Field trials with a large group of cassava germplasm were conducted at the seasonally-dry and hot environments in southwest Colombia to investigate photosynthetic characteristics and production under drought conditions. Measurement of net photosynthetic rate (P N ), photosynthetic nitrogen use efficiency (PNUE), mesophyll conductance to CO 2 diffusion (g m ), and phosphoenolpyruvate carboxylase (PEPC) activity of upper canopy leaves were made in the field. All photosynthetic characteristics were significantly correlated with final dry root yield (Yield). Correlations among the photosynthetic traits were also significant. PEPC activity was highly significantly correlated with P N and PNUE, indicating the importance of the enzyme in cassava photosynthesis and productivity. Among a small selected group from the preliminary trial for yield performance, the second year Yield was highly significantly correlated with P N measured on the first year crop. Thus variations in the measured photosynthetic traits are genetically controlled and underpin variations in yield. One short-stemmed cultivar M Col 2215 was selected for high root dry matter content, high harvest index, and tolerance to drought. It was tested under the semi-arid conditions of the west coast of Ecuador; participating farmers evaluated cultivar performance. This cultivar was adopted by farmers and officially released in 1992 under the name Portoviejo 650.

Diurnal variations in net photosynthetic rate (P N ), transpiration rate (E), stomatal conductance (g s ), internal CO 2 concentration (C i ), and water use efficiency (WUE) were studied on individual leaves of coffee plants to determine the effect of climatic factors on photosynthetic capacity. P N and E showed bimodal behaviour with the maximum values of P N at mid-morning. At noon, under saturating photosynthetically active radiation (PAR) and high leaf temperature (T l ), P N declined. In the afternoon (14:00), P N slightly recovered in association with a decrease in T l and in leaf-to-air vapour pressure deficit (VPD). Reductions in E during the morning were associated with decreases in g s . Higher WUE in the morning was related to higher P N and lower E. The reverse occurred in the afternoon. Goudriaan's simulation model, adapted for coffee canopy photosynthesis, was tested at the level of whole plant (P pl ). Three methods were used: (a) Whole plant net photosynthesis (P pl ) under semi-controlled conditions in a chamber. (b) P pl estimation following Goudriaan's method (Gaussian integration) of instantaneous P N in single leaves at three canopy depths and at three different hours assuming a photosynthesis unimodal behaviour. (c) P pl using Goudriaan's method but at five different hours according to the bimodal behaviour reported above. Results of P pl estimates using Goudriaan's model adapted for coffee canopy confirm the observed P pl bimodal behaviour with high fitness degree of the measured whole plant photosynthesis. The high fitness found among observed and simulated data indicates that the modified model may be used as a subroutine for the general simulation model of coffee crop growth.Additional key words: Goudriaan's model; leaf gas exchange; net photosynthetic rate; whole plant photosynthesis.

No abstract

<p>El modelo de simulación de crecimiento y producción de soya (Glycine max. (L) Merr) Soygro V5-42, fué validado a nivel del trópico con datos experimentales de un ensayo en el cual se evaluaron dos genotipos Soyica P-33 e ICA-Ariarii-1, con diferente hábito de crecimiento, bajo dos densidades de plantas, en el Centro de Investigación Palmira de la Corporación Colombiana de Investigación Agropecuaria Corpoica, localizado a 3°32' de latitud norte y 76°17' de longitud oeste, en un Mollisol, clasificado como Isohipertérmico Aquico Hapludoll. Las salidas del modelo fueron sensibles a los ajustes en trece coeficientes genéticos, los cuales se calibraron sistemática e iterativamente. La calibración del modelo presentó una estrecha relación entre lo observado y lo simulado para las principales variables de respuesta. Su validación con base en datos de experimentos de campo anteriores, presentó una relación muy estrecha R&gt; = o.86, entre lo observado y lo simulado; lo cual indicó, que el modelo explica acertadamente la variación en las épocas de siembra y las densidades de plantas. Como resulta do de la simulación del efecto de las épocas de siembra con variaciones de clima, bajo diferentes ambientes (3 localidades), se encontró que el modelo se ajusta a las condiciones de siembra utilizadas por los agricultores en Colombia.</p><p> </p><p><strong>Evaluation of a Growth Simulation Model Applied to Soybean Genotypes (<em>Glycine </em><em>max</em><em>.</em>L. Merr) Under Tropical Conditions</strong></p><p>The crop growth simulation model of soybean ( <em>Glycin</em><em>e </em><em>max </em>(L) Merr) Soygro V5.42, was tested under tropical conditions, using experimenta l data from a field experiment in which two genotypes (Soyica P-33 and Soyica Ariari-1) with different growth h abit, were grown under two plant densities, at the Palmira Research Centre of the Corporación Colombiana de Investigación Agrop ecua ria, located at 3°32' north latitude and 76°17' west longitude, on a Mollisol soil classified as Isohyperterthermic Aquic Hapludoll. Thirtheen genetic coefficients were calibrated and showed sensitivity to the model. The validation of the model was made using previous experimental data from the same genotypes. The narrow line observed vs predicted seed yield was close to the 1:1 slope, indicating that, the simulating was not skewed. Thus, the model explained well the variation due to planting dates and plant spatial distribution in the field. The simulation of planting dates effects in relation to climatic variation at three different soybean production regions, is in agreement with the traditional planting dates used by the soybean growers of Colombia.</p>