SEARCH

SEARCH BY CITATION

Keywords:

  • emergence patterns;
  • establishing canopy;
  • Galinsoga parviflora Cav.;
  • germination;
  • light environment;
  • thermal regime;
  • weed seeds

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Knowledge of the regulatory effects of the crop canopy on weed seed germination is necessary to understand fully the behaviour of weed seed banks during a crop cycle. It is well known that canopy presence interferes with seed germination through modifications to the light and thermal environment, but the changing effect of a growing canopy has not been assessed, thus precluding the identification of shading-intensity thresholds for triggering the different types of canopy-detection mechanisms in different seed populations.
  • 2
    Field experiments were performed with artificially modified thermal and light environments, using two types of seed banks (seeds buried or located at the soil surface). Conditions below a wheat canopy were modified to match light and thermal conditions prevailing on bare soil (i.e. soil without vegetation).
  • 3
    Most weed emergence patterns during the early stages of crop establishment were not modified by the thermal regime produced by the incipient canopy compared with a bare soil control. However, the reduction of the red : far-red ratio from the bare soil value of 1·2–0·9 below the wheat canopy reduced germination of some weed species located at the soil surface, and the effect could by reversed by far-red filters.
  • 4
    The weed Galinsoga parviflora developed two generations during the crop cycle. The modified light and thermal environments beneath an establishing wheat canopy was not sufficient to inhibit the germination of Galinsoga parviflora, even if seeds were located at the soil surface. Only for seeds of the second generation, dispersed from seeds of the first plant generation, was there sufficient modification of the photothermal environment below the wheat canopy to interfere with dormancy termination.
  • 5
    Synthesis and applications. Understanding seed responses to modifications in the photothermal environment below a crop canopy (e.g. wheat crop) should allow us to improve weed management strategies by manipulating crop canopy attributes. This could be achieved by modification of sowing date, crop density, spatial arrangement and genotype. For example, increasing the crop plant density would diminish the number of weeds emerging during the first phase of crop establishment. This strategy would be appropriate where weed seeds are predominantly located at the soil surface, typically found in a no-till cropping system.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Many weed seeds possess dormancy mechanisms that may be responsible for the survival of the seeds in the soil because they would allow germination only when conditions could be favourable for establishment of a new plant (Ghersa & Roush 1993). Once having reached a low dormancy level, most weed seed populations require termination of dormancy through light (Górski 1975; Fenner 1980; Silvertown 1980; Scopel, Ballaré & Sánchez 1991; Batlla, Kruk & Benech-Arnold 2000) and/or fluctuating temperatures (Thompson, Grime & Mason 1977; Benech-Arnold et al. 1988, et al. 1990; Ghersa, Benech-Arnold & Martinez-Ghersa 1992; Bewley & Black 1994). Light and fluctuating temperature requirements have been linked with the possibility of detecting canopy gaps as well as depth of burial (Holmes & Smith 1977; Thompson & Grime 1983; Benech-Arnold et al. 1988; Ghersa, Benech-Arnold & Martinez-Ghersa 1992; Deregibus et al. 1994; Batlla, Kruk & Benech-Arnold 2000). Fluctuating temperature requirement has been regarded as an effective mechanism for spreading germination over a long period of time (Benech-Arnold et al. 1990) or for forming an ‘emergence window’ (Bouwmeester & Karssen 1993; Vleeshouwers & Bouwmeester 2001).

