Fruit Development in Trillium

Abstract

Leaves are the main source of carbon for fruit maturation in most species. However, in plants seeing contrasting light conditions such as some spring plants, carbon fixed during the spring could be used to support fruit development in the summer, when photosynthetic rates are low. We monitored carbohydrate content in the rhizome (a perennating organ) and the aboveground stem of trillium (Trillium erectum) over the entire growing season (May–November). At the beginning of the fruiting stage, stems carrying a developing fruit were harvested, their leaves were removed, and the leafless stems were maintained in aqueous solution under controlled conditions up to full fruit maturation. These experiments showed that stem carbohydrate content was sufficient to support fruit development in the absence of leaves and rhizome. This is the first reported case, to our knowledge, of complete fruit development sustained only by a temporary carbohydrate reservoir. This carbohydrate accumulation in the stem during the spring enables the plant to make better use of the high irradiances occurring at that time. Many other species might establish short-term carbohydrate reservoirs in response to seasonal changes in growing conditions.

Carbon allocation in plants changes over the course of the season. In most herbaceous perennials there are large changes in underground biomass at the end of the growing season, reflecting the accumulation of carbohydrates for the next year's growth (Bradbury and Hofstra, 1977; Chapin et al., 1986; Cyr et al., 1990). However, spring ephemerals that senesce with the closing tree leaf canopy show a rapid accumulation of carbohydrates early in the season (Risser and Cottam, 1968), indicating that carbohydrate reserves can be quickly replenished and that the timing of carbohydrate accumulation might be related to growing conditions. Spring plants that senesce late in the summer experience two types of contrasting light conditions. Photosynthetic rates are high at the beginning of the growing season under the high light conditions before the tree canopy closes and much lower over the rest of the growing season, which is over by July or August (Sparling, 1967; Taylor and Pearcy, 1976). Under such conditions, plants might accumulate carbohydrates much earlier than plants growing under constant low (understory) or high (field) light conditions. The first goal of this study was to determine the pattern of carbohydrate accumulation and use in trillium (Trillium erectum L.), a spring herb with a long growing season.

Despite the fact that T. erectum produces only one fruit per plant, with a biomass of 2 to 8% of total plant biomass, fruit abortion is frequent (L. Lapointe, personal observation). Aerial parts of the plant senesce quickly when the plant aborts its fruit, most frequently in late June, whereas nonaborting plants keep their leaves up to the time of complete fruit maturation, suggesting that leaves are required to support fruit development. We completely defoliated plants at different times during fruit development to determine if fruit abortion could be induced later in the season, and to what extent carbohydrate reserves would be used to support the fruit during its development in the absence of leaves (L. Lapointe and A. Deslauriers, unpublished data). Contrary to our original expectations, complete defoliation during fruit development did not induce fruit abortion. Furthermore, carbohydrate analysis showed that there was no loss of reserves from the rhizome in defoliated, fruiting plants. These results suggested that there may be a temporary carbohydrate reservoir in the stem that can support fruit development and maturation. Stems have been shown to act as temporary reservoirs of carbon in some species, but usually contribute less than 50% to fruit development (Wardlaw, 1990; Schnyder, 1993).

Along with carbohydrate content of the rhizome, that of the stem was also measured during the entire growing season in fruiting plants of T. erectum. Previous work showed that at the end of the growing season, T. erectum rhizomes contained high levels of starch. Considering the large size of the rhizome and the small biomass of the fruit, it was possible that fruit maturation in defoliated plants would not reduce the level of rhizome carbohydrates to any great extent (Primack and Hall, 1990). The aim of this study was to determine whether the stem was capable of supporting fruit development by itself. Stems carrying a developing fruit were harvested, defoliated, and maintained under controlled conditions in liquid medium. If complete fruit maturation occurred this would show that the stem of T. erectum had accumulated sufficient carbohydrates to support the development of fruit in the absence of rhizome and leaves. Whereas some species use stems as temporary carbon reservoirs (Pate et al., 1983; Yamagata et al., 1987; Wardlaw, 1990), only Jerusalem artichoke (Helianthus tuberosus) appears to rely extensively on this temporary reserve for tuber formation at the end of the growing season (Incoll and Neales, 1970). Cereal plants rely on stem carbohydrates for fruit development under severe stress, which leads to reduced yields (Gallagher et al., 1976; Schnyder, 1993). If the stem can support normal fruit development in T. erectum, this would be the first reported case, to our knowledge, of a complete dependence on temporary carbohydrate reserves for fruit development.

