While all eyes in crop production research are typically focused on assessing, manipulating, and propagating variability in yield among crop cultivars, all too often very little is known about cultivar differences in growth habit, phenology, and ontogeny that are causal agents of that yield variability. This is certainly the case among today's cultivated peanut (Arachis hypogaea L.) genotypes where almost no information is available regarding cultivar differences in aboveground growth habit, leaf size, flowering potential, tissue nutrient contents, harvest index, nor on the differences in development of these traits through the season. On the surface it may appear that this information has relatively little value to either scientist or producer. However, in the ever-narrowing gap between economic viability and bankrupted ruin in the U.S. farm environment, these traits can have very significant implications on final peanut production and can increase the chance of economic success. Cultivar choice can be a “make or break” decision for a producer and having information about a cultivar's growth habit can be a critical component of this decision. In today's southeastern U.S. peanut production environment, there are roughly 8–10 cultivars that are available for a producer to choose from. Up front, this means that information regarding variability among cultivars in growth habit and reproduction could help a grower choose cultivars that will yield optimally under particular field conditions. Beyond these producer benefits, science can also reap the rewards of further information about cultivar variation because it represents essential input into both breeder programs and crop model development.

The existence of peanut cultivar variability in growth and development is questionable because of the increasingly narrow gene pool foundation of current peanut germplasm as a result of industry and grower preferences. The cultivars used in the southeastern U.S. peanut production are almost solely runner market types because of the preference of the processing industry towards their shelling, size, and flavor characteristics. In addition, there is an absolute necessity for the sole use of Tomato spotted wilt virus resistant runner peanut genotypes, often derived from similar parental germplasm, because of the devastating economic effects of this plant virus. Both of these conditions have significantly narrowed the genetic background and variability among peanut cultivars making them genetically quite similar in many ways (Kottapalli et al., 2007). Therefore, any variation among current cultivars is essential to document and quantify.

Of course environmental growth conditions can have an impressive influence on the expression of inherent genetic differences in peanut cultivars. One such influence is the application of supplemental irrigation during the growing season. Nearly 50% of U.S. peanut production is under irrigation and this percentage is expected to expand as the global climate becomes warmer and droughts become more frequent and severe in many of these production regions (Lamb et al., 2004). Surprisingly, within the southeast U.S., irrigation is vital to production of economically sustainable yields despite the widely presumed image of the area as receiving adequate rainfall. Although irrigation can increase yields by as much as 50% (Lamb et al., 1997; Lamb et al., 2004), skyrocketing fuel prices ($30.00 U.S. per hectare – Nathan Smith pers. comm.) necessitate that irrigation application be as efficient as possible or it will quickly lose its cost effectiveness.

Currently there are three established and utilized irrigation scheduling methods for peanut: 1) Irrigator Pro (IP), 2) the EASY Pan (Evaporation-based Accumulator for Sprinkler-enhanced Yield - EZ), and 3) the University of Georgia's Extension check book method (UGA-EXT). Irrigator Pro is based on monitoring soil temperatures and incrementing water application with crop growth stage (Lamb et al., 1993; Davidson et al., 1998). The EASY Pan method uses a galvanized washtub and float with different sized mesh screening over the top of the tub to simulate evaporation losses affected by different levels of crop cover (Thomas et al., 2004). The University of Georgia's Extension checkbook method is an incrementally increasing model linked to crop stage and has now been updated by J.P. Beasley (unpublished data). To date, no studies have actually compared the efficacy of these methods on peanut production. Variation among peanut cultivars as to how they respond to irrigation scheduling methods is most likely due to: 1) differences in growth habits; 2) variation in the time period required for fruit maturation; and 3) genetic differences in water use strategies and water-use efficiency (Rowland and Lamb, 2005). Irrigation scheduling method could have critical effects on the plasticity of these traits that translate into long-lasting effects on final crop yield.