Seeds perceive the light environment through photoreceptors, especially the phytochrome family (Bliss & Smith 1985; Hartmann & Nezadal 1990; Casal, Sánchez & Botto 1998). Phytochromes have two photo-interconvertible forms: Pfr (usually the active far-red absorbing form), with maximum absorption at 735 nm, and Pr (the inactive red-absorbing form), with maximum absorption at 665 nm (Borthwick et al. 1954; Taylorson & Borthwick 1969; Mancinelli 1994). Exposure of seeds to light with a high red (R) to far-red (FR) ratio (R : FR) leads to the formation of Pfr : P ratios that might be high enough to trigger germination, depending on the dormancy level in the seed population. Phytochrome-mediated responses can be classified physiologically into three classes (Kronenberg & Kendrick 1986). Two of these classes, the very low fluence responses (VLFR) and the low fluence responses (LFR), are characterized by the correlation between the physiological effect and the level of Pfr formed by the light exposure. They differ in that an extremely low level of Pfr saturates VLFR, while higher Pfr levels are needed to bring about LFR (Mancinelli 1994; Casal, Sánchez & Botto 1998). High irradiance responses (HIR) are the third class, showing no simple relationship between Pfr levels, and may involve additional components of the phytochrome system (Heim & Schäfer 1982, 1984; Mancinelli 1994). The HIR of germination has maximum activity at 710–720 nm (Hartmann 1966; Hendricks, Toole & Borthwick 1968; Mohr 1972) and the inhibitory effect of continuous FR can be secured for R-promoted seeds, even after lapse of the escape time (Mohr & Appuhn 1963; Hartmann 1966; Frankland & Taylorson 1983).

The presence of a crop canopy reduces the photon flux of all wavelengths relative to full daylight, much more in the photosynthetically active part of the spectrum (400–700 nm) than in the near infra-red (700–1000 nm), because of strong absorption by chlorophylls (Taylorson & Borthwick 1969; Smith 1982; Pons 1992). Hence canopy shade light is rich in FR and poor in R. For example, unfiltered daylight has an R : FR ratio of approximately 1·2 and leaf canopies may reduce this value to 0·2, depending on the leaf area index (LAI; Federer & Tanner 1966; Holmes & Smith 1977; Pons 1983). Moreover, the presence of a crop canopy reduces the thermal amplitude of the soil, thus preventing germination of species that require fluctuating temperatures to terminate dormancy (Thompson & Grime 1983; Benech-Arnold et al. 1990). Although it is well known that canopy presence interferes with seed germination through modifications of the above-mentioned environmental cues, the issue has only been investigated through static studies (i.e. the effect of an established canopy). The changing effect of a growing canopy has not been assessed previously, thus precluding the identification of shading-intensity thresholds for triggering the different types of canopy-detection mechanisms in different seed populations.

Knowledge of the regulatory effects of the crop canopy on weed seed germination is necessary to understand fully the behaviour of weed seed banks during a crop cycle. In this context the following questions arise. (i) Do seeds from the different predominant weed species require light and/or fluctuating temperatures to have their dormancy terminated? (ii) With what LAI is a crop able to diminish light quality and/or thermal amplitude to prevent the germination of seeds requiring these factors for dormancy termination? Location of the weed seeds in the soil, among other things, will determine changes in the factors that terminate dormancy. For example, modifications in the light environment would only be perceived by seeds that are located at the soil surface or very close to it (Tester & Morris 1987; Kasperbauer & Hunt 1988), whereas variations in the thermal amplitude must be registered independently of the position that seeds occupy in the soil profile (Ghersa, Benech-Arnold & Martinez-Ghersa 1992). In this work we tried to clarify whether the germination–emergence pattern of weed species, accompanying a wheat canopy, can be modified by alterations in the light and thermal environments within this crop. For this purpose experiments under field conditions were carried out, to modify artificially the thermal and light environments under a wheat canopy, until light and thermal conditions prevailing at bare soil (i.e. soil without vegetation) were matched. It was also intended to elucidate the mechanisms through which germination of the different species is regulated by the presence of the crop.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

