Previous PageTable Of ContentsNext Page

Modelling the effect of temperature, soil moisture and photoperiod on growth and development of bambara groundnut (Vigna subterranea (L.) Verdc.): BAMGRO model

Asha Karunaratne1, Sayed Azam-Ali1, Abu Sasey2, Hans Adu-Dapaah3 and Neil Crout1

1 Division of Agriculture and Environmental Science, School of Biosciences, University of Nottingham, Nottingham, NG7 2RD, UK , Email
Botswana College of Agriculture, Private bag 0027, Gaborone, Botswana, Email
Crop Research Institute, Kumasi, Ghana, Email


The model described in this paper -BAMGRO, integrates data from contrasting landraces and locations in Africa and UK. The model is based on the established CROPGRO model taking into account some features of previous bambara groundnut models, BAMnut and BamGro. To explore the potential productivity of bambara groundnut landraces in specific agro-ecological environments, it is necessary to understand the link between physiological and environmental factors in the capture and use of resources by contrasting landraces. A robust crop model that links the growth and development of landraces to environmental factors will help to identify suitable landraces and cultural practices for specific environments. This model predicts the effect of drought, heat and cold stress independently and collectively on the growth, development and yield of bambara groundnut landraces. In addition, BAMGRO estimates the effects of photoperiod on growth and development. Whilst its predecessors modelled bambara groundnut as a species, BAMGRO identified landrace responses to major abiotic stresses.

Key words

BAMGRO model, landraces, heat stress, cold stress, drought stress


The leguminous crop bambara groundnut (Vigna subterranea (L.) Verdc.), is an indigenous, underutilized secondary food crop in semi-arid Africa. It produces protein rich seeds which are eaten unripe or ripe. In the absence of established varieties, marginal and subsistence farmers in Africa grow locally adapted landraces. Although there are many growth simulation models for a range of major crops, few attempts have been made to model underutilised species for which factors controlling growth and development are not well understood and the general literature is sparse. Quantification of the influence of temperature, soil moisture and photoperiod on growth and development with a suitable crop model will help agronomists in selection of landraces for their specific environments and management decision making.


Model data sets: The parameters and relationships needed to build the functions in the model are derived from various field experiments in Africa and controlled environment experiments at Tropical Crops Research Unit (TCRU), University of Nottingham, UK (Table 1). The details of experimental design, plant sampling procedure, irrigation treatments and standard measurements for TCRU experiment have been previously explained in (Mwale et al., 2007). Over the summer months of 2006 (April to September 2006), two bambara groundnut landraces UniswaRed (Swaziland) and S19-3 (Namibia) were grown in five glasshouses with each house having UniswasRed plot and S19-3 plot under controlled temperature regimes. Two temperatures 2350C (LT) and 3350C (HT) imposed to the five glasshouses randomly. Soil moisture in each house was non-limiting with weekly irrigation to field capacity up to harvesting. The treatments were allocated according to split-plot design that combined two bambara groundnuts landraces and two different temperature regimes. The same experimental protocol was repeated over summer 2007, with non-limiting irrigation only up to 77 days after sowing. BAMGRO was validated for field trials at Botswana College of Agriculture, Gaborone, Botswana and Crop Research Institute, Kumasi, Ghana. The model validation for different photoperiod levels was done using in field sites of Botswana and Ghana by adjusting the sowing dates to capture different photoperiod levels.

Table 1. Summary of experiments used for model data sets


Location and year

Major abiotic stress

TCRU (Glass house)

University of Nottingham, UK (2003)


TCRU (Glass house)

University of Nottingham, UK (2006)

Heat and cold

TCRU (Glass house)

University of Nottingham, UK (2007)

Heat, cold and drought


Botswana College of Agriculture (2006-2007)

Heat, cold, drought and photoperiod


Crop Research Institute, Ghana (2006-2007)

Heat, cold, drought and photoperiod


University of Nottingham, UK (2008)

Various stress

Model description

The new crop model developed in this work is BAMGRO and is based on the established CROPGRO model (Boote et al. 2002a). The features of previous bambara groundnut models, BAMnut (Azam-Ali et al. 2001, Bannayan 2001a) and BamGro (Cornelissen 2005) are considered in BAMGRO which offers a significant improvement to capture landrace variability due to major abiotic stress factors.