The current study was aimed at documenting and quantifying any variability among several currently grown peanut cultivars in traits that could have significant impact on physiological functioning of the crop and thus influence yield. Secondly, the differential effects on peanut development of the three most commonly used irrigation scheduling methods (IP, EZ, and UGA-EXT) were tested. The following specific objectives were addressed: 1) is there any variation in morphology, ontogeny, and reproduction among currently grown peanut genotypes? and 2) what is the effect of irrigation scheduling method on these traits, especially on carbon isotope discrimination as a representation of seasonal water-use efficiency?

There is true value in studies that examine growth and developmental differences among crop cultivars because they describe genetic variability that can either be utilized in breeding programs or crop models (Bell et al., 1991b; Bell et al., 1993a; Craufurd et al., 2002; Baterng et al., 2003; Kiniry et al., 2005). For breeding programs, a description of the differences among cultivars is essential for exploiting available genetic variability during the development of new germplasm. For crop modelers, knowledge of how cultivars can differ in their growth and reproductive responses can be essential to the success of the model's predictive ability. For example, the crop model PNUTGRO requires over 40 cultivar-specific inputs for successful simulation of growth and yield in peanut (Hammer et al., 1995). Previous work has shown that there is variability in growth and reproduction in peanut but comparisons have typically been made among widely different peanut species and botanical types (Bell et al., 1991a,b; Bell et al., 1993a,b; Bell et al., 1994; Craufurd et al., 2002; Banterng et al., 2003). But the existence of variability among current peanut cultivars whose genetic pedigrees have become increasingly similar due to the type of traits that have been favored for several years is questionable at best. However, this study has documented and quantified several important growth and reproductive characteristics that differ among currently cultivated peanut genotypes. It also appears that there are differences among cultivars in resource allocation strategies for canopy development that can translate into effects on pod production.

Despite expectations of a kind of genetic dilution to phenotype variability among cultivars, genotype differences were quite strong for growth traits. Previously documented cultivar differences in leaf weight and harvest index do exist, but again these were in widely diverse botanical types (Bell et al., 1993a; Craufurd et al., 2002; Banterng et al., 2003), making the current study's documentation of growth variability among closely related cultivars unique. Although year had a significant effect on these traits indicating an effect of climate on peanut physiology, there were largely no interactions between year and genotype indicating that the changes caused by climate were relatively low in magnitude and did not alter the overall genetic pattern of development for each cultivar. A lack of an effect of irrigation scheduling method or interaction between irrigation scheduling method and cultivar for growth traits also illustrated the fairly strong genetic control over aboveground biomass development among the cultivars in this study.

Overall, the pattern and magnitude of LAI measurements across the season in this study were typical for other peanut cultivars (Banterng et al., 2003; Kiniry et al., 2005), but some studies have documented a more parabolic pattern for LAI development (Kar and Kumar, 2007) as well. Patterns of growth in the current study indicated that late-maturing cultivars had higher late-season growth than mid-maturing cultivars. Trade-offs in allocation patterns among LAI, stem, and leaf growth indicated cultivars had different strategies for canopy development. For example, although the late-maturing cultivar Georgia-01R had high partitioning to leaves and stems, its short internode lengths maintained a mid-level LAI. Two mid-maturing cultivars alternated between large internode lengths and thus high LAI (Carver) and short internode lengths and low LAI (Georgia-03L). Variability in partitioning strategies was also illustrated by similar harvest indices that were arrived at by coupling low below- and aboveground biomass in Tifrunner, while Georgia-01R coupled high below- and aboveground biomass. Biomass partitioning especially between leaves and stems or between vegetative and reproductive traits can vary among cultivars in other legumes such as cowpea (Vigna unguiculata (L.) Walp.) and chickpea (Cicer arietinum L.), and some diversely related peanut cultivars (Bell et al., 1993b; Bell et al., 1994; Krishnamurthy et al., 1999; Banterng et al., 2003; San José et al., 2004).