design and location of the experiments

Two experiments were carried out under different conditions. In experiment 1 the behaviour of a natural soil seed bank, existing in a wheat field located in the Instituto Nacional de Tecnologia Agropecuaria (INTA) Balcarce Experimental Station (37°45′S, 58°15′W), was studied. The soil was subjected to tillage before sowing, and therefore the seeds were distributed in the arable soil profile when the experiment was started. The wheat was sown to establish a density of 350 plants m−2 at the beginning of September 1995. Experiment 2 was conducted the following year, with the objective of investigating the behaviour of an artificial seed bank, composed of seeds from the most frequent weed species accompanying wheat crops in the area where experiment 1 was performed. In this case, the experiment was carried out in the experimental field of the Facultad de Agronomía (Universidad de Buenos Aires, Buenos Aires, Argentina; 34°25′S, 58°25′W). In contrast to experiment 1 all seeds were located at the soil surface. Therefore, wheat plots (0·8 × 0·45 m) with three inter-rows were sown, establishing a density of 350 plants m−2, with a distance of 15 cm between north–south orientated rows. Prior to sowing of wheat (28 August 1996), the upper 5 cm of the inter-row soil was replaced with soil that had been oven-sterilized at 90 °C for 7 days to kill containing seeds. At crop emergence (visible first leaf tip), 100 seeds of each species (Amaranthus quitensis H.B.K., Carduus acanthoides L., Galinsoga parviflora Cav., Portulaca oleracea L. and Raphanus sp.) were randomly distributed on the soil surface of each inter-row (300 seeds plot−1). The experimental design and procedure, as described in the following subsection, was the same for both experiments.

treatments

Because of soil heterogeneity a randomized complete block design with four replications was selected for statistical design. Within each block the following treatments were allocated: (i) thermal and light environments below the establishing wheat canopy (W); (ii) bare soil (B); (iii) modification of the light environment (i.e. R : FR ratio) below a wheat canopy by eliminating FR with CuSO4-filled filters (BLWT); (iv) modification of the thermal environment below a wheat canopy by means of heating racks to match soil temperature of bare soil (WLBT); (v) modification of the light and thermal environments below the establishing crop canopy (BLBT); and (vi) as treatment (i) but with filters filled with distilled water and units buried in the soil to mimic the presence of the heating system that is described below (WLWT, control treatment).

Modification of the soil temperature below the crop canopy

To evaluate the modification of the soil thermal amplitude responsible for a possible reduction in weed germination below the crop, a soil-heating system was designed. This system produced a soil thermal regime beneath the wheat canopy resembling that experienced by the weed seeds buried at bare soil. The system consisted of three alveolar polycarbonate racks (0·8 × 0·12 m) with an electrical resistance inside. The resistances were turned on as soon as the soil temperature beneath the canopy was lower than that registered at bare soil. Automatic turn on was regulated by a thermostat connected to the resistance net and to a temperature sensor located at the bare soil. Before wheat sowing, alveolar polycarbonate racks with electrical resistances inside were buried 5 cm deep in each plot (except the bare soil plot) and either the natural seed banks (experiment 1) or the sterilized soil (experiment 2) were placed back over the polycarbonate racks. To avoid the formation of temperature gradients in the experimental plots the polycarbonate racks were buried in horizontal form. Wheat seeds were sown in the 0·02-m gap between the polycarbonate sections. Empty polycarbonate units were buried in the control treatment (vi).

Modification of the light environment below the crop canopy

We wanted to clarify to what extent a differential modification of the R : FR ratio of the light reaching the soil surface below the wheat canopy was responsible for a possible reduction of weed germination. Modification of the R : FR ratio under the established canopy was carried out by filtering light through solutions of CuSO4 contained in alveolar polycarbonate racks. CuSO4 is known to absorb a great fraction of the FR radiation without affecting photosynthetically active radiation (PAR; Casal & Sánchez 1994). The CuSO4 solution was calibrated to obtain a 1·2 R : FR ratio at the soil level of a crop canopy with a LAI around 4, using concentration and procedures as described by Ballaré, Scopel & Sánchez (1991). Polycarbonate racks, containing distilled water instead of CuSO4, where used in the case of the control treatment (vi).

determinations

In both experiments the following determinations were carried out in all treatments during the wheat crop cycle. (i) Weed seedling emergence was recorded every 20 days, and then the weeds were removed. (ii) Soil temperature was measured at 1-cm depth using sensors connected to a data-logger (LI-COR 1015 and LI-COR 1000; LI-COR Inc., Lincoln, NE). (iii) R : FR ratio at the soil surface was measured at midday with a Skye sensor (SKR 100; Skye, Llandrindod, Wells, UK). (iv) In experiment 2, non-destructive measurements of the crop LAI were carried out with a LI-COR LI2000 Canopy Analyser (LI-COR Inc.).

