)||0 (0%)|
|Mali||8 (6.5%)||5 (7.1%)||3 (5.6%)|
|Mauritania||2 (1.6%)||1 (1.4%)||1 (1.9%)|
|Multiple||10 (8.1%)||6 (8.6%)||4 (7.4%)|
|Niger||4 (3.2%)||3 (4.3%)||1 (1.9%)|
|Nigeria||17 (14%)||9 (13%)||8 (15%)|
|Senegal||24 (19%)||11 (16%)||13 (24%)|
|Sierra Leone||9 (7.3%)||4 (5.7%)||5 (9.3%)|
|1970-79||7 (5.6%)||4 (5.7%)||3 (5.6%)|
|1980-89||4 (3.2%)||2 (2.9%)||2 (3.7%)|
|1990-99||9 (7.3%)||4 (5.7%)||5 (9.3%)|
|2000-09||28 (23%)||22 (31%)||6 (11%)|
|2010-19||62 (50%)||31 (44%)||31 (57%)|
|2020-||14 (11%)||7 (10%)||7 (13%)|
|Repeat study visits - Yes||41 (33%)||29 (41%)||12 (22%)|
|Level of geolocation|
|No geolocation||6 (4.8%)||1 (1.4%)||5 (9.3%)|
|Specific trap site||11 (8.9%)||9 (13%)||2 (3.7%)|
|Study||31 (25%)||22 (31%)||9 (17%)|
|Study site||76 (61%)||38 (54%)||38 (70%)|
|Method of speciation|
|Molecular||9 (7.3%)||7 (10%)||2 (3.7%)|
|Morphological||65 (52%)||40 (57%)||25 (46%)|
|Morphological and molecular||37 (30%)||19 (27%)||18 (33%)|
|Not stated||13 (10%)||4 (5.7%)||9 (17%)|
|Use of a species accumulation curve - Yes||14 (11%)||14 (20%)||0 (0%)|
|Use of a measure of species diversity - Yes||26 (21%)||24 (34%)||2 (3.7%)|
|Reporting of trapping effort|
|Incomplete||43 (35%)||19 (27%)||24 (44%)|
|No||39 (31%)||15 (21%)||24 (44%)|
|Yes||42 (34%)||36 (51%)||6 (11%)|
|Reporting of potential pathogens|
|No||62 (50%)||62 (89%)||0 (0%)|
|Rodent pathogen||8 (6.5%)||7 (10%)||1 (1.9%)|
|Yes||54 (44%)||1 (1.4%)||53 (98%)|
The detailed aims of the included studies are shown in Figure 2. Biodiversity of rodents was a shared aim between the two themes with 37 ecological studies and 2 zoonoses studies investigating this.