Effects of elevated CO2 and temperature on seed quality.

Successful crop production depends initially on the availability of high-quality seed. By 2050 global climate change will have influenced crop yields, but will these changes affect seed quality? The present review examines the effects of elevated carbon dioxide (CO2) and temperature during seed production on three seed quality components: seed mass, germination and seed vigour. In response to elevated CO2, seed mass has been reported to both increase and decrease in C3 plants, but not change in C4 plants. Increases are greater in legumes than non-legumes, and there is considerable variation among species. Seed mass increases may result in a decrease of seed nitrogen (N) concentration in non-legumes. Increasing temperature may decrease seed mass because of an accelerated growth rate and reduced seed filling duration, but lower seed mass does not necessarily reduce seed germination or vigour. Like seed mass, reported seed germination responses to elevated CO2 have been variable. The reported changes in seed C/N ratio can decrease seed protein content which may eventually lead to reduced viability. Conversely, increased ethylene production may stimulate germination in some species. High-temperature stress before developing seeds reach physiological maturity (PM) can reduce germination by inhibiting the ability of the plant to supply the assimilates necessary to synthesize the storage compounds required for germination. Nothing is known concerning the effects of elevated CO2 on seed vigour. However, seed vigour can be reduced by high-temperature stress both before and after PM. High temperatures induce or increase the physiological deterioration of seeds. Limited evidence suggests that only short periods of high-temperature stress at critical seed development stages are required to reduce seed vigour, but further research is required. The predicted environmental changes will lead to losses of seed quality, particularly for seed vigour and possibly germination. The seed industry will need to consider management changes to minimize the risk of this occurring.

INTRODUCTION

With global change, atmospheric carbon dioxide (CO2) concentration is predicted to rise from today's value of c. 370–550 ppm by 2050 and could reach between 730 and 1010 ppm by 2100 (Solomon et al. 2007). This, combined with other atmospheric changes, is projected to increase global mean temperatures by 1·4–5·8 °C (Houghton et al. 2001). Jaggard et al. (2010) concluded that CO2 enrichment was likely to allow yield increases of c. 13% in most C3 crops, but yields of C4 crops are not expected to change. However, increasing temperatures may negate these increases in C3 crops, particularly if they occur during reproductive growth (Allen & Boote 2000; Wheeler et al. 2000). Gornall et al. (2010) noted that extreme weather events are more likely to occur in the changed climate of the future, and predicted that over much of the world's crop land, maximum daily temperature highs may be increased by around 3 °C by 2050.

A major challenge ahead for those involved in the seed industry, therefore, is to provide cultivars that can maximize future crop production in a changing climate (Ainsworth et al. 2008; Bruins 2009; Ceccarelli et al. 2010). Ainsworth et al. (2008) considered that this will be possible within a decade.

Successful crop production in any environment depends initially on the quality of the seed being sown. The term ‘seed quality’ is used in practice to describe the overall value of a seed lot for its intended purpose (Hampton 2002), and includes the components of species and cultivar purity, seed mass (size), physical purity, germination, vigour, moisture content and seed health. The present review examines the effects of increased CO2 and increased temperature on three of these seed quality components, seed mass, germination and vigour.

SEED MASS

Within the seed industry, seed size is commonly denominated by the mean seed weight, often expressed as ‘thousand seed weight’, the weight of 1000 seeds of the seed lot. However, seed size refers to volume, while seed weight and seed mass refer to density, which are different traits (Castro et al. 2006). Seed mass in crop cultivars is considered the least variable of the seed yield components because of plant breeding for increased seed uniformity (Almekinders & Louwaars 1999) and the removal of small seeds during the seed cleaning processes. Factors affecting seed mass, including genetic factors, water availability and nutrient availability were reviewed by Castro et al. (2006).