BAMGRO consists of different components or sub modules that deal specifically with plant development, crop growth, the soil water stress, temperature stress and photoperiod. The crop growth component, predicts different growth and developmental stages. The temperature module calculates the thermal time and phenochrons for the developmental processes of the crop using weather data and cardinal temperatures. Quantitative description of shoot development, is calculated using the phenochron concept as described by (Mathews and Stephens 1998a).The carbon balance includes daily inputs from photosynthesis, conversion of carbon into crop tissues and losses due to abscised parts. The simulation of growth includes leaf addition, leaf area expansion, pod addition and senescence. The main time step in BAMGRO is one day but thermal time and canopy level photosynthesis are calculated hourly. The photoperiod module uses the response planes approach (Brink, 1997). The soil water module explains the root growth, distribution and water uptake on daily basis under variable climatic conditions. The model links the size and distribution of root system to the capture of water over the growing period (King et al. 2003). The particular feature of BAMGRO-soil water module is its reliance of minimum number of parameters (each with a clear biological meaning).The soil water balance calculates inputs through rainfall and irrigation and various means of water losses through the system. The main inputs are daily weather data: minimum temperature, maximum temperature, solar radiation and rainfall in a text file in standard DSSAT format, soil characteristics: as bulk density, soil texture and landrace coefficients. This new approach to model bambara groundnut responses to major abiotic stress factors provides a platform for easily incorporating other biotic and abiotic factors and extending the model to more landraces. The major parameters used in BAMGRO are explained in Table 2.

As explained in Robertson et al., (2001), phenology is simulated from sowing through 7 stages (1) germination, (2) emergence, (3) end of vegetative phase, (4) floral initiation, (5) flowering, (6) pod filling, (7) maturity. Accumulation of predetermined number of thermal units decides each growth stage. Germination is set to occur the day after sowing in all landraces. The accumulated thermal units produce the first leaf. Leaf area is produced as a function of leaf number. Daily plant growth is computed by converting daily intercepted Photosynthetically Active Radiation (PAR) into plant dry matter using a crop-specific radiation use efficiency parameter. Light interception is computed as a function of leaf area index (LAI), radiation use efficiency (єs), light extinction coefficient (K). The amount of new dry matter available for growth in each day is modified by the most limiting of soil moisture or air temperature. Above ground biomass demands the major part of the carbohydrates produced each day and at the end of the day carbohydrates not used for above ground parts are allocated to roots. Roots receive growth stage dependent minimum amount of carbohydrates.

Pod numbers per plant are computed during the flowering stage based on the landrace’s genetic potential, water stress, heat stress, cold stress and photoperiod level. Potential pod number is inversely proportional to the leaf number. Once the grain filling is started, the model computes the daily growth rate of pods based on user defined landrace inputs. Daily growth rate of pods is modified by temperature stress, water stress and photoperiod level. If the daily available photosynthates are insufficient to achieve the potential growth rate of pods, a fraction of carbohydrates can be remobilized from vegetative parts to reproductive sinks each day.

Table 2. Summary of key model parameters




Method of estimation


Radiation use efficiency

1.5-2.0 g/MJ

TCRU -2003


Transpiration equivalent

1.9-2.2 g/mm

TCRU -2003


Light extinction coefficient




Bulk density of soil

1.3 g/cm



Base temperature

9.9-12 0C

Massawe 2003


Optimum temperature

28-31 0C

Massawe 2003


Ceiling temperature

42-15 0C

Massawe 2003

*The values are lower limit and upper limit for landraces

Results and Discussion

Model simulation results closely matched with the corresponding measured values from the Glasshouse experiment, Nottingham and field sites in Botswana in terms of canopy development (Figure 1) and pod weight (Figure 2).