Unlike growth traits, the separation among cultivars in reproductive traits was not related to maturity class. On average, mid- and late-maturing cultivars had similar flower, peg, and pod production. With elevated canopy growth but equal reproductive rates, late-maturing cultivars were expected to have lower harvest indices than mid-maturing genotypes. This was certainly the case, especially for the later maturing cultivars Georgia-01R, Georgia-02C, and Tifrunner which had the lowest harvest indices among all eight cultivars. This pattern of lowered harvest indices in late-maturing cultivars indicates a somewhat lower efficiency of yield return for the amount of energy invested in aboveground canopy structures. These results concur with differences in harvest index that have previously been found between peanut cultivars of different maturity classes with late maturing cultivars having low relative harvest indices (Bell et al., 1991b; Bennett et al., 1993). This is true across all grain/seed producing food crops, legumes, or cereals.

Aside from the contribution of genetics, harvest index can be significantly affected by various environmental factors that affect peanut growth and reproduction including: temperature (Vara Prasad et al., 2000); heat stress (Vara Prasad et al., 1999); irradiance and photoperiod (Bagnall and King, 1991a,b); ozone stress (Booker et al., 2007; Burkey et al. 2007); and even planting pattern (Lanier et al., 2004) Of course one of the greatest influences on harvest index can be irrigation amount and timing (Reddy et al., 2003). That is why perhaps the most surprising result from this study was the lack of impact of irrigation scheduling method on peanut growth and reproduction. Even though there was over a 100 mm difference in the total water applied in both years between the top and bottom irrigation scheduling methods, the crop performed well under all systems and was not significantly affected. This is likely due to the fact that all three irrigation scheduling methods in both years provided more than enough water to meet maximal cultivar yield potential. Based on sap flow water-use measurements in southeastern peanut, seasonal water-use can be approximately 710–735 mm (unpublished data). For both 2004 and 2005, water receipt (from both irrigation and precipitation combined) topped these levels indicating that irrigation application could have been reduced while still meeting crop demand.

Although the irrigation environment did not affect growth and reproduction to an appreciable degree, yearly variation did. Looking at the climatic data, the most likely factor was the timing and amount of rainfall, with 2005 having the highest rainfall total during the period of flowering and fruit maturation. However, the crop did not utilize this higher water availability in reproduction (as evidenced by the lack of significance for year in peg number and pod number and weight). The effect of this difference in water application between years was in the growth characteristics; namely stem and leaf weight (Table 2) which were actually lower across cultivars in 2005 (data not shown). Additional water “availability” coupled with lower leaf and stem production might indicate some degree of water over-saturation in the soil. Tissue nitrogen was also lower in 2005 than in 2004 (data not shown) which indicate deleterious effects on nitrogen fixation under hypoxic soil conditions. Either way, water availability appeared to be more than adequate for the crop in both years.

Harvest index measured in these eight cultivars showed relatively low values across years in comparison to 14 studies examining harvest index in peanut reviewed by Kiniry et al. (2005). The low harvest indices in the current study are even more surprising given all treatments were irrigated because harvest index is predicted to be as high as 0.58 under efficient irrigation (Kiniry et al., 2005). This result also provides evidence that all three irrigation scheduling methods used in this test were likely applying too much water to the crop, thereby increasing canopy growth without a concomitant increase in yield and thus leading to low harvest indices. Based on these results, it appears that irrigation scheduling methods for southeastern U.S. peanut have a lot of room to lower total water application during the season and increase efficiency of crop water use.