Increased atmospheric CO2 concentrations might be expected to increase seed mass because of increased plant assimilate availability (Jablonski et al. 2002), but the reported effects of elevated CO2 are highly variable among species (Miyagi et al. 2007; Hikosaka et al. 2011). Different studies have reported seed mass to increase (Musgrave et al. 1986; Baker et al. 1989; Dijkstra et al. 1999; Steinger et al. 2000; Quaderi & Reid 2005), show no change (Edwards et al. 2001; Prasad et al. 2002) and decrease (Huxman et al. 1998; Smith et al. 2000; Wagner et al. 2001) in response to elevated CO2. Jablonski et al. (2002) conducted a meta-analysis of 184 CO2 enrichment studies from 79 species and found a mean 4% increase in seed mass, with the response being greater in legumes (+8%) than non-legumes (+3%), and absent in C4 plants. Considerable variation in seed mass in response to elevated CO2 was also reported within species. Hikosaka et al. (2011) reported the enhancement ratio of seed mass per plant (seed mass in elevated CO2/seed mass in ambient CO2) ranged from 0·75 to 4·45 in rice (Oryza sativa L.), from 0·93 to 1·87 in soybean (Glycine max (L.) Merrill), and from 0·88 to 2·07 in wheat (Triticum aestivum L.).

Jablonski et al. (2002) found a 14% reduction in seed nitrogen (N) in response to elevated CO2 averaged across all 79 species in their analysis, although there was significant variation; seed N was not reduced in legumes, but was reduced in non-legumes. Hikosaka et al. (2011) suggested that seed mass could only increase when N became more available at elevated CO2 concentrations. Legumes may use increased carbon (C) gain under elevated CO2 for increased nitrogen fixation (Allen & Boote 2000), and can therefore increase seed mass without decreasing seed N. In non-legumes, seed mass increases may result in a decrease in seed N concentration. In some species, this decreased seed N may be at the expense of seed quality (Fenner 1991; Andalo et al. 1996).

Increasing temperature can negate the response to elevated CO2 (Prasad et al. 2002) and may reduce seed mass (Spears et al. 1997) because of the resulting acceleration in seed growth rate (dry matter accumulation) and reduction in the duration of seed filling (Weigand & Cueller 1981; Young et al. 2004). However, a reduction in the rate of seed dry matter accumulation can also occur (Gibson & Paulsen 1999) and seed mass has also been reported not to change, or sometimes increase, in response to temperature increase (Peltonen-Sainio et al. 2011). A reduced seed mass for a seed lot does not necessarily mean a loss in other seed quality attributes. Many studies have shown no relationship between seed mass and germination (Castro et al. 2006) or seed mass and seed vigour (Powell 1988).

GERMINATION

For high-quality seed lots, germination (defined as the process that begins with imbibition and which is completed by the production of a normal seedling; ISTA 2011) is desired by the seed industry to be as close to 100% as possible. The germination of a seed lot can be negatively affected by the conditions the seeds are exposed to during harvesting, drying, cleaning and storage, but can also be reduced by unfavourable environmental conditions in the field during seed growth and development (Dornbos 1995), particularly temperature, rainfall and relative humidity (Egli et al. 2005).

Seed germination in response to elevated CO2 has been reported to decrease (Farnsworth & Bazazz 1995; Andalo et al. 1996; Quaderi & Reid 2005), show no change (Huxman et al. 1998; Steinger et al. 2000; Thomas et al. 2009; Way et al. 2010) or increase (Wulf & Alexander 1985; Ziska & Bunce 1993; Edwards et al. 2001). The responses vary among species (Ziska & Bunce 1993) and genotypic variation has also been reported (Andalo et al. 1996).

Elevated CO2 has been shown to increase the C/N ratio in seeds (Huxman et al. 1998; Steinger et al. 2000; He et al. 2005) and in non-legumes, seed N reduction can occur when seed mass is increased by elevated CO2 (see previous section). High seed N is an advantage for germination rate (Hara & Toriyama 1998), but not germination per se. However, a change in C/N ratio can lead to a decrease in seed protein content, resulting in a reduction in the ability of the seed to supply the amino acids required for the de novo protein synthesis necessary for embryo growth in the germinating seed. This could reduce seed viability (Andalo et al. 1996).

Elevated CO2 also increases ethylene production (Esashi et al. 1986) and Ziska & Bunce (1993) suggested that an increased availability of ethylene may have been the reason for the stimulated germination they reported. Ethylene is implicated in the promotion of germination of non-dormant seeds of many species (Leubner-Metzger 2006).