Water stress was prevented by means of irrigation, with soil water content maintained at near field capacity throughout the experiment.

statistical analysis

Two-way anova was performed using GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego, CA; http://www.graphpad.com; accessed April 2003).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

modifications in the soil thermal environment by a growing wheat canopy

In both experiments, and during most of the growing cycle, the soil thermal amplitude at bare soil was higher than that registered under a wheat canopy (i.e. maximum temperatures occurring at bare soil were higher than those under the crop canopy; minimum temperatures were similar for both situations). As expected from other results (Huarte & Benech-Arnold 2003), the soil heating system produced a thermal environment for the soil beneath the canopy that was similar to that registered at bare soil (Fig. 1c). Soil temperatures in the control treatment (i.e. racks without resistances buried in the soil beneath the crop canopy) were similar to those registered in the soil under the canopy (Fig. 1c).

image

Figure 1. (a) Development of the LAI within a wheat canopy; (b) measured R : FR ratios below the wheat canopy (open circles), at bare soil (closed squares) and below the wheat canopy with light filtering (closed circles); (c) maximum (upper line) and minimum (lower line) soil temperatures (°C) below the wheat canopy (open circles), at bare soil (closed squares) and below the wheat canopy with a soil heating system (closed circles). Data for experiment 2 as a function of days after crop emergence on 9 September 1996.

Download figure to PowerPoint

modification of the light environment under a growing wheat canopy

Results from experiment 2 only are shown, because the proximity to the experimental site allowed more frequent measurements. In any case, there was no reason to assume that the experimental system used to modify the light environment behaved differently in experiment 1. Crop LAI and R : FR ratios measured at the soil surface after wheat emergence are shown in Fig. 1a,b. LAI of the crop, measured 3 weeks after crop emergence, was close to 1 (Fig. 1a). This LAI caused the R : FR ratios measured in the inter-rows to be lower than those registered at bare soil (0·9–1·0 and 1·2, respectively). With crop development the LAI increased to a value of 4 and R : FR ratios fell to a value of 0·4 (Fig. 1a,b). In contrast, in the inter-rows subjected to the effect of the FR filtering system (i.e. polycarbonate racks filled with CuSO4), R : FR ratios increased to 3–4 while the LAI remained low. However, as the LAI increased, the R : FR ratio measured below the filters gradually decreased but never became lower than 1·2, which was the value measured at bare soil during the whole crop cycle (Fig. 1b). R : FR ratios higher than 1·2, equivalent to those measured under the filters at the beginning of the crop cycle (i.e. low LAI), should not have modified seed behaviour because the amount of Pfr produced by the perception of a R : FR ratio equal to 1·2 is known not to increase further with exposure to R : FR ratios higher than 1·2 (Smith 1982). Indeed, if the R : FR ratio > 1·5, the phytochrome photoequilibrium is maintained nearly constant with a Pfr : Ptotal typical of sunlight (Smith 1982). These results show that the experimental system used behaved satisfactorily and was able to reproduce the light environment typical of bare soil below the wheat canopy.

germination of a natural weed seed bank during the wheat crop cycle (experiment 1)

During this experiment six weed species were identified: Amaranthus quitensis, Anagallis arvensis L. var. arvensis, Chenopodium album L., Portulaca oleracea, Galinsoga parviflora and Polygonum aviculare L. Figure 2 shows the emergence of seedlings for each species at bare soil and under the wheat canopy, dependant on days after crop emergence. Because of the small seedling emergence of some species the ordinates of the graphics were adjusted to the maximum number of seedlings emerged for each species.

image

Figure 2. Emerged weed seedlings per square meter in experiment 1, as a function of time after crop emergence at bare soil (closed squares) and under the wheat canopy (open circles) for the natural seed bank of (a) Amaranthus quitensis, (b) Anagallis arvensis var. arvensis, (c) Chenopodium album, (d) Portulaca oleracea, (e) Polygonum aviculare and (f) Galinsoga parviflora. Vertical bars give SEM. Date of crop emergence was 20 September 1995.

