There is abundant evidence that sucrose is unloaded from the phloem in the pedicel and hydrolysed to glucose and fructose by acid invertases before being taken up by rapidly filling maize kernels (Shannon, 1968, 1972; Shannon and Dougherty, 1972; Felker and Shannon, 1980; ap Rees, 1984; Doehlert and Felker, 1987; Griffith et al., 1987; Porter et al., 1987; Miller and Chourey, 1992; Thomas et al., 1993; Xu et al., 1995; Cheng et al., 1996). This also seems to be true when ovaries alone are present, because Zinselmeier et al. (1995b, 1999) found high activities of acid invertases prior to fertilization. The invertases began to be active when the ovaries became susceptible to abortion.
The hexose products of the invertase reaction were either converted to starch and stored in the ovary walls or were converted to other constituents necessary for ovary development (Zinselmeier et al., 1999). When low w occurred, the starch disappeared from individual ovaries destined to abort (cf. Fig. 1D, F). Glucose followed suit and decreased as the starch disappeared (McLaughlin and Boyer, 2004). It appears that the starch supplied glucose to the developing ovaries when the sucrose stream diminished at low w, buffering against low glucose. But the starch is a small pool and quickly depleted. It would be sufficient to supply glucose for a night when photosynthetic products may be in short supply but could not make up for days of inhibited photosynthesis that occur at low w (McLaughlin and Boyer, 2004).
Schussler and Westgate (1991) reported that ovaries excised from plants having low w absorbed sucrose less rapidly than controls. This suggests that the uptake process may have been affected at low w. Because invertase hydrolyses sucrose to hexoses, its activity may play a part in uptake. The decreased invertase activity might then account for the slower sucrose uptake these workers observed at low w.
Schussler and Westgate (1994, 1995) further investigated whether stored reserves such as sugars and starch affected the tendency of ovaries to abort. Plants with high or low reserves were grown and subjected to low w. Regardless of reserve status, abortion occurred similarly in each treatment and the investigators concluded that the flux of photosynthate was more important than the reserve status of the parent plants for triggering abortion. It would have been interesting to determine the status of the starch in individual ovaries about to undergo abortion. It seems possible that ovary starch could be broken down despite high reserves in the rest of the parental tissues.
Recent reviews have proposed intriguing interactions between ABA and sugars that might regulate developmental processes in well-watered plants (Finkelstein and Gibson, 2002; León and Sheen, 2003). Setter et al. (2001) reported increasing sugar and ABA concentrations in maize florets during a water deficit. Reports that these compounds increase in concentration in florets of water deficient maize (Schussler and Westgate, 1995; Zinselmeier et al., 1995b, 1999; Setter et al., 2001) raise the possibility that such changes (or their interactions) might signal abortion processes to commence. The difficulty in assigning a ‘signal’ value to these observations is that it is unclear how much of the increases in sugar and ABA concentrations were attributable to lower water contents in the florets. Zinselmeier (1991) reported that a water deficit prior to pollination decreased ovary water content by about 3–4% (87% to 83% moisture) and was severe enough to cause abortion of nearly all pollinated florets. The observed change in concentrations therefore depends on whether water contents or photosynthate flux was more affected, which would be likely to vary between experiments.
Thus, it is difficult to establish a link between changes in concentrations of, say, ABA and sugars, when all other constituents have increased as well because of water loss. In support of this concept, when Wang et al. (2002) measured catabolic activity for ABA at low w, there was a large increase on a fresh-weight basis but less when expressed on a per-kernel basis (see Fig. 5 in Wang et al., 2002). Similarly, carbohydrate concentrations increased on a fresh-weight basis, but the accumulation of carbohydrate had actually decreased when expressed on a floret basis (see Figs 6 and 7 in Setter et al., 2001).
In an experiment similar to Setter et al. (2001), but conducted under field conditions, Andersen et al. (2002) reported some ovaries aborted as low w developed, but aborting and developing ovaries displayed no difference in ABA concentration expressed on a dry-weight basis. Like ovary water content, the dry-matter content of the ovaries is highly dependent on the photosynthate flux delivered to them. Sugars and starch can comprise as much as 70% of the ovary dry weight (Zinselmeier et al., 1999). As the delivery of sugars diminishes, the dry weight decreases relative to that of well-watered plants. Consequently, changes in ABA concentration depend on whether ABA or dry weight is more affected by the water deficit.
Andersen et al. (2002) found that sucrose, hexose, and starch concentrations were depleted in some low w treatments, but not in others. These investigators did not determine these concentrations in individual ovaries and instead sampled ears having a mixture of aborting and normal ovaries. Thus, it was inevitable that variable numbers of normal ovaries were included in the samples. Andersen et al. (2002) could detect no relationship between ovary abortion and ovary sugar and starch concentrations, but agreed with Schussler and Westgate (1994, 1995) and Zinselmeier et al. (1999) that abortion was controlled by the photosynthate flux to the ovaries.
When Zinselmeier et al. (1999) fed sucrose to the stems of maize, the photosynthate flux increased at low w and the starch pool was maintained in the ovaries (Fig. 1E). The presence of starch thus indicated whether ovary demand for sucrose (and glucose) was being met. Because the sucrose feeding prevented much of the abortion (Fig. 1B), Zinselmeier et al. (1999) concluded that aborting ovaries were starving for sugars at low w, and re-establishing the sugar stream prevented the abortion.
Zinselmeier et al. (1999) found that ovary starch, while high in the plants fed sucrose at low w, remained less than in the controls. Invertase activity was also not fully maintained by sucrose feeding. The downstream intermediates leading to starch biosynthesis were depleted at low w and only partially replenished by the feeding. Glucose and fructose are the products of invertase, and glucose concentrations were only partially maintained with sucrose feeding (McLaughlin and Boyer, 2004). The sucrose feeding rescued 40–80% of the otherwise aborting ovaries, and Zinselmeier et al. (1999) attributed the lack of 100% recovery to the remaining low invertase activity and depletion of downstream pools of metabolites.