In different plant species, sometimes even small differences in temperature during seed development and maturation can have an influence on germination (Gutterman 2000). High temperatures during seed filling frequently disrupt normal seed development, which increases the proportion of seeds that are shrivelled, abnormal and are of lower quality (Spears et al. 1997). However, it has been shown that after removal of these seeds, the germination of the remaining seeds decreases as mean maximum temperature during seed filling increases (Khalil et al. 2001, 2010; Egli et al. 2005; Thomas et al. 2009; Table 1). High-temperature stress before the developing seeds achieve physiological or mass maturity (PM – the end of the seed filling phase) is likely to inhibit the ability of the plant to supply the seeds with the assimilates necessary to synthesize the storage compounds required during the germination process (Dornbos & McDonald 1986), and/or the seeds suffer physiological damage (see McDonald & Nelson 1986; Coolbear 1995; Powell 2006) to the extent that the ability to germinate is lost.

Table 1.

Effect of temperature during seed development on seed germination and seed vigour of two soybean cultivars (adapted from Spears et al. 1997)

	Temperature regime and duration*
	27/22 °C		33/28 °C		38/33 °C
	R5-PM	R5-R8	R5-PM	R5-R8	R5-PM	R5-R8
Germination (%)
McCall†	100	100	98	100	94	63
Hutchenson	100	100	99	98	57	14
Germination (%) after accelerated ageing*‡
McCall†	98	100	73	86	8	7
Hutchenson	98	95	43	86	2	1
Conductivity‡ (μS/cm/g)
McCall†	5·5	5·4	7·0	6·3	16·1	23·9
Hutchenson	6·5	6·2	9·1	10·6	29·8	35·9

Day/night temperatures with 10 h at the day temperature; R5=beginning of seed fill; PM=physiological maturity; R8=harvest maturity.

Soybean cultivars; McCall=indeterminate growth habit; Hutchenson=determinate growth habit.

Seed vigour tests.

High-temperature stress after PM can also sometimes reduce germination (Green et al. 1965; Table 1), but more often reduces seed vigour (see next section).

The relationship between temperature during seed development and subsequent seed germination requires further investigation. For example in soybean, temperatures (32–38 °C) that reduced the germination of some cultivars in controlled environments did not vary during seed filling, in contrast to field temperatures which can vary substantially (Egli et al. 2005), and the plants were at these temperatures from anthesis until seed harvest. However, there may be critical periods during seed development when seeds are particularly sensitive to temperature (Egli et al. 2005; Shinohara et al. 2006). This was investigated for pea (Pisum sativum L.) by Shinohara et al. (2006), who showed that when plants were exposed to a day/night temperature of 30/20 °C for 4 days (=240 °C h above a base temperature (Tb) of 25 °C) at the beginning of seed filling and then returned to the field until seed harvest, germination was significantly reduced in one of two cultivars (Table 2). Exposure to these conditions at later stages of seed development did not affect germination.

Table 2.

Effect of high temperature (30/25 °C) for 4 days at different stages of seed development and maturation in two cultivars of pea (Pisum sativum L.) on seed quality components (adapted from Shinohara et al. 2006b)

Stage at treatment*	Mean seed weight (g)		Germination (percentage)		Hollow heart (proportion)		Average conductivity† (μS/cm/g)
	Alderman‡	E. Onward‡	Alderman	E. Onward	Alderman	E. Onward	Alderman	E. Onward
Control	379	350	95	93	0·04	0·00	220	363
S1	309	289	84	97	0·01	0·03	315	368
S2	326	262	86	92	0·24	0·04	367	374
S3	391	345	86	96	0·08	0·00	339	467
S4	378	362	88	98	0·01	0·00	371	475
S5	376	337	92	96	0·04	0·00	424	420
s.e.d. (d.f.)	11 (18)	9 (18)	6·7 (18)	3·9 (18)	0·054 (18)	0·020 (18)	31 (140)	123 (140)

S1=beginning of seed filling (810 mg/g SMC); S2=rapid seed filling (700 mg/g SMC); S3=PM (630 mg/g SMC); S4=beginning of desiccation (440 mg/g SMC); S5=harvest maturity (230 mg/g SMC); SMCs are mean of the two cultivars.

Data are the average of 25 results in the single seed conductivity vigour test.

Pea cultivars.

s.e.d. (between cultivars)=10 (mean seed weight), 5·4 (germination), 0·041 (hollow heart) and 71 (average conductivity).