MATERIALS AND METHODS

Plant Harvests

Every week from May to August, and every 2 weeks from September to November, 1994, eight reproductive plants of trillium (Trillium erectum L.) were harvested from a woodland area located near Québec city, Canada. The plants were washed, separated into rhizomes, roots, stems, leaves, and reproductive structures, dried (heat-killed at 100°C, dried at 70°C), and weighed. The rhizomes and stems were then ground to a fine powder. Plants harvested during the summer tended to be larger than plants harvested in the spring, so for size comparisons we also collected a series of fruits from smaller plants during the summer. These fruits were also dried and weighed.

Carbohydrate Analyses

Carbohydrates were analyzed as described previously (Lapointe and Molard, 1997), with the following modifications. Dry-ground tissues were reground with a polytron homogenizer (model PT 3100, Kinematica AG, Littau, Switzerland) in a mixture of methanol, chloroform, and water (12:5:3, v/v). The mixture was then left to macerate for 2 h at 0°C before being centrifuged to separate soluble sugars from nonsoluble residues. The supernatant was analyzed before and after invertase digestion to estimate reducing sugar and Suc content. The nonsoluble residues were first heated at 100°C for 90 min to gelatinize starch, then incubated at 55°C for 1 h in presence of amyloglucosidase. All reducing sugars were measured colorimetrically at 415 nm after reaction with p-hydroxybenzoic acid hydrazide.

Cut-Stem Experiment

Sixty T. erectum flowering plants were harvested in mid-June, 1996. By this time, plants that would have aborted their fruit had already started to senesce and only nonsenescing plants were chosen. Each stem was cut at the base with a razor blade and immediately placed in water. Back in the laboratory, the stems were measured, re-cut under water, and all three leaves were removed. Each stem was placed in a test tube containing 10 mL of one of the following: distilled water, nutrients, HQS, or nutrients and HQS. The nutrient solution was made up of 5 mm Ca(NO₃)₂, 5 mm KNO₃, 2 mm KH₂PO₄, and 1 mm MgSO₄. The concentration of the HQS solution was 200 mg L⁻¹. Stems were maintained in growth chambers (G30, Conviron, Winnipeg, Canada) that simulated understory conditions in July, with an irradiance of 25 μmol m⁻² s⁻¹ and a 14-h light period at 25/15°C day/night temperatures. Stems were cut every week and solutions were replaced twice a week.

Fruits were harvested when they reached maturity and easily detached from the pedicel. Mature seeds (large seeds bearing an eliaosome), aborted seeds (any seed larger than the unfertilized ovule but smaller than mature seeds and without an eliaosome), and unfertilized ovules were counted for each harvested fruit. These three categories of seeds were used to estimate the percent seed set and the percentage of ovules fertilized.

1

2

Total dry fruit biomass, which included carpels, unfertilized ovules, and aborted seeds, and mature seed dry biomass were measured.

We labeled 34 T. erectum flowering plants in the same woodland area in early spring. When the fruits were mature we recorded stem height and harvested the fruit. Fruits were dissected and weighed as described above. The plant height and fruit characteristics of cut-stem plants (four treatment groups) and of understory plants were compared using one-way analysis of variance.