Variation among cultivars in δ¹³C showed there is room to develop peanut germplasm with increased seasonal water-use efficiency as well. The larger canopies of late-maturing cultivars necessarily spell a larger area for transpirational water loss. For this reason, it might be expected that water-use would be higher in late- than mid-maturing cultivars and thus lead to decreased water-use efficiency. However, patterns of δ¹³C did not reflect lower seasonal water-use efficiency for late-maturing cultivars. In fact two of the cultivars exhibiting the highest seasonal water-use efficiencies (Georgia-02C and Tifrunner) were late-maturing cultivars. Seasonal water-use efficiency appears to be under tight genetic control in U.S. peanut cultivars because irrigation scheduling method had no effect on δ¹³C composition. Isotopes are often affected by irrigation regime, but the current results agree with previous results that found very little effect of irrigation on the isotope composition of southeastern cultivars (Rowland and Lamb, 2005). The only slight indication of irrigation scheduling changing the pattern of peanut water-use was the differing δ¹³C levels in the EZ method between years. In 2004, the EZ treatment showed much higher water-use efficiency in the crop while in 2005, this trend was opposite. Therefore, irrigation scheduling appears to have the potential to affect crop water-use efficiency but is highly dependent on prevailing climatic conditions.

The existence of varying partitioning strategies among cultivars in this study necessitates the assumption that reproduction might be predicted from patterns in certain growth traits throughout the season. Although an early study indicated that peanut yield was determined more by genetic pedigree than growth habit (Norden and Lipscomb, 1974) more recent studies show a strong link between growth and reproduction for other legumes (Krishnamurthy et al., 1999) and other peanut culitvars (Bennett et al., 1993; Banterng et al., 2003). In the current study, traits relating to canopy development (and thus assimilation capacity) as well as mid-season reproductive potential appeared to be causally linked to final reproductive output in the late season. LAI and leaf weight had some correlative relationships with the yield-determining traits of pod number and pod weight. As expected, pod number and weight were highly correlated with one another; a result that has recently been supported in different peanut cultivars (Haro et al., 2007). This relationship between pod number and weight indicates there is no real trade-off between pod number and pod size likely due to the indeterminate nature of peanut. Aside from the relationships between pod number and weight, the strongest predictors of reproductive output were late-season traits including leaf weight and LAI. However, the correlation pattern was different for these two canopy traits; there was a negative relationship between LAI and pod weight, but for leaf weight, there was a positive relationship with pod number. This disparity may reflect an allocation trade-off where the more energy put into an overall larger canopy (LAI) had no real pay-off in production, while increased leaf production alone does have benefits because assimilation potential would be expected to increase. Mid-season traits that were important predictors of pod production were peg and pod numbers. These relationships really illustrate that the reproductive potential of the crop is determined at that mid-season point. Therefore, although the overall growth habit of peanut is indeterminate, any energy allocated to reproductive structures by the crop later in the season goes to waste because the plant cannot adequately mature these fruits.

This study documented no large differences among cultivars in nutrient accumulation or allocation of a magnitude that would have biological relevance. This result has importance to the application of the nutrient predictor model that uses the nutrient levels of leaf tissue in the late-season to predict optimum peanut harvest maturity (Rowland et al., 2008). Because genotype appears to have little effect on nutrient levels in leaf tissue across the season, application of this maturity model across different peanut genotypes is likely to give reasonable predictions of peanut maturity. The overall parabolic pattern of tissue nutrient levels confirms what Rowland et al. (2008) found for the cultivar Georgia Green. A decline in vegetative nitrogen late in the season found in these eight peanut cultivars shows a possible remobilization of nitrogen from vegetative parts to developing fruits which consequently drops nutrient levels in the late season as the crop matures (Bell et al., 1994).

While yield is the basis on which genotypes are compared and evaluated in peanut breeding trials, much less information is known about differences in growth and developmental traits because many of these characteristics are largely ignored (Banterng et al., 2003). By selecting for yield alone and not examining other traits that are causally related to yield may be a less successful strategy in breeding programs because these traits can be important predictors of yield (Wallace et al., 1993). Information on the physiological basis to yield variation among cultivars can be used to explain the mechanisms behind such variation and even determine alternate management strategies in the field that can optimize a particular genotype's performance (Baterng et al., 2003). Therefore, knowing the causes behind yield variability can ultimately aid in maximizing it.