Wheat growth and development is often described in terms of the Zadoks decimal scale which helps to standardise the growth stages observed during wheat plant development (Zadoks et al. 1974). A number of publications discuss wheat development in Australia with reference to this scale (Bowden et al. 2008; GRDC 2013c; GRDC 2015c).
Wheat does not reproduce vegetatively, so all reproduction is sexual.
The time and duration of flowering is dependent upon geographical location. Sunny weather and temperatures of at least 11 – 13 °C are required for flowering (OECD 1999). Florets on the spike of the main tiller open first and flowering commences in the middle of each spike and proceeds synchronously towards the tip and the base. In a study of wheat sown in May in Narrabri (northern NSW) the time from sowing to flowering was approximately 105 - 120 days and the time from flowering to maturity was approximately 35 - 45 days, based on a ten year period from 1990 - 2000 (in the ten year period 1990-2000; Sadras & Monzon 2006).
4.2 Pollination and pollen dispersal
Flowers may be described as chasmogamic - opening to expose flowers and stamens to the air - or cleistogamic - remaining closed and thus necessarily self-fertilising (Sethi & Chhabra 1990). Considerable differences in flower opening occur amongst varieties and species of wheat, and De Vries (1971) noted that 80-90% of bread wheats showed open flowers. The extent of flower opening is an important factor in influencing cross-pollination and potential gene flow during anthesis. Floral structure, anthesis and anther dehiscence patterns in wheat make it predominantly self-pollinating with low rates of out-crossing (Waines & Hegde 2003). Generally, wheat flowers lack nectaries to attract insects (Eastham & Sweet 2002) and the role of insects in cross-pollination is considered to be minimal (Glover 2002). Any out-crossing that may occur is facilitated by wind dispersal.
In wheat, the stamens are smaller and produce fewer pollen grains (1000-3800 pollen grains per anther; 450,000 pollen grains per plant) than other cereal grasses. This compares to approximately four million pollen grains per ear for rye (Secalecereale L.) and 18 million pollen grains per maize tassel (Zea mays L., see de Vries 1971).
In general, approximately 80% of the pollen from an anther protruding from the spikelet is dispersed into the air, but pollen shedding capacity may depend on the cultivar. In a three year field study of wheat varieties, the number of anthers emerging from flowers was found to vary from 14 to 80% of anthers per ear (D'Souza 1970). In a study of 22 wheat cultivars, shedding ranged from 3 to 80% of the pollen produced was shed, with a strong influence of cultivar - tall varieties with more pollen grain/anther and longer filaments shed greater quantities of pollen outside the florets (Beri & Anand 1971).
The pollen load in the air at a given time is a function of the amount of pollen produced per anther, the amount of anther extrusion and the number of anthers per unit in a given area (discussed in (Virmani & Edwards 1983).
However, physical movement of pollen does not necessarily result in gene flow (Waines and Hegde 2003). Cross-pollination rates are usually less than 1% but rates up to 6% or higher can be observed, depending on cultivars and environmental conditions (Hucl 1996; Hucl & Matus-Cádiz 2001).
Wheat pollen grains are relatively heavy compared to other grasses, short-lived (up to 30 minutes under optimal field conditions (20 °C, 60 % relative humidity)) and typically travel very short distances in still air (Lelley 1966; de Vries 1971). A majority of studies suggest that more than 90 % of wheat pollen falls within three metres of the source (Hegde & Waines 2004). Cross-pollination due to insects is deemed unlikely as wheat flowers have no nectaries and produce relatively small quantities of pollen (de Vries 1971; Treu & Emberlin 2000). Any outcrossing occurring is facilitated by wind dispersal of pollen (Treu & Emberlin 2000). Studies in small pollinator blocks have shown that wheat pollen grains can travel up to 60 m from the pollen source (reviewed by (Waines & Hegde 2003). However, gene flow remains limited as less than 1 % average cross-pollination is observed beyond 6 m (D'Souza 1970; de Vries 1974). However, a number of researchers have also reported long distance pollen movement. At the field scale (Matus-Cádiz et al. 2004) and commercial scale (Matus-Cádiz et al. 2007) long distance pollen dispersal has been observed at trace levels (see section 9.1). Laboratory experiments have shown that pollen can travel a distance of about 60 m at a height of one metre (D'Souza 1970). In field experiments, Wilson (1968) found 10 % seed set on male sterile wheat plants 30 m from the pollen donor plants.