RESULTS

Carbon Allocation during the Growing Season

Starch was the main carbohydrate in T. erectum rhizomes year-round (Fig. 1). Starch concentrations were low at the beginning of the season, but after 3 weeks growth rapidly increased and reached maximum levels within 5 weeks. The starch content of the rhizome then remained constant throughout the summer until senescence of the aboveground parts (late August for the plants sampled). Starch content decreased drastically throughout autumn and reached early spring levels by mid-November. The levels of Suc and reducing sugars did not show seasonal variations. Reducing sugar content was slightly but consistently higher than Suc content.

Figure 1.

Starch (▾), reducing sugar (•), and Suc (○) content (mg g⁻¹ dry mass) of T. erectum rhizome during the growth season and until late November. Data are represented as mean ± se from eight individuals.  

The pattern of stem carbohydrate content was different from that of the rhizome. In the stems, reducing sugars were the most abundant carbohydrates throughout the season (Fig. 2). For the first 3 weeks of growth, reducing sugar content steadily increased, and then decreased slowly over the summer until there was a rapid decrease when fruit reached maturity. Starch and Suc content were low in the stems. Starch content was highest from late May to late June. Suc content gradually decreased throughout the spring then increased slowly over the summer.

Figure 2.

Starch (▾), reducing sugar (•), and Suc (○) content (mg g⁻¹ dry mass) of T. erectum stem during the growth season. Data are represented as mean ± se from eight individuals.

To compare changes in fruit biomass during development and changes in stem and rhizome carbohydrates, the total carbohydrates (starch, Suc, and reducing sugars) in the whole rhizome and in the whole stem were estimated (Fig. 3). Rhizomes are large compared with stems and contain more carbohydrates. Week-to-week variations in rhizome total carbohydrates reflected large variations in plant size. Variations in fruit dry mass or total stem carbohydrates were much less than for total rhizome carbohydrates. Until the beginning of July, carbohydrates in the stem were equivalent to a higher biomass than the total mass of the developing fruit. After this period, fruit dry mass was higher than stem carbohydrate biomass, but both showed similar weekly variations until fruit maturation.

Figure 3.

Total carbohydrates contained in the rhizome (A) and stem (B) of T. erectum during the growth season. Flower (from May to mid-June) and fruit dry mass (○) are also presented in B for direct comparison with total stem carbohydrate content (•). Data are represented as mean ± se from eight individuals.

The total amount of carbohydrates in the stem at the end of spring was greater than that required for fruit development during the summer (Table I). Total leaf area, which is a good estimate of plant size in T. erectum (r = 0.74; P < 0.001), was similar for plants harvested in the spring for the carbon-allocation study (197.5 ± 10.1 cm²) and for plants harvested for fruit in late summer (206.1 ± 13.8 cm²). Since the bud for the following year's flower has been formed by the end of the season, the mean bud biomass was subtracted from fruit biomass to estimate the cost of fruit development for the current season. Fruit biomass amounted to about 100 mg or 65% of the stem total carbohydrates. These calculations do not take into account growth and maintenance or fruit respiration during development.

Table I.

Total carbohydrates accumulated in the stem during the spring compared with flower and fruit biomass in T. erectum

Total Carbohydratesa	Flowerb	Fruitc
mg
155  ± 7d	61  ± 3	161  ± 20

^aEstimated from plants harvested from May 16 to June 27, when sugar levels were stable in the stem (Fig. 3). 

^bEstimated from plants harvested from May 2 to June 14, when flower biomass was stable and fruit development was not initiated yet (Fig. 3). 

^cEstimated from 27 fruits harvested in the understory the same year (1994). 

^dData are presented as mean ± se for 56 individuals (stem sugar and flower biomass), and 27 understory plants (fruit biomass). 