SEED VIGOUR

While the term germination has long been used to describe the planting value of a seed lot (ISTA 2011), when conditions in the seed bed are less than optimal the germination test is a poor predictor of field emergence (Dornbos 1995), suggesting that a further physiological aspect to seed quality exists – seed vigour (Powell 2006). Seed vigour is defined by ISTA (2011) as ‘the sum of those properties that determine the activity and level of performance of seed lots of acceptable germination in a wide range of environments’, or more simply, the ability of a high germination seed lot to emerge under seed-bed stress.

While there have been reports that elevated CO2 increases or decreases seedling vigour (i.e. growth rate or biomass production) because of the effect on seed mass (Huxman et al. 1998; Steinger et al. 2000), there have been no reports on the effects of elevated CO2 on seed vigour.

Seed vigour is reduced by high-temperature stress before PM (Spears et al. 1997; Egli et al. 2005; Shinohara et al. 2006; Table 1) and after PM (TeKrony et al. 1979, 1980; Gibson & Mullen 1996; Hampton 2000). Shinohara et al. (2008) examined the relationship between vigour test results for 262 garden pea seeds lots produced in New Zealand and climate data in five regions over four consecutive production seasons, and while regional and seasonal variation for vigour occurred, these variations were significantly associated with temperature during seed development – generally the higher the temperature, the lower the seed vigour.

The susceptibility of seeds to loss of vigour following high-temperature stress depends on the stage of development (Shinohara et al. 2006). Using hollow heart, a physiological disorder of germinating pea seeds (Halligan 1986) as a seed vigour indicator (Castillo et al. 1993), Shinohara et al. (2006) found in a field study that hourly thermal time (HTT, measured in degree hours, °Ch, Tb=25 °C) when seeds were at the green-wrinkled pod stage (700–800 mg/g seed moisture content (SMC)) was significantly correlated with hollow heart incidence at harvest, and that 100 °Ch were required to induce the condition. There was no such relationship between HTT and hollow heart after PM. While there were cultivar differences, for one cultivar there was a linear increase in hollow heart incidence as the degree hours (°Ch) increased. In a follow-up controlled environment study, Shinohara et al. (2006) confirmed this result, by demonstrating that exposure to day/night temperatures of 30 and 25 °C, respectively, for 4 days (240 °Ch, Tb=25 °C) at the green-wrinkled pod stage induced hollow heart, but exposure to the same conditions at the beginning of seed fill (>800 mg/g SMC), PM (550–650 mg/g SMC) or after PM did not (Table 2). Single-seed conductivity (which is an indicator of cell membrane integrity – see Powell 2006) was increased only after exposure of the developing seeds to the high temperature at or after PM, and not before (Table 1).

Seed vigour loss is associated with seed physiological deterioration (Powell 1988; Hampton & Coolbear 1990), and lipid peroxidation is the most frequently cited cause (McDonald 1999). Lipid peroxidation causes cellular degeneration through free radical assault on important cellular molecules and structures (Wilson & McDonald 1986). McDonald (1999), in his model of seed deterioration, proposed four types of cell damage, viz. mitochondrial dysfunction, enzyme inactivation, membrane degradation and genetic damage.

Grass & Burris (1995) reported that high-temperature stress of the parent plant caused mitochondrial degeneration and reduced adenosine triphosphate (ATP) accumulation, energy levels and rates of oxygen uptake in imbibing wheat embryos (Table 3), providing clear evidence for metabolic changes at the mitochondrial level in early seed germination in response to heat stress during seed development and maturation. High temperatures during reproductive growth increase seed cell membrane damage (Nilsen & Orcutt 1996; Shinohara et al. 2006) so that electrolyte leakage from seeds is increased (Castillo et al. 1994; Spears et al. 1997; Shinohara et al. 2006). High leachate conductivity in pea has been associated with dead/deteriorating tissue on the abaxial surface of the cotyledons (Powell 1985; Shinohara et al. 2006), and on the adaxial cotyledonary surface for hollow heart (Don et al. 1984; Shinohara et al. 2006). However, temperature stress also results in damage to the shoot apical meristem of the embryonic axis (Fu et al. 1988; Senaratna et al. 1988). Membrane disorganization would reduce mitochondrial efficiency and may allow the release of peroxidative enzymes capable of causing subsequent cellular damage after imbibition has begun (McDonald 1999).

