Supporting and consolidating the Participatory Experimental Network (REP)

  1.  → 
  2. Cacao Forest in Dominican Republic
  3.  → Participatory Experimental Network

What's the REP ?

Réseau Expérimental Participatif
In the Dominican Republic, cacao trees are cultivated together with other plants. It is common practice for Dominican producers to plant a variety of species on their plots. This cultivation method, which encourages biodiversity, is called agroforestry when at least one of the associated cultivated species is a woody perennial plant. This subject lies at the heart of our project. Agroforestry is a way of diversifying production and harvesting a wide variety of fruits, vegetables and timbers for consumption by the producer’s family and their animals, or for sale. In addition, this diversity of cultivated plants is very beneficial to the ecosystem balance.

The Participatory Experimental Network (REP) is essentially a three-stage scientific experiment that draws on the local knowledge of cacao growers and their families to optimise the benefits of agroforestry.

Stage one

The first stage involved bringing together groups of producers from different regions around the country to develop new cacao cultivation models in a bid to improve the income generated by the crops that the farmers grow on their plot.

Stage two

The second stage consisted of setting up the REP and making use of each model established on the farms of the producers taking part, under an experimental framework that includes repeated use of models and reference parcels.

Stage three

The current stage, which involves monitoring the economic, agronomic and ecological parameters of each model in the local context and documenting every change made to each starting model, through an ongoing dialogue with the producers concerned.
The REP can be described as a scientific trial that involves the producers’ families and which makes use of the more than 60 plant species that are currently being cultivated in Dominican plantations to create and trial innovative agroforestry systems that generate higher incomes for farmers.
Réseau Expérimental Participatif

But how did this experimental process come about, and what are the different agroforestry models?

Between 2014 and 2017, at the request of several French chocolate makers that buy their cocoa in the Dominican Republic, the French Agricultural Research Centre for International Development (CIRAD) carried out an audit of Dominican cacao plantations and compiled a record of cocoa production in four provinces: San Cristóbal, Duarte, Hato Mayor and Seibo.

This audit revealed five main characteristics of Dominican cocoa production:


The great majority of farms are no bigger than three hectares and focus on cacao cultivation


All the parcels under cacao cultivation use agroforestry and are thus home to a wide diversity of cultivated plants grown alongside the cacao trees. More than 60 different species have been identified


The Dominican plantations are not very profitable. Much of the produce grown on them is not used and left to rot


More than 70% of producers are aged over 60, and the younger generation is reluctant to take over


The level of education and technical expertise among producers is low

This audit was of great interest (and concern) to the six French chocolate producers (Valrhona, Weiss, Révillon, Voisin, Carambar and AlterEco) and the pâtisserie industry association Relais Desserts, raising three main questions:
How do we associate the benefits of the great biodiversity that characterises Dominican cocoa with its image?
Will the income generated by Dominican plantations ensure the continuation of cocoa production in the medium to long term?
How much certainty is there regarding the long-term survival of the aromatic quality of Dominican cocoa?
Can it be preserved as it is today?
It is evident that the future of Dominican cacao cultivation depends on its capacity to generate sufficient income for cocoa producers. If this happens, younger producers will be more willing to take over cocoa production and cocoa farming will acquire a better image. It was this context that gave rise to the Participatory Experimental Network set up by the Cacao Forest project. This scheme, which is essentially an experiment in cultivated biodiversity on Dominican cacao plantations, consisted of two phases:
The first phase, which included more than 30 workshops held over a period of a year, took place between 2017 and 2018 and consisted of a participatory scheme involving groups of producers and technical experts in cacao cultivation who developed four agroforestry models combining a wide variety of crops in a bid to generate a higher income. These four models are described in the REP’s technical and scientific guide, which was produced at the end of phase one.
The second phase consisted of applying these models on the ground and implementing them on the farms of producers who had signed up to the scheme, where they could be compared with other models in the scheme and with the ‘control’ reference models – simple cacao cultivation systems combining cacao with just banana, and traditional cacao-growing plots.

Two aspects of the Participatory Experimental Network are original and unique:

1. The producers are involved

This experimental scheme is being carried out on producers’ farms and not in a ‘controlled environment’. Participating producers sign an agreement with the project in which they undertake to allocate an area on their farm where the experiment will be conducted on one or two models of their choice. In this agreement, they undertake to adhere to the set-up protocol on each parcel and to provide the labour required to complete the project tasks. In addition, the producers support the Cacao Forest technical team by keeping a record of their income (harvests) and expenditure (input costs and labour) relating to the parcels concerned. For its part, the project undertakes to help with technical assistance, the agricultural and scientific training required to collect the data, and the resources required to conduct the experiment (young plants, fertiliser and tools).

2. The REP is tailored to their reality

This network is flexible because the Cacao Forest project is designed to adapt the established models to take into account the reality of life in rural parts of the Dominican Republic and the various challenges that producers face, which range from treating disease, managing shade and maintaining their plots, to marketing the cocoa and associated products they produce. Every alteration made to a model is documented and recorded. The models developed are given a period of between three and five years to collect enough data to obtain precise information about income, costs and benefits, enabling them to be assessed from an economic, agronomic and ecological point of view.

