Canola (Brassica napus), commonly grown for seed oil, shows great potential for use in phytoremediation of selenium-rich soils (Bañuelos et al. 1997a). Canola is considered a secondary accumulator of selenium since this crop plant typically shows selenium concentrations of several hundred mg Se/kg dry weight (DW) when grown in soils with moderate levels of selenium.

Selenium has been shown to be persistent, bio-accumulative, and toxic to most organisms (Brooks 1998; Herda 1999). The effects of selenium toxicity (selenosis) became well known after studies at Kesterson National Wildlife Refuge in California revealed that high concentrations of soluble selenium had been causing deformities and deaths in waterfowl inhabiting the area (Ohlendorf et al. 1986). Ironically, trace amounts of selenium are considered essential for proper growth and development in humans and other animals (Combs 1994). Selenium is also toxic to many plants at high concentrations (Brooks 1998), although several plant species are well known for their ability to tolerate this element at high levels (Johnson and Larson 1999).

Within the United States, soils reported to have high levels of selenium are found in the high plains of Nebraska, the Dakotas, California, and other regions as well (Powers 1996). It is therefore conceivable that many crop plants are in fact grown in selenium-rich soils. The effects that selenium may have on canola, and possibly other crops, are relevant to farmers who may be growing plants in selenium-rich soil and to consumers eating food products with potentially high levels of selenium. Thus, the impacts of dietary selenium, the role of selenium in plant growth, and the use of plants for phytoremediation of selenium-rich soil continue to be active areas of research.

Our understanding of selenium metabolism in higher plants and the use of crop plants for phytoremediation and as a source of dietary selenium have increased dramatically over the past 10 years (Bañuelos et al. 1990, 1992, 1993, 1997a, 1997b, 1998; Terry et al. 2000). While much is known about the interactions between canola and selenium-rich soil, there appears to be no published studies that report the effects of selenium on reproductive events (e.g., seed formation) in canola. The studies presented here aim to increase our understanding of selenium accumulation in canola and the effects that this element has on plant growth and development. In particular, the goals of this study are: 1) to identify which plant tissues accumulate high levels of selenium; 2) to test whether selenium accumulates in canola seeds and oil; 3) to examine the effects of selenium on vegetative growth and development; and 4) to examine the effects of selenium on reproductive events in canola including flowering and seed production.

Seeds from the control and selenium-treated plants were also collected for oil extractions. Oil was extracted by crushing 25 seeds in 10 mL hexane. After thorough grinding, the hexane-oil mixture was decanted and placed in a scintillation vial. The vials were left open overnight to allow the hexane to evaporate. One oil sample each from the control and selenium-treated seeds was analyzed for selenium quantifications (described below).

An independent laboratory (Dr. J.W. Finley, USDA Human Nutrition Research Center, ND, USA) quantified selenium concentrations in various plant organs by atomic absorption spectrometry as described in Finley et al. (1996). A single sample of oil obtained from plants in both the control and the selenium-treated groups was analyzed directly for selenium quantifications. For all other tissues, plants were dried to constant weight as described above. Three replicate dry weight samples of each tissue were then processed for selenium quantification.

Statistical analyses were conducted using the General Linear Model of the Statistical Package for the Social Sciences. Time course study data were analyzed using repeated measures analysis. Differences between one-time averages, for example between number of seeds formed per fruit in the control and selenium-treated groups, were tested using the Student’s t-test.

Table 1 . Mean Selenium Concentration in Various Tissues* *Mean concentration (standard error) of three samples of hydroponically grown canola plants †A single oil sample for each of the two treatments was analyzed directly

Figure 1 . Effect of selenium on plant height in canola. The plants were grown hydroponically without any added selenium (control) or with 2 ppm selenium added in the form of sodium selenate. Data points are means ± standard errors.

Leaf Formation — Plants in the experimental group produced significantly fewer leaves than those grown without any added selenium (P < 0.015, Figure 2). This difference was not evident until plants in the control group started flowering at high rates. In general, there were no visual differences in leaf morphology between the control and selenium-treated plants. However, there was a difference in the leaves produced during the reproductive growth phase and the vegetative growth phase. Vegetative growth phase was period of time growth occurred prior to flowering. Specifically, the leaves produced during the onset of flowering were much smaller than those formed during the vegetative growth phase. Although we did not measure surface area, it is possible that the formation of the numerous, small leaves during the reproductive phase may not be associated with a significant increase in the amount of total leaf surface area.

Figure 2. Effect of selenium on leaf production in canola.

Transpiration of Nutrient Solution — During the first 45 days of growth, there was virtually no difference in transpiration rates between the control and experimental plants (Figure 3). However, after that time the selenium-treated plants displayed significantly higher transpiration rates than did controls (P < 0.004, Figure 3). The lower transpiration rate in the control group, compared to that of the selenium-treated plants, was correlated with the time at which the control plants began to flower and set seed.

Figure 3. Effect of selenium on transpiration rates in canola.

Flowering — Plants in the experimental group displayed significantly slower flowering rates than those grown without any added selenium (Figure 4). There was approximately a 45-day lag time before even one-fourth of the selenium treated plants began to flower. By this same time all of the plants in the control group had flowered. Even after 108 days of growth, only 12 of the 24 selenium-treated plants showed evidence of flowering. Thus, selenium caused a significant delay in the onset of flowering.