Under the Glasshouse condition, Swaziland landrace, UniswaRed showed significantly higher LAI and lower pod yield under high temperature irrespective of the moisture regime indicating heat stress. Similar results were obtained for Namibian landrace (S19-3) only for LAI. Pod yield was not significantly different between two temperature regimes under irrigated condition and it was higher with high temperature under drought. These results explain the model predictions for two landraces originated in two contrasting environmental conditions: Swaziland (cold), Namibia (warm).

The results from field experiment in Botswana showed the effect of photoperiod on growth and development of UnswaRed landrace. The crop was grown at semi arid condition in Botswana showed significantly lower canopy development and pod yield when compared with Glasshouse data in Nottingham. The results explain the growth and yield of UniswaRed under different photoperiod levels that captured in 3 sowing dates December 21, January 18 and February 1.

Figure 1. Measured and simulated LAI for the cumulative thermal time in TCRU glasshouse experiments at 2 different temperatures (3350 C, 2350C) for UniswaRed-irrigated (a1), UniswaRed-droughted (a2), S19-3-irrigated(b1), S19-3-droughted (b2) and 3 sowing dates for UniswaRed in field experimets in Botswana (c).

Figure 2. Measured and simulated pod weight for the cumulative thermal time in TCRU glasshouse experiments at 2 different temperatures (3350 C, 2350C) for UniswaRed-irrigated (a1), UniswaRed-droughted (a2), S19-3-irrigated(b1), S19-3-droughted (b2) and 3 sowing dates for UniswaRed in field experimets in Botswana (c).


This paper reports a crop simulation model for an indigenous crop bambara groundnut and develops a suitable model framework. The model BAMGRO gives good predictive capabilities for bambara groundnut phenology, LAI, biomass production, dry matter partition and pod yield. This develops landrace specific relations to calculate development and yield for a specific environment. The model also calculates each of the growth processes under major abiotic stress factors of moisture, heat and cold under different photoperiod regimes.


Bannayan M (2001). BAMnut: a crop simulation model for bambara groundnut. Agricultural Sciences & Technology 15, 101-110.

Boote KJ Minguez, MI and Sau, F (2002). Adapting the CROPGRO legume model to simulate growth of Faba bean. Agronomy Journal 94, 743-756.

Brink M (1997). Rates of progress towards flowering and podding in bambara groundnut (Vigna subterranea) as a function of temperature and photoperiod. Annals of Botany 80, 505-513.

Cornelissen RLEJ (2005). Modelling variation in the physiology of bambara groundnut (Vigna subterranea (L) Verdc.). School of Biosciences, PhD theis, pp. 200. Sutton Bonnington: University of Nottingham, United Kingdom.

King J, Gray A, Bradely RG, Binham I, Folkes J, Gregory P and Robinson D (2003). Modelling cereal root system for water and nitrogen capture: Towards an economic optimum. Annals of Botany, 91, 383-390.

Mathews RB and Stephens W (1998a). CUPPA TEA: A simulation model describing seasonal yield variation and potential production of tea. 1. Shoot development and extension. Experimental Agriculture 34, 345-367.

Massawe FJ, Azam-Ali SN and Roberts JA (2003). The impact of temperature on leaf appearance in bambara groundnut landraces. Crop Science 43,1375-1379.

Mawale SS, Azam-Ali SN and Massawe FJ (2007). Growth and development of bambara groundnut (Vigna subteranea) in response to soil moisture 1. dry matter and yield. European Journal of Agronomy 26, 345-353.

Robertson MJ, Carberry PS, Chauhan YS, Ranganathan R and O'Leary GJ (2001). Predicting growth and development of pigeonpea: a simualtion model. Field Crops Research 71, 195-210.

Previous PageTop Of PageNext Page