Cut-Stem Experiment

None of the cut stems in aqueous solutions, but 21% of the labeled plants in the understory, aborted their fruit (data not shown). The plants chosen for the cut-stem experiment tended to be larger (higher stem height), but on average produced somewhat smaller fruits compared with plants in the understory; however, these differences were not statistically significant (Table II; P = 0.069). The proportion of fertilized ovules, the proportion of fertilized ovules that matured into seeds, and the total number of seeds per fruit were not significantly different for any of the treatments or growing conditions. Mean seed mass was lower in treatments in which HQS was included in the stem maintenance solution. Otherwise, there were no significant differences in mean seed mass of fruit developed on cut stems compared with mean seed mass of fruit developed on control plants. Carpel dry mass, which included all parts of the fruit except mature seeds, showed that the slightly lower fruit biomasses from cut stems were mainly due to lower mean seed mass; carpel dry masses were similar across all groups (P = 0.140). Fruits developed on cut stems were more fleshy, as was reflected by their lower ratio of dry to fresh mass. This parameter was also influenced by the presence of nutrients in the solution that tended to reduce the amount of water in the fruit.

Table II.

Characteristics of the fruits harvested from cut stems maintained in different aqueous solutions and from control plants growing in the understory

Treatment	Stem Height	Seed No.	Ovules Fertilized	Seed Set	Dry Mass			Fruit Dry-to-Fresh Mass Ratio
					Total fruit	Mean seed	Carpel
	cm		%		mg
Distilled water	27.8  ± 0.7	31.7  ± 5.2	74.2  ± 4.5	52.0  ± 5.6	99.3  ± 12.3	2.39  ± 0.20 ab	30.3  ± 3.5	0.109  ± 0.005 ab
HQS	27.6  ± 0.8	31.5  ± 4.1	71.5  ± 3.1	52.3  ± 4.2	84.8  ± 10.3	1.86  ± 0.12 a	28.4  ± 3.5	0.106  ± 0.007 a
Nutrients	27.3  ± 0.6	32.1  ± 5.3	64.6  ± 5.1	54.4  ± 5.6	101.0  ± 9.6	2.36  ± 0.18 ab	33.7  ± 3.2	0.153  ± 0.014 bc
HQS + nutrients	27.1  ± 0.9	32.9  ± 4.6	76.3  ± 3.0	62.3  ± 4.3	90.2  ± 11.7	1.87  ± 0.15 a	29.7  ± 3.2	0.125  ± 0.010 ab
Understory	24.0  ± 1.0	28.3  ± 3.4	71.8  ± 2.9	57.8  ± 3.8	149  ± 21	3.73  ± 0.30 b	43.1  ± 5.3	0.202  ± 0.017 c
Analysis of  variance results	P = 0.336	P = 0.782	P = 0.305	P = 0.536	P = 0.069	P < 0.001	P = 0.140	P < 0.001

Stem height is an indication of plant size. One-way analysis of variance results are presented for all parameters. Data are presented as mean ± se for 15 individuals per treatment and 27 understory plants. Statistical differences (P ≤ 0.05) after a multiple comparison procedure was performed (Tukey test) are indicated by different letters (a, b, and c).

DISCUSSION

The present study strongly suggests that stems are used as temporary carbohydrate reservoirs for fruit maturation in T. erectum. A complete defoliation treatment during fruit maturation under natural conditions has already suggested the possibility of carbohydrate accumulation in the stem (L. Lapointe and A. Deslauriers, unpublished data). Stem tissue was previously found to be a temporary carbon reservoir in cereals (Schnyder, 1993) and other agricultural plants (Incoll and Neales, 1970; Pate et al., 1983; Yamagata et al., 1987); however, in most species stored carbohydrates support only part of the fruit development (Rawson and Evans, 1971; Bonnett and Incoll, 1992). In the current study of T. erectum it appeared that stem carbohydrate content was sufficient to support complete fruit development. Establishing a temporary carbohydrate reservoir would allow spring ephemerals to maximize photosynthesis when irradiances are high in early spring and to develop mature fruit more often or to produce larger fruit than plants relying on low summer light for photosynthesis to support fruit development.