Physical movement of pollen does not necessarily result in gene flow (Waines & Hegde 2003) and long distance pollen movement does not necessarily result in a proportional increase in gene flow. Pollen grains quickly desiccate after release from the anthers (Heslop-Harrison 1979) and under field conditions, the viability of pollen grains may be less than 30 minutes (D'Souza 1970). Under optimal conditions of 5 °C and 60 % relative humidity, however, pollen can remain viable for over 90 minutes (D'Souza 1970). Field conditions including temperature, relative humidity and wind intensity have a great influence on pollen viability and pollen movement. Extreme cold or hot temperatures are unfavourable for pollination and fertilisation, and weather conditions also play an important role. Humid weather makes the pollen heavy, limiting dispersal distance from the plant, while dry weather causes desiccation and loss of viability (D'Souza 1970).
Heslop-Harrison (1979) reported that after release, wheat pollen attaches to the stigmatic branches and water is absorbed by the pollen grain through gaps in the stigma cuticle. This process enables the pollen tube to grow, which in turn facilitates fertilisation. Pollen tube growth is initiated 1-2 hours after pollination and fertilisation takes place after an additional 30 - 40 hours (de Vries 1971). The duration of stigma receptivity is an important consideration in understanding wheat reproductive biology. Estimates of the duration of receptivity vary from a few days to up to 13 days and estimates vary not only due to experimental and environmental conditions but also due to the methods used to determine receptivity (de Vries 1971). An extensive review of wheat flowering biology is available (de Vries 1971).
4.3 Fruit/seed development and seed dispersal
The rate of endosperm cell division is influenced by light intensity, water stress, temperature and genotype (Wardlaw 1970; Brocklehurst et al. 1978). Starch deposition begins 1 - 2 weeks after anthesis and initiates a 2 - 4 week period of linear increase in kernel dry weight. This process is also influenced by water stress, temperature and genotype (Simmons 1987). The growth and final weight of an individual kernel depends on the spikelet and floret position (Kirby 1974), the kernels formed in central spikelets and proximal florets within an individual spikelet are usually largest (Simmons & Crookston 1979). Each wheat ear can produce approximately 30 to 50 kernels while the number of ears a wheat plant produces depends on the number of tillers produced and the number of tillers that produce a mature ear (Setter & Carlton 2000a).
Wheat is generally considered to have lost its natural seed dispersal mechanisms with domestication. The genes that control seed dispersal have been characterised in domesticated wheat and both modern durum and bread wheats were found to have the genotype brbrtgtgQQ (Li & Gill 2006): Br controls rachis brittleness, Tg controls glume toughness, and Q controls seed threshability. In wild ancestral wheats, shattering is caused by a brittle rachis, which is conferred by a dominant Br allele. A recessive mutant allele br at this locus in modern wheats produces a non-brittle spike (Li & Gill 2006).
When rain coincides with harvest, pre-harvest sprouting can occur i.e. grains may germinate while still on the ear of the parent plant. Thus, in cereal crops some degree of dormancy during seed development can be advantageous. Kernels that mature under cool conditions are more dormant than those ripened under warm conditions (Austin & Jones 1975). In Australia, rising temperatures late in the development of the wheat crop, particularly after heading, are considered an important yield–limiting factor (Wardlaw & Moncur 1995). However, wheat cultivars vary in their response to high temperature during kernel filling and the relationship may not be a simple temperature effect (Stapper & Fischer 1990; McDonald et al. 1983). Wardlaw and Moncur (1995) reported a significant drop in kernel dry weight at maturity, with significant variation in response, ranging from a 30-60% decrease in kernel dry weight at maturity, for a rise in temperature from 18/13C (day/night) to 30/25 C (day/night).