Table 3.

Effect of temperature during seed development and maturation on nucleotide content, mitochondrial respiration rate and adenylate energy charge (AEC) of excised wheat embryos after 4 h imbibition (adapted from Grass & Burris 1995)

	Temperature regime*
	20/15 °C	28/21 °C	36/29 °C
Nucleotide content† (nmol/g)
AMP	400	370	448
ADP	571	518	412
ATP	452	393	141
Total	1422	1180	1001
State 3 mitochondrial respiration rate (nmol O2/min)	9·7	6·9	5·3
AEC‡	0·52	0·51	0·35

Day/night with 8 h day temperature and 16 h night temperature.

AMP, adenosine monophosphate; ADP, adenosine diphosphate; ATP, adenosine triphosphate.

AEC expressed as the ratio (ATP+0·5ADP/ATP+ADP+AMP)=energy status of the seed (Atkinson 1968).

If heat stress leads to mitochondrial dysfunction and membrane damage, it may also result in reduced enzyme activity (e.g. decreased α-amylase – Shephard et al. 1996) and genetic damage (e.g. decreased DNA synthesis – Cruz-Garcia et al. 1995). Whether these and other seed deteriorative changes (see McDonald 1999) occur following heat stress during seed development and maturation is yet to be determined.

CONCLUSIONS

The environment during seed development and maturation can significantly reduce seed quality (Dornbos 1995; Gusta et al. 2004; Egli et al. 2005; Shinohara et al. 2008), particularly seed vigour. How likely is it that elevated CO2 levels and temperature increases of up to 3 °C by 2050 will further increase this loss of seed quality? To answer this question accurately will require substantially more research in order to determine the critical periods during seed development when seeds are sensitive to environmental stresses, and for temperature, how this interacts with the duration of exposure to elevated temperatures which are deleterious to seed quality. For example, Shinohara et al. (2006) found that during the rapid seed filling stage in pea, a temperature of 30/25 °C for 2 days (120 °Ch at Tb=25 °C) did not induce hollow heart, but 4 days induced hollow heart in one cultivar (see Table 2), and 6 days (360 °Ch) induced the condition in both cultivars used (0·43 in cvar Alderman and 0·23 in cvar Early Onward).

From the information that is available, it can be concluded that predicted environmental changes will lead to the increased occurrence of loss of seed quality, particularly seed vigour and possibly germination. While seed mass will also change, this does not necessarily imply any negative effect on germination or vigour. To minimize the risk of reductions in seed quality the seed industry will therefore have to consider:

Moving seed production to the limits of adaptation either in latitude (northern or southern) or in elevation (highland and mountainous regions) in order to reduce the chances of environmental stress (Egli et al. 2005; Shinohara et al. 2008).

Changing sowing date so that seed filling occurs at lower temperatures (Castillo et al. 1994; Egli et al. 2005; Shinohara et al. 2006). The latter authors demonstrated that for the pea cultivar Alderman, HTT (Tb=25 °C) during the rapid seed filling stage was 198, 106 and 21 °Ch for sowings at the same site in September, October and November, respectively, and the number of hours during this stage when temperature exceeded 25 or 30 °C also reduced as sowing date was delayed (Table 4).

Table 4.

Effect of sowing date at the same field site on HTT (Tb=25 °C) and the number of hours of exposure to temperature exceeding 25 and 30 °C during the period when SMC was between 700 and 800 mg/g for garden pea cvar Alderman (adapted from Shinohara et al. 2006).

Sowing date	SMC 700–800 mg/g				PM*
	Period†	HTT	No. of hours
		(°Ch)	>25 °C	>30 °C
26 Sep	31 Dec–4 Jan	198	45	19	11 Jan
21 Oct	8 Jan–13 Jan	106	38	8	21 Jan
19 Nov	1 Feb–7 Feb	21	5	0	18 Feb

Dates when seeds were adjudged to have reached PM.

Period when SMC was between 700 and 800 mg/g.

Exploring genotypic differences in the ability to acquire and retain good seed quality in stressful environments, firstly among existing cultivars (Spears et al. 1997; Shinohara et al. 2006), and in the breeding of new cultivars (Ainsworth et al. 2008).