Rhizome carbohydrate reserves were replenished quickly in the spring and leaves did not translocate significant amounts of carbohydrate to the rhizome over the summer. In many other perennials, carbohydrate accumulation in underground parts occurs much later, in August or September (Bradbury and Hofstra, 1977; Cyr et al., 1990; Zasada et al., 1994). In some arctic plants, rhizome starch contents are replenished quickly (early July), well before the end of the growing season (Fonda and Bliss, 1966; Chapin et al., 1986). But there are other arctic plants with typical late-season starch accumulation (Chapin et al., 1986). The accumulation of carbohydrate reserves early in the season might reflect the short spring period, when ephemerals can photosynthesize at a maximum rate before conditions change. Sixty percent of the rhizome reserves was replenished by current photosynthesis, the rest appeared to be long-term carbohydrate reserves. Long-term reserves may only be used after a catastrophe such as fire, severe defoliation, or freezing (Whigham, 1984), or may be used to allow fruit set in perennial species (Stephenson, 1981).

Buds in T. erectum were visible at the end of the growing season and contained partly developed leaves and flowers. A large fraction of the rhizome starch content appeared to be translocated to the bud throughout autumn, and by mid-November, rhizome starch content was the same as in early spring. Only a small fraction of the autumn decrease in starch content may be attributed to the increased Suc content (23%) and rhizome respiration. Soil temperature decreased rapidly in the autumn (data not shown), which would minimize carbon respiration losses. Early carbon translocation to the bud probably accelerates shoot growth in the spring, which then only requires water uptake for cell elongation.

Sugar accumulated in stem tissue of T. erectum in the spring. Cereals also accumulate maximum carbohydrate in stems under high irradiance. Shaded cereals (30% of full sunlight) could only accumulate about one-half of the carbohydrates accumulated by full-sun plants (Judel and Mengel, 1982). In this study irradiance in the understory over summer, 1% of full sunlight (data not shown), was much lower than that used in a previous study (Judel and Mengel, 1982). We found no increase in stem carbohydrate content in T. erectum over the summer, suggesting that the carbohydrates had accumulated in stems only during the conditions allowing high photosynthetic rates.

The idea that the stem in T. erectum was used as a temporary carbohydrate reservoir was supported by several observations. First, we found more carbohydrate in the stem than is required for stem tissue respiration. Second, most of these carbohydrates were reducing sugars and not Suc, which suggests that carbohydrates were not mobile sugars that shifted from the leaves to the rhizome, but rather were stored carbohydrates. Third, when irradiances decreased in early June (Vézina and Grandtner, 1965), stem sugar content did not decrease, acting as a temporary carbon reservoir and not actively translocated between leaves and rhizome. There is evidence from sequential ¹⁴C labeling in wheat (Bell and Incoll, 1990) for separate carbohydrate pools for transport and storage in stems. The total amount of carbohydrates stored in the stem was low compared with that in the rhizome, but was sufficient to allow complete fruit development (Table I). The percentage of biomass in the stem slowly decreased during fruit maturation (L. Lapointe and A. Deslauriers, unpublished data), which suggests that, similar to wheat (Bell and Incoll, 1990), the decrease in stem dry biomass during fruit maturation in trillium is a reflection of source and sink exchanges.

The cut-stem experiment confirmed that the carbohydrate reserves in the stem were sufficient to support fruit development. Seeds were slightly smaller than seeds matured on complete plants (rhizome and leaves present), but all cut stems matured a fruit. In defoliation experiments using double-stem T. erectum plants, the leafless stem produced smaller seeds when the nondefoliated stem senesced early compared with when it senesced late (A. Deslauriers and L. Lapointe, unpublished data). Therefore, leaves may play a role during fruit maturation, but are not essential for fruit development. Very low irradiances in the understory mean leaf photosynthetic rates are low during the summer (L. Lapointe and A. Deslauriers, unpublished data). Stem photosynthesis is likely to be limited, since the stem is vertical and shaded by the leaves. The fruit is always dark red and cannot fix carbon. A major source of carbon for fruit development in T. erectum must be the carbohydrates stored in the stem during spring.