Download figure to PowerPoint

In general, the highest emergence of seedlings occurred during the first 30 days after crop emergence (Fig. 2). The presence of the crop canopy had no regulatory effect on the emergence of the weed species because seedling emergence was similar for both treatments (Fig. 2). The only exception was Galinsoga parviflora, the emergence of which continued beyond the first 30 days after crop emergence; after this point it appeared that the wheat canopy was able to exert some regulatory effect on the emergence of this weed (Fig. 2f). These results suggest that the environmental modification produced by a wheat crop during the first 30 days after emergence is not sufficiently strong to modify the emergence pattern of most of the weeds typically observed at bare soil. This could be the reason why the different treatments (i.e. soil heating and FR filtering) did not produce any modification of the emergence patterns. Consequently, the results for these treatments collected during the first 30 days after crop emergence are not shown.

Galinsoga parviflora was the only species the emergence pattern of which appeared to be modified by the presence of the wheat canopy, although these differences, as mentioned above, were only observed towards the end of the crop cycle (e.g. beyond c. 90 days after crop emergence; Fig. 2f). Moreover, Galinsoga parviflora was the only species that produced two different generations during the crop cycle. The seeds that gave origin to the second generation were dispersed about 90 days after crop emergence, and the germination of the seeds from this second generation appeared to be regulated by the presence of the crop. Therefore, the emergence of this second generation was analysed separately from its first generation and from the other species (Fig. 3). It should be noted that seeds from the second generation of Galinsoga parviflora were located mainly at the soil surface, because no soil perturbation took place after its dispersal and before the end of the crop cycle. In contrast, seeds from the first generation of Galinsoga parviflora and of the other species had been buried in the arable soil profile during seed bed preparation. A high number of Galinsoga parviflora seedlings emerged at bare soil while the crop significantly reduced the final number of emerged seedlings from this second generation (Fig. 3). However, modifying both the light and the thermal environments beneath the wheat canopy to levels close to those observed at bare soil eliminated the inhibition of the germination exerted by the presence of the crop (Fig. 3). Modification of the light environment alone was also able to overcome the inhibition of germination, but to a lesser extent (Fig. 3). In contrast, soil heating by itself did not improve the germination observed under the wheat canopy (Fig. 3). The results of this experiment show that a wheat canopy can modify the light and the thermal environments sufficiently to interfere with the germination of seeds from the second generation of Galinsoga parviflora via dormancy termination.

image

Figure 3. Cumulative number of emerged seedlings per m2 of the second generation of Galinsoga parviflora, as developed from seeds dispersed by plants grown together with the wheat within the experimental unit of experiment 1 under differing light and thermal environments: W, thermal and light environments imposed by the establishing wheat canopy; B, bare soil; BLWT, modified light environment (i.e. R : FR ratio) beneath a wheat canopy by eliminating FR with CuSO4-filled filters; WLBT, modified thermal environment beneath a wheat canopy by means of heating racks until soil temperature matches that of bare soil; BLBT, modification of the light and thermal environments beneath the establishing crop canopy; WLWT, below the establishing wheat canopy but with filters filled with distilled water and racks buried in the soil to mimic the presence of the heating system. Vertical bars give SEM.

Download figure to PowerPoint

germination of an artificial weed seed bank during a wheat crop cycle (experiment 2)

In this experiment weed seeds were dispersed on the soil surface to detect modifications in the light environment below a canopy.

