Do traditional practices and wildlife values influence ape presence near communities in eastern DR Congo?

Raised of $4,500 Goal
Ended on 6/12/21
Campaign Ended
  • $1,130
  • 26%
  • Finished
    on 6/12/21



Three villages will be selected based on their proximity to ape habitat, accessibility and willingness to participate. Sociocultural data will be collected from members of selected households, traditional leaders, hunters and cultivators using questionnaires and conducting interviews. Ecological data will be collected by conducting reconnaissance walks to gather direct and indirect signs of both apes and humans in near and far sample areas in forest areas bonding the villages. Camera traps will be placed between the near and far forest sample areas to check clips for apes. Specifics about the described methods are below.

Household Member Demographics: Data will be collected from household members including: ethnic group, tribal affiliation, name of village, estimated distance from the forest, estimated distance to the protected area, number of people in the household, main sources of income and main dietary sources. In addition tribal identity, age, gender, level of education, employment (if any) of participants and how long they’ve lived in the village will also be collected.

Attitudes Towards Apes:  Open-ended questions will be used to assess attitudes towards apes and include questions about if, when and where participants have seen apes, how they feel when they see them, and if they know anything about their behavior. 

Traditional Leaders will be selected using purposive sampling as they have specialized knowledge of local traditions. Four traditional leaders will be selected per village/cluster (total of 16). A research assistant will conduct semi-structured interviews to learn ape related traditional practices, if they are passed on, if any animals are considered sacred and if they would be willing to share these stories with others outside the village. 

Reconnaissance Walks will assess the presence and proximity of apes near villages. Stratified sampling will be done in two forested areas (near and far) from each of the study sites (villages). Ten recce walks will be conducted in the near-village sample area at 0 km to 1 km distance from the village. The ‘near’ recce walks will start at the edge of the village towards forested areas and adjacent protected area. Ten recce walks will also be done in the far-village sample area at 4 km to 5 km distance from the village. This 1 km ‘far’ recce band will be closer to the adjacent protected area. Recce walks will be repeated seven weeks later, starting and ending in the same location to account for new signs and for seasonal variations (Campbell et al., 2016). Data recorded during recce walks will include direct and indirect signs of apes, direct and indirect signs of humans, habitat type, slope, dominant tree species and man-made features. Indirect signs of apes (distinguishing gorilla from chimpanzee when possible) include nests (location, quantity, age) foot/knuckle trace, feces, and food remains. Direct and indirect human signs will include visual observations of encounters with people, human footprints, shell casings, snares, hunting camps and rubbish. Indirect signs of large mammals will also be recorded such as visual sightings, feces and foot traces. Habitat type and slope will be recorded every 250 meters. Habitat type categories are primary forest, secondary forest, swamp, bamboo and cultivated (Basabose, 2005). Man-made features such as roads, mines, logging and hunting camps, etc. Human disturbance and potential ape response to human activities will also be documented such as species hunted, where hunting takes place, crop-raiding occurrences, and presence of roads and fences.

Camera Traps  (24) will be placed in the near and far sample area bands and the area in between the bands. The camera trap grid will be 4 rows of 6 traps. Each row will be placed at 0 km on the edge of the village and every 1 km after, with the camera placed at 5 km from the village. Cameras will overlap with the recce near and far sample areas to get a continuous distance gradient of ape and large mammal presence to compare with recce data. Cameras will be checked every two weeks for 5 weeks. 


As far as data challenges It is possible that the traditional knowledge of villagers that I hope to gather may no longer be relevant. It is also possible that I may not find signs of apes, or very few due to population decline or because they are staying far away from the villages. This information will still be useful. Heavy rains may also delay data collection or make it difficult to see. I will try to adjust the timeline a few days to account for especially rainy days. Getting to the field sites/villages may also be challenging due to poor roads and weather. Alternative field sites may have to be chosen or eliminated from the data collection plan. 

Pre Analysis Plan

Hypothesis: Apes are presence at closer distances (0-5 km) to villages that maintain traditional practices such as hunting taboos or have positive association with apes.

Alternative hypothesis 1: Apes are not within 0 - 5 km of villages that maintain traditional practices such as hunting taboos or other traditions that regard apes.

