Knowledge of such information how J. curcas root system architecture varies across a range of soil types will facilitate an evaluation of its suitability for revegetation in soil conservation efforts, will be relevant for biofuel purposes, and may also help determine if both aims can be achieved simultaneously. The objective of this study was to quantitatively describe the shoot and root structural variation of J. curcas seedlings in three different soils that are characteristic of the Mexican tropics. 2. Materials and Methods2.1. Biological MaterialNative Mexican seeds of J. curcas were collected in Papantla, in southeastern Mexico (20.2558�� N, 97.2600�� W, 77 masl) during August 2010. Seeds were selected from the middle of their weight distribution for sowing; average �� standard deviation (SD) measures were mass: 758 �� 97mg, length: 8.

4 �� 1.0mm, width: 10.4 �� 0.50mm, and thickness: 9.0 �� 0.5mm. 2.2. Soil SelectionSoils were selected based in their textural characteristics and because they represented prominent soils of the eastern Mexican tropics. The sandy soil was an arenosol, the sandy-loam was a regosol, while the clay-loam was a phaeozem; typologies were based on previous research performed in the region [13]. Sandy-loam and clay-loam soils were obtained from the premises of the Colegio de Postgraduados in Veracruz (19.1954�� N, 96.3389�� W), while sandy soil was obtained from a dune near the city of Veracruz (19.2093�� N, 96.2597�� W). The upper 50cm of soil was collected and homogenized; one subsample (500g) was taken from each soil type for physical and chemical analyses.

Textural characterization was performed following Bouyoucos [14] and classified according to NRCS [15]; bulk density was estimated by the gravimetric method. Analysis of pH was conducted using an electronic potentiometer in a 1:1 slurry, organic matter content was determined by the Walkley-Black method, extractable phosphorus was determined following Olsen and Sommers [16], and exchangeable calcium and magnesium concentrations were determined using methods based on Diehl et al. [17], all adapted for Mexican soils [18]. 2.3. Experimental ConditionsThe AV-951 experiment was conducted outdoors in Veracruz, Mexico (19.1988�� N, 96.1522�� W, 2masl) and was carried out using a completely randomized design, with 15 replicates per soil type (clay-loam, sandy-loam, and sand; n = 45 plants). Seeds were sown in early September 2010 and were uprooted three months after germination (when they were in the juvenile life stage). The maximum, minimum, and average temperatures recorded at a local meteorological station (Skywatch Geos no. 11) during the period were 29.2, 19.4, and 23.7��C, respectively. The average relative humidity was 75.3%.