In the cut-stem experiment we added HQS to control bacterial growth (Rogers, 1973; Ketsa and Boonrote, 1990). It appeared that HQS was slightly harmful to the fruit and seed development. Cut stems fed with nutrients did not produce larger seeds or fruits compared with stems maintained in distilled water. The only effect of the nutrient treatments was on the ratio of dry to fresh mass of the fruit. We noticed that the fruits produced by cut stems were fleshier than the fruits harvested from the field. This could have been because of higher RH in the growth cabinets and the absence of wind. Since nutrients played such a minor role in fruit and seed development, it may be that the stem also accumulates nutrients during spring. Spring ephemerals may take advantage of the nutrient flushes right after snow melt (Muller, 1978; Hicks and Chabot, 1985). T. erectum leaves contain high levels of nitrogen (L. Lapointe, unpublished data). Under natural conditions these leaves slowly senesce during fruit maturation, with the possibility of translocation of nutrients from leaves to fruit, as has been shown for wheat (Waldren and Flowerday, 1978). However, this postulated translocation does not seem to be important enough to affect seed biomass in leafless plants. Stem nutrient content may be sufficient to allow normal seed development to take place.

Stems have previously been shown to be a temporary carbon reservoir in cereals (Schnyder, 1993) and other grasses (Warringa and Kreuzer, 1996), in soybean (Yamagata et al., 1987), cowpea (Pate et al., 1983), Plantago major and Urtica dioica (den Hertog et al., 1996), and Jerusalem artichoke (Helianthus tuberosus; Incoll and Neales, 1970). However, in tulips, the only other spring ephemeral in which stem carbohydrate content has been investigated, there was no accumulation of carbohydrates in stem tissues (Ho and Rees, 1976). Stem contribution to ear growth in cereals varies from 3 to 40% in unstressed plants (Rawson and Evans, 1971; Austin et al., 1977; Bonnett and Incoll, 1992), but can reach 70 to 100% in stressed plants (Gallagher et al., 1976; Scott and Dennis-Jones, 1976). In T. erectum, stem carbohydrate contribution to fruit development was high and resembled the situation in stressed cereals. The period of carbohydrate accumulation is shortened in stressed plants and the importance of carbohydrates stored in stems may be related to the time the plant had to accumulate carbohydrates during periods of high photosynthetic rates before fruit development began.

In spring ephemerals the establishment of a temporary carbon reservoir enables the plant to benefit from high irradiances in the spring. The presence of a temporary carbon reservoir may alleviate some of the sink limitations present before fruit development in fast-growing species such as cereals (Bell and Incoll, 1990). It seems probable that the development of a temporary carbohydrate reservoir could alleviate some of the sink limitations of slow-growing species as well, since their growth is often also strongly restricted by sink activity. Carbohydrate accumulation in stems would not compete with rhizome carbohydrate storage in T. erectum as it does in cereals, where there does not seem to be competition between grain filling and stem carbohydrate accumulation (Schnyder, 1993). T. erectum leaves were capable of replenishing rhizome carbohydrate reserves within 3 weeks after full development, when plants were still flowering, and accumulation of carbon in the stem made better use of the high spring irradiances. Stephenson (1981) suggested that perennial species may use reserves accumulated in previous years to support fruit maturation. In T. erectum, and probably in other species as well (Sohn and Policansky, 1977), long-term carbohydrate reserves are probably aimed at future growth and are not used for fruit development. These plants develop a short-term carbohydrate reservoir that supports current-year reproduction.

Abbreviation:

HQS

8-hydroxyquinone