The wheat crop was able to reduce the germination of Raphanus sp. (P < 0·07; Fig. 4a), Portulaca oleracea (P < 0·01; Fig. 4c), Amaranthus quitensis (P < 0·09; Fig. 4g) and Carduus acanthoides (P < 0·04; Fig. 4i) in relation to that observed at bare soil. In agreement with the results obtained from experiment 1, the germination of all weed species in all treatments had almost finished 30 days after the crop emergence, except for Portulaca oleracea (Fig. 4c) and Amaranthus quitensis (Fig. 4g), where the germination was inhibited for 30 more days. The inhibition of the germination of the other species by the crop canopy was already accomplished 20 days after crop emergence (Fig. 4). R : FR ratios recorded beneath the canopy at that time were slightly lower (0·9) than those recorded at bare soil (1·15) and the crop LAI was almost 1 (Fig. 1a,b). On the other hand, the soil thermal regime beneath the crop canopy had not changed significantly compared with bare soil (Fig. 1c).

image

Figure 4. Emerged weed seedlings per square meter from an artificial seed bank in experiment 2 as a function of time after crop emergence. Dispersed seeds on the soil surface: (a, b) Raphanus sp., (c, d) Portulaca oleracea, (e, f)* Galinsoga parviflora, (g, h)* Amaranthus quitensis and (i, j) Carduus acanthoides. Left side, seeds exposed to the light and thermal environments below a wheat canopy (open circles) and bare soil (closed squares); right side, seeds exposed to the soil heating system (––) and seeds exposed to soil without a heating system (––; control). Open symbols, light environment below the wheat canopy filtered through water; closed symbols, light environment below the wheat canopy but FR eliminated with CuSO4-filled filters. Vertical bars give SEM. Date of crop emergence was 9 September 1996. *Data partly taken from Batlla, Kruk & Benech-Arnold (2000).

Download figure to PowerPoint

The inhibition of those species, whose final number of emerged seedlings was reduced by the presence of the crop, could be totally or partially eliminated through FR filtering for Raphanus sp. and Portulaca oleracea (Fig. 4b,d). The lack of response to the combination of FR filtering and soil heating might result from some thermal inertia that induced secondary dormancy in seeds of these species. In contrast, the inhibition of germination caused by the presence of the canopy could not be eliminated by any of the applied treatments for Amaranthus quitensis and Carduus acanthoides (Fig. 4h,j).

Interestingly, Galinsoga parviflora was not among the species whose germination was reduced by the crop during the initial stages of its cycle (i.e. 25 days after crop emergence; Fig. 4e,f). However, and in agreement with the results obtained from experiment 1, the plants completed their cycle well before the end of the wheat cycle, and they dispersed seeds that originated a second generation. It was the germination of this second generation that was reduced by the presence of the crop. Indeed, significant differences between the number of seedlings of this species emerged at bare soil and beneath the crop canopy were not detected until 60 days after crop emergence (P < 0·05; Fig. 4e,f). But FR filtering mostly eliminated the inhibition caused by the canopy (Figs 4f and 5). Moreover, the combination of FR filtering and soil heating further increased the number of emerged seedlings, as also observed in experiment 1 (Figs 4f and 5). These results confirmed that the modifications in the light and thermal environments caused by a wheat crop during its initial stages were not able to inhibit the germination of seeds from Galinsoga parviflora, even if they were located at the soil surface. However, during the germination of the second generation of Galinsoga parviflora, a wheat canopy could modify environmental light quality and thermal amplitude to interact with the fulfilment of the requirements of the species to terminate dormancy.

image

Figure 5. Cumulative number of emerged seedlings of the second generation of Galinsoga parviflora, as developed from seeds dispersed by plants grown together with the wheat within the experimental unit of experiment 2, under different light and thermal environments. W, thermal and light environments below the establishing wheat canopy; B, bare soil; BLWT, modified light environment (i.e. R : FR ratio) below a wheat canopy by eliminating FR with CuSO4-filled filters; WLBT, modified thermal environment below a wheat canopy by means of heating racks to match the soil temperature of bare soil; BLBT, modification of the light and thermal environments below the establishing crop canopy; WLWT, control below the establishing wheat canopy with filters filled with distilled water and racks buried in the soil to mimic the presence of the heating system. Vertical bars give SEM.