Alternative Hypothesis 2: Traditional practices are not consistently or at all so do not appear to influence ape presence near or far from communities. 

Recce walk data will be analyzed using general or generalized mixed effects (with recce as a random effect) models in R statistical package to assess relationships between 1. Ape sign frequency of occurrence every 250m (numeric response) and habitat type (predictor categorical), 2. Ape sign frequency of occurrence every 250m (numeric response) and human sign frequency of occurrence every 250m (numeric predictor) with recce as a random effect, 3. Ape sign frequency of occurrence every 250m (response numeric) and dominant tree species (predictor categoric) with recce as a random effect. 4. Compare ape and human sign frequency of occurrence in near and far sample areas (numerical). Distance from village will also be included in all models as well as a random effect for recce ID to account for spatial autocorrelation. The type of error structure to use in the model (logistic, negative binomial, or linear regression) will be determined based on the distribution of the response variable. 

Camera trap data will be used to validate recce data and to evaluate a distance gradient of ape and large mammal presence from focal villages and to analyze the relationships between: a) ape and other large mammal captures with potential ape presence (based on the frequency of occurrence of direct and indirect signs of apes found in the recce walk data) and b) ape captures and hunter and cultivator stated locations of apes (marked on participatory map). For each of the 24 cameras, data will be quantified including a) number of trap nights, b) frequency of focal species captures (gorillas, chimpanzees, other large mammals) in each camera (every 1km) and c) Frequency of captures of ape species and large mammals in each habitat type. These will be analyzed using Generalized Linear Mixed Effects Modeling with each camera as the sampling unit and recce as a random effect. The number of chimpanzee, gorilla and large mammal 30-second video clips captured each trap day will be the response variable. Habitat type, human-made features and frequency of occurrence of human signs at each camera station will be co-variate predictor variables. Detection probability will be estimated by calculating the frequency of all detections per camera every two weeks divided by total number of detections of all cameras in a sample forest site. Detection probability will be calculated the same for each of the village forest sites. Cameras will be placed for five weeks at each site. Capture frequencies of chimpanzees, gorillas and large mammals in each habitat will also be calculated (Figure 4). Analysis of camera traps and recce walk data will be combined using GIS mapping of recce way-points and camera trap location and frequency of large mammal captures, correlation regression and mixed effects modelling. 

Sociocultural Data      

The household questionnaire data will first be analyzed separately using descriptive statistics. Frequency tables will be made to calculate: 1. The number of males and females according to their age. 2. Main food source/s per household 3. Education level per household. 4. Main source/s of income per household. Percentages, median, mode will be calculated from frequency tables. Age categories are: 0-5, 6-10, 11-15, 15-20, 20-30, 31-50, >50. Education levels categories are: None,  1-5 years, 6-10 years, more than 10 years. Main source of income categories are: Forest plants, forest animals (type), domestic meat (type), cultivated plants (type). Logistic regression will be used to analyze the relationship between demographics and attitudes. The demographics of age, gender, tribal affiliation, education level, main food source and main source of income will be used as categorical predictor variables for attitudes towards apes and attitudes towards conservation.

Attitudes towards apes: The participant responses will be coded and classified by common use of words or emotions and then put into two categories of positive and negative according to gender, age, tribal affiliation and ethnic affiliation. Each attitude category will be defined to clarify categorization. These response categories will be analyzed using descriptive statistics, narrative summary, ANOVA, Chi Square and Mixed Effects Model.

Data collected from the semi-structured interviews with hunters/cultivators and traditional leaders include audio recordings and written notes. The notes will be annotated, and the recordings will be transcribed. Both will be coded according to basic themes. Themes from the interviews with traditional leaders will be placed in categories according to wildlife tradition type: story/myth, totem, taboo, dance, medicine. Once the themes are indexed in these main categories, the traditions will be further divided into two other secondary categories: 1) traditions that protect or exploit wildlife (protection/exploitation) and 2) traditions that are used currently or not. (current/past). The frequencies of the types of traditions will be calculated. The frequencies of the secondary categories will also be calculated. These data will be used to compare between sites and as predictor variables to assess the relationship between ape presence and knowledge and culture (response variable). These data will be analyzed using descriptive statistics, narrative summary, ANOVA, Chi-square, mixed effects model.


This project has not yet shared any protocols.