LCMsymm = 0 if all main laterals originate at the same point and 1 if roots originate from opposite extremes of the stump. Note that this definition of LCMsymm differs from that of Reubens et al. [3], insofar as they considered a departure from the fixed value of 2.5cm for all seedlings in their study. As this value depends on the soil type and how deeply the seed was sown, we only evaluated the similarity of the length to root base collar (LC) from each lateral root.The final similarity measurement we considered was the angle from which the taproot deviated from a vertical line (Tapsymm):Tapsymm=[(?90��)?TapIncAngle]90��,????????????????(7)where TapIncAngle is the inclination angle between the taproot and the vertical at the level of the root stump (Figure 4(e)). Tapsymm = 0 if the taproot is vertically oriented and 1 if the taproot is horizontally oriented.We computed a composite metric for the degree to which J. curcas plants adhered to the idealized model plant (SI):SI=(��symm+??Dsymm+??��symm+??LCMsymm+Tapsymm)5??.??????(8)SI = 0 for root systems perfectly matching the model and 1 for complete lack of adherence.We used an index of phenotypic plasticity (PI) [23] to quantify the magnitude of the morphological response to varying soil types. For each variable, PI uses the mean response for individuals grown in each treatment to evaluate the greatest change displayed by the species among treatments:PI=(Maximum?value?Minimum?value)Maximum?value.(9)PI ranges from 0 to 1, with 1 representing the greatest possible plasticity.2.8. Statistical AnalysisDifferences in parameter means among soil types were statistically compared using one-way analysis of variance (ANOVA) in SigmaPlot 10.0. Tests of residual normality and equal variance were conducted. Post hoc comparisons were made for normally distributed parameters with a Tukey test, while non-normally distributed parameters were analyzed with Dunn’s Method, all with a 95% confidence level.3. Results3.1. Soil AnalysisAll substrates were found to be slightly alkaline. However, the sandy soil had very low organic matter content, being 2�C4% of the amount in the other soils (Table 1). The sand also had 10�C26% of the P, 23�C44% of the Ca, and 29�C53% of the Mg found in other soils (Table 1). Table 1Soil characteristics for the three soil types in which J. curcas seedlings were grown for three months.3.2. Above- and Belowground Response to Soil TypesPlants grown in the sandy-loam and clay-loam soils had, on average, approximately twice the height and three times wider collar diameter than plants grown in sandy soil. Stem volumes, numbers of leaves, and leaf areas were more than five times greater for plants grown in these soils than for plants grown in sandy soil (Table 2). All plants grown in sandy soil survived, but 62% were completely defoliated by the conclusion of the experiment; none of those grown in sandy-loam or clay-loam soils lost all leaves.