Download figure to PowerPoint

The emergence phase of the weeds is shown in Fig. 6, referring to the wheat phenology and its light-filtering properties. The emergence of Carduus acanthoides and Raphanus sp. occurred very early within the crop cycle, and a R : FR ratio of 0·9 below the crop was able to inhibit the germination of these species. In contrast, the germination of Portulaca oleracea, Amaranthus quitensis and Galinsoga parviflora (second generation) took place at later stages of the crop cycle, namely 30 days after wheat emergence for the first two species and after crop anthesis for Galinsoga parviflora. This means that the germination was inhibited by R : FR ratios smaller than 0·7.

image

Figure 6. The R : FR ratio as recorded below a growing wheat crop and its inhibitory effect on the emergence of different weed species.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

It is well known that crops accompanying the weed community can modulate environmental factors such as light and fluctuating temperatures that terminate dormancy (Holmes & Smith 1977; Deregibus et al. 1994). These modifications in the seed environment are dynamic during the crop cycle and their influence varies with crop structure and the different phenological stages (Cumming 1963). In weed species requiring environmental signals to terminate dormancy, the strength of the regulation of the emergence by the presence of a crop canopy will depend on the overlap between the ‘emergence window’ of the weed (i.e. the period at which the seed population displays minimum dormancy; Kruk & Benech-Arnold 2000; Vleeshouwers & Kropff 2000) and the density of the crop canopy modifying these signals. The capacity to detect these changes in the thermal and light environments will also depend on the position of the seeds within the soil (Bliss & Smith 1985; Ghersa, Benech-Arnold & Martinez-Ghersa 1992), which will determine the probability of a seed germinating, emerging successfully and its relative time of emergence (Grundy, Mead & Burston 2003).

Two seed locations in the soil were studied in this work: (i) a buried seed bank in experiment 1 and (ii) seeds at the soil surface in experiment 2. In experiment 1 no modifications were observed in the emergence pattern of the different species beneath a crop canopy if related to that observed at bare soil in a buried seed bank (Fig. 2). This could be because during the crop establishment stages (i.e. when most of the weed emergence takes place) the changes produced in the thermal environment were not sufficiently pronounced to modify the emergence pattern of the different species. Galinsoga parviflora presented a particular behaviour: the life cycle of this weed is very short (Rai & Tripathi 1983) and the plants emerging together with the crop were able to disperse seeds when the crop was at anthesis (Fig. 2f). Ivany & Sweet (1973) had reported that freshly dispersed seeds from this species show poor dormancy and germinate in the field during spring and summer, immediately after dispersal. In experiment 1 the seeds dispersed at crop anthesis were located at the soil surface, so they should have been able to perceive the light environment below the crop canopy. The number of seedlings produced by this second generation of seeds at bare soil was higher than that recorded beneath the crop canopy (Fig. 3). Filtering FR in the inter-rows by means of CuSO4 solution produced a similar amount of emergence as that recorded at bare soil, thus eliminating most of the inhibition caused by the canopy (Fig. 3). In contrast, polycarbonate racks containing distilled water did not produce the same effect, thus suggesting that this influence reflects changes in the R : FR ratio of the light that penetrates the crop canopy. In other words, the situation prevailing during the initial crop stages does not allow buried seeds to perceive environmental light changes and the crop canopy does not produce important modifications in the thermal environment. However, canopy structure at anthesis markedly modified the light environment and this could be perceived by seeds located at the soil surface, inhibiting their germination. This conclusion is supported by the fact that Galinsoga parviflora seeds have been reported to respond to light in LFR mode (Batlla, Kruk & Benech-Arnold 2000).