Agroforestry models

The agroforestry models are the focal point of project on which the REP is based. Each model associates different varieties of cacao with a combination of cultivated plant species, all of which are selected by the producers and technicians taking part in the set-up workshops. In the province of San Cristóbal, two models have been proposed by the producers and technicians.

These models are based on a total of eight varieties of cacao, four of which are allocated to Model No.1 (ICS-39, ML-4, ML-103 and UF-221) and four to Model No.2 (GS-36, ICS-95, ML-22 and UF-676). In addition, these models include a total of 16 crop species grown alongside the above-listed cacao varieties. Nine of these species are grown in Model No.1 (avocado, breadfruit, coconut, soursop, mandarin, bitter orange, banana, cocoyam and sapote), and 10 species are grown in Model No.2 (avocado, African mahogany, ginger, castor oil plant, Persian lime and creole lime, gmelina, banana, creole oak and cocoyam). The producers and technicians in Duarte province have proposed two other models, based on a total of 15 varieties of cacao, 10 of which are used in Model No.3 (ICS-1, ICS-39, IML-44, IML-53, ML-66, ML-103, RZ-12, RZ-44, RZ-83 and RZ-100) and seven in Model No.4 (ICS-39, ICS-95, IML-53, IML-119, ML-22, ML-105 and UF-296).

In addition, the models include a total of 13 crop species grown alongside the above-listed cacao varieties. Eight of these species are grown in Model No.3 (avocado, achiote, coconut, sweet orange and bitter orange, Cuban piñon, banana and cocoyam), and nine species are grown in Model No.4 (avocado, turmeric, ginger, pepper, Cuban piñon, banana, sweet and bitter orange, and cocoyam).

Lastly, each experiment must be accompanied by a control that serves as a reference against which to compare the results of the four models, not only between each other but also against (i) the cacao cultivation model traditionally used in each province (known as the business-as-usual – BAU – model), and (ii) the conventional cacao cultivation model as recommended in most technical guides to Latin America: the simple agroforestry model, which combines cacao with plantain.

Model application methods

Two methods are used to evaluate each model on the ground:


The development method (F), in which the model is created on new agricultural land that meets three criteria: the previous land use must not have been for forest of any kind; the parcel must be easily accessible and in an ecological zone appropriate for cacao cultivation; and the soil characteristics must be suitable for cacao cultivation.


The rehabilitation method (R), in which an old and/or unproductive cacao plantation is turned into an agroforestry model of the producer’s choice through a gradual process lasting for three to four years. The old cacao grove selected must meet two criteria: the parcel must be easily accessible; and the parcel’s productivity (cacao and associated crops) must be low.

For each method, three identical set-ups of each model are created on the producers’ farms. Lastly, in each province, the following are set up:

3 identical parcels of the simple agroforestry control model comprising all of the cacao varieties present in the models set up in that province (Models 1 and 2 in the province of San Cristóbal, and Models 3 and 4 in Duarte and Hato Mayor provinces), along with plantain;
3 identical BAU (business-as-usual) control parcels, consisting of the agroforestry plantations typically found in that province.

The participatory experimental scheme

Each experimental parcel set up on the farm of a producer who has joined the scheme covers an area of 0.25 hectares. To date, a total of 54 parcels have been set up, allocated as shown in the table below:
Province Method Agroforestry
No. of parcels
in Dec. 2019
No. of parcels
in Dec. 2020
No. of parcels
in Dec. 2021
SAN CRISTOBAL Development No.1 3 1 3
No.2 3 1 3
Control 3 1 3
Rehabilitation No.1 3 1 3
No.2 3 2 3
Control 3 2 3
DUARTE Development No.3 3 2 3
No.4 3 2 3
Control 3 3 3
Rehabilitation No.3 3 3 3
No.4 3 3 3
Control 3 3 3
TOTAL 36 24 36
P: number of producers involved 17 15 23
C: number of producers joining – leaving 17-0 7-9 12-4

Future possibilities and scope for international cooperation

The Participatory Experimental Network is of interest to French international cooperation organisations including the Agence Française de Développement (AFD) in the context of its work supporting public and private institutions in the Dominican chocolate industry. The AFD, along with French and Dominican chocolate makers, are very interested in the results and the data produced by the REP in helping to introduce a future programme aimed at rehabilitating and renovating cacao plantations in the Dominican Republic.
Innover ensemble pour la cacaoculture de demain

Subscribe To Our Newsletter

Abonnez-vous à notre lettre d'information et tenez-vous informé des dernières innovations dans le domaine de la cacaoculture.

Votre inscription a été prise en compte !

Innover ensemble pour la cacaoculture de demain

Subscribe to our newsletter

Join our mailing list to receive the latest news and updates from our team.

You have successfully subscribed! Please check your inbox to confirm your subscription.