Kangaroos (Macropus spp.), rabbits (Oryctolagus cuniculus), mice (Mus musculus) and rats (Rattus spp.) are known pests of wheat (Hill et al. 1988; AGRI-FACTS 2002) and therefore potential distributors of viable wheat seeds. Small dormant seeds are more likely to survive chewing and digestion (Malo & Suárez 1995). White wheats have large seeds with low dormancy and a thin seed coat and are therefore expected to be easily broken down in the digestive system of mammals (Hansen 1994). Intact seed may make up to 30 % (wheat) or 15 % (barley) of dry matter in the faeces of cattle fed grain (Beauchemin et al. 1994), however, the germination rates of this seed were not measured. In other studies wheat seeds have been shown to germinate in the dung of cattle and sheep (Ovis aries), but not donkeys (Equus asinus), after consumption (Seman 2007). This indicates the potential for livestock to disperse viable wheat seed after consumption. Wheat seeds can also be dispersed in the wool of sheep (Ryves 1988).
Although rabbits are known pests of wheat plants, viable wheat seeds have not been found in rabbit dung (Malo & Suárez 1995). In a study that looked at the germination of seeds on dung from cattle, red deer (Cervus elephus), sheep, hare (Lepus capensis), rabbit and red grouse (Lagopus scotica), the number of seeds germinating was least on rabbit dung (Welch 1985). Similarly, a study that looked at viable grass seeds in dung from cattle, pronghorn (Antilocapra americana) and rabbit, found few seedling populations of any species emerged from rabbit dung (Wicklow & Zak 1983). Rodents may eat seeds, thus destroying them, at the seed source or they may hoard seed elsewhere and disperse the seed (AGRI-FACTS 2002).
Emus (Dromaius novaehollandiae) have been shown to disperse seeds (Calvino-Cancela et al. 2006), however germination rates are generally very low (Rogers et al. 1993; McGrath & Bass 1999). Viable seed from Avena sativa L., a grass from the same subfamily as wheat (Pooideae), was detected in emu droppings (Calvino-Cancela et al. 2006). It has been stated that seeds of wheat will also germinate after passage through an emu’s digestive system, although no experimental evidence was provided (Davies 1978).
The white wheat varieties have a thin seed coat (Hansen 1994) and are readily digested by birds (Yasar 2003). An unpublished study conducted under laboratory conditions showed that when wheat was fed to corellas and galahs (Eolophus roseicapillus), some wheat seeds remained intact following passage through the digestive tract, but at very low numbers (2.3 % and 0.7 %, respectively) (Woodgate et al. 2011). For that intact seed, the germination rate was 87.5 % from corellas and 100 % for galahs (Woodgate et al. 2011), such that overall germination rates under laboratory conditions were 2 % and 0.7 % respectively. In another study, seed of four crop species tested (maize, barley, safflower and rice) did not remain intact after passage through the digestive tract of birds (mallard duck, Anas platyrhyncos; ring-necked pheasant, Phasianus colchicus, red-winged blackbird, Agelaius phoeniceusand rock pigeon, Columba livia) (Cummings et al. 2008). However, the authors noted that seed remained intact within the oesophagus/crop and gizzard for several hours and this could be a mechanism for dispersal, i.e. if the birds were killed within hours of consuming the seed. Similarly, dispersal could occur via intact seed found on the muddy feet/legs (but not the feathers) of a few birds (Cummings et al. 2008). Ring-necked pheasants, mallard ducks, and rock pigeons have all been introduced into Australia (Atlas of Living Australia, accessed 23 February 2015). An extensive search of the literature did not identify any reports of birds transporting and dispersing wheat seed by taking panicles containing viable seed or seedlings from wheat crops.
A variety of insects are likely to feed on the wheat crop, but it is unlikely that most of these would contribute to the dispersal of seeds beyond the field. It is possible that ants may remove seeds for underground storage, but to depths where germination is highly unlikely. Although there are differences in ant behaviour and territory size across species, seed dispersal occurs at a local scale, such that seeds are usually only moved a few metres (Cain et al. 1998; Peters et al. 2003). Maximum seed dispersal distances by ants in Australia and the rest of the world are typically less than 40 m, with a mean dispersal distance of 0.96 m (Berg 1975; Beattie 1982; Gómez & Espadaler 1998).
It is important to remember that in Australia wheat is cultivated on about 14 million ha and produces as much as 26 million tonnes annually (Figure 1). Production on this scale involves considerable movement and loss of seed during transport, cultivation, harvest, storage, and processing; but also during distribution of animal feed stock, hay and straw. Wheat seeds have been dispersed on clothing (Ansong & Pickering 2014; Huiskes et al. 2014) and in the seed of other crop plants and grass seed (Conn 2012). Thus the greatest dispersal of wheat seed is likely through human intervention.