In experiment 2 the seeds were located at the soil surface and had been exposed to the thermal and light modifications caused by the canopy. The germination of most of the studied species was inhibited by the presence of the establishing crop and, interestingly, this influence occurred very early in the crop cycle, with a crop LAI smaller than 1 and an R : FR ratio around 0·9 (Figs 1a, 1b, 4 and 6), which means a moderate modification of the light environment. Germination of Portulaca oleracea and Amaranthus quitensis took place at later stages of the crop cycle, also below the crop canopy (c. 30 days after crop emergence; Fig. 4). At that moment the LAI of the crop was near 2 (Fig. 1a) and the R : FR ratio beneath the canopy was down to 0·4–0·6 (Figs 1b and 6). As concluded from experiment 1, modification of the thermal regime by an incipient canopy is not enough to alter the weed emergence patterns observed at bare soil. On the other hand, R : FR ratios of the light reaching the soil beneath the canopy (c. 0·9–1·0) should be unable to inhibit germination via phytochrome shift towards its inactive form (Pr). For example, Frankland & Poo (1980) have shown that germination of Plantago major seeds is inhibited when the R : FR ratio caused by the canopy of Sinapis alba drops below 0·6, resulting from a LAI near 1·4. Deregibus et al. (1994) found that a R : FR ratio threshold of 0·5–0·8 inhibited the germination of Lolium multiflorum below a canopy. These R : FR ratios require a relatively dense canopy. However, the inhibiting influence of the growing canopy was eliminated for those treatments when the FR was filtered out by CuSO4 solution (Fig. 4), indicating that an increase of the R : FR ratios nullifies the inhibition. This is good evidence that the inhibition of germination by the crop was caused by small modifications of the light quality arriving at the soil surface. Batlla, Kruk & Benech-Arnold (2000) reported that, besides moderate FR enrichment, prolonged exposure is needed to inhibit germination; indeed, exposure to canopy-filtered light pulses of 1 h, presumably saturating the low fluence response of phytochrome (LFR), did not inhibit but rather promoted germination compared with dark controls.

These results, together with the relatively high R : FR ratios measured below the early stages of established canopy, suggest that an HIR might be also be involved here, because the light environment can be perceived by different plant photoreceptors. All cases so far investigated are consistent with the notion that phytochrome A mediates VLFR and HIR while phytochrome B, and in some cases other stable phytochrome(s), mediate LFR (Shinomura et al. 1996; Casal, Sánchez & Botto 1998; Hartmann & Mollwo 2002).

Photocontrol of weeds by soil cultivation in darkness is a preventive weed control method. The idea is to reduce the germination of light-sensitive weed seeds by excluding light during soil disturbance (Hartmann & Nezadal 1990). However, the results of this method are very variable (Hartmann et al. 2003; Juroszek & Gerhards 2004). Another possibility for photocontrol of weed germination is shading of seeds located on the soil surface by vegetation or mulches to avoid their germination (Ghersa et al. 1994).

Understanding the seed responses to modifications in the photothermal environment below the crop canopy (e.g. wheat crop) should allow us to improve weed management strategies in manipulating crop canopy attributes. This could be achieved by modification of sowing date, crop density, spatial arrangement and genotype. For example, increase in the crop plant density would diminish the number of weeds emerging during the first phase of crop establishment. This strategy would be feasible in a situation with weed seeds predominantly located at the soil surface, typically found in a no-till cropping system (Yenish, Doll & Buhler 1992). This soil surface layer, which represents the interface between the seed bank and the environment, is a crucial and variable zone in weed population dynamics (Grundy, Mead & Burston 2003). Furthermore, it is vulnerable to rapid changes in moisture and temperature (Fenner 1985). Ormerod et al. (2003) reported that agricultural production is more complex because of the need to balance global food security, optimum production, technological innovation, preservation of environmental functions and protection of biodiversity. The ecological knowledge achieved in the present study enables considerable compliance with most of those objectives. Furthermore, this information could be incorporated into weed-seed germination models as proposed by Grundy, Mead & Burston (1999).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Silvina Enciso and Jorge Cullen for their contribution in the field experiments. This research was financially supported by Fundación Antorchas (project A-13359/1-000046), SECYT-CONICET (BID 802/OC-AR PIC No. 1/344 and BID 1201/OC-AR PICT No. 08/06651).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References