Figure 4. Effect of selenium on flowering in canola.

Fruit and Seed Set — As expected, based on the flowering data in Figure 4, experimental plants displayed reduced fruiting rates compared to those in the control group (Figure 5). Within 45 days after formation of the first fruit, nearly all of the control plants showed evidence of fruit set. However, less than one-fourth of the selenium-treated plants had fruit by this same time.

Figure 5. Effect of selenium on fruiting in canola.

Most of the selenium-treated plants never did produce fruits; those that did formed more fruits per plant than plants in the control group (Figure 6). However, the relatively high number of fruits formed on the selenium-treated plants did not compensate for the fact that so few plants ultimately set seed.

Figure 6. Effect of selenium on fruit set in canola. This graph includes only plants that had seed pods.

There was no significant difference in number of seeds per fruit formed in the control and selenium-treated groups (Figure 7). By the end of the experiment, the control group produced 7.33 ± 0.5 seeds per fruit while the selenium-treated group produced 7.27 ±1.36 (mean ± standard error). Pollination was not controlled in any way and it is assumed that both the control and selenium-treated plants had equal pollen loads.

Figure 7. Effect of selenium on seed production in canola.

Germination Rates — It was found that the seeds from the plants in the experimental group had significantly lower germination rates than seeds from the control group (Figure 8). Within two weeks of planting, all the seeds from the control plants had germinated and formed normal, healthy-looking seedlings. However, only 60% of the seeds from the selenium-treated plants germinated. This maximum germination rate for both groups of seeds was achieved approximately 12 days after planting.

Figure 8. Effect of selenium on seed germination in canola. Seeds were subsequently collected and tested for germinability. N=100 seeds for each of the two treatments.

Biomass — On a per-plant basis, the DW of leaves, fruits, and seeds were not significantly different between the selenium-treated and control plants (Table 2). The leaf DW was significantly higher in the selenium-treated plants than in the control group. The fruit and seed DWs were not significantly different between the two treatments, but there was a trend toward higher DWs in the control group. There was no significant difference in the amount of stem or root tissue produced in the control and selenium-treated plants.

Table 2. Dry weights of canola plants grown in differential selenium media* *All values are mean ± standard error. All samples were dried to constant weight over the course of four days.

Selenium in Canola — Our results reveal that leaves contained the highest selenium concentration of all tissues assayed (1565.5 mg Se/kg DW, Table 1). Comparatively, the selenium level in seed pods was only 63% that in the leaves, while the seeds contained 31%, roots 30%, stem tissue 19%, and oil 3%. In general, these results were expected and are consistent with previous studies indicating high selenium levels in leaf tissue. Bañuelos et al. (2000) grew canola plants in selenium-laden soil and subsequently harvested them for analysis. When comparing selenium levels in leaf and stem tissue, they found the selenium levels of stems were only 37% of that in leaves. Similar patterns were observed in a field study with high selenium Turlock soil (Bañuelos et al. 1998). The same group in 1998 reported selenium concentration in roots of canola plants was only about 27% of that found in leaves while the stem concentration was only about 18% of the leaf levels. Similar results have been shown for other Brassica species as well. For example, shoots of B. juncea contained 1092 mg Se/kg DW while roots of those same plants had only 287 mg Se/kg DW after growing in selenium-enriched water cultures (Bañuelos et al. 1997b).

The relatively high concentrations of selenium in leaves and seed pods are expected when one considers the movement of water-soluble selenium throughout the xylem of a plant and the larger number of vacuoles in the leaves. The leaves and fruits represent the ultimate termination of the vascular tissue and thus the regions where selenium would accumulate the greatest. One might also expect the seeds to contain levels of selenium comparable to the seed pods. However, seeds are not the main transpiring organs of the plant.

Seed oil from selenium-treated plants had higher levels of selenium than seed oil from the control plants. This result is intriguing when one considers the consumption of canola oil by humans. There may be at least the potential for selenium in commercially produced canola oil. Selenium is part of an enzyme called glutathione peroxidase that has been shown to reduce the rates of cancer in humans (Clark et al. 1996). Other researchers (Finley et al. 1996, 1998) are attempting to introduce selenium into human diets through broccoli and wheat grown on high-Se soils.

Canola oil represents another possible crop plant that can be used to add selenium to the human diet at healthy levels. A healthy level of selenium in adult men and woman over the age of 19 is 55 micrograms per day (National Institutes of Health 2003). A study by Clark et al. (1996) showed that the administration of 200 micrograms of selenium per day resulted in lower incidents of certain types of cancer. Significantly exceeding these levels can result in selenium toxicity. Conversely, daily levels below 55 micrograms can result in selenium deficiency symptoms in humans. Within the United States, areas that have selenium-deficient soils are the pacific northwest and the northeast. The southeast has variable levels of selenium in the soil and the middle states contain adequate levels as well as areas where accumulator plants contain greater than 50 ppm on a DW basis (Hoagger Goat Supply, 2003).

It is likely that canola plants grown under natural conditions in high-selenium soil would accumulate less selenium than what was observed in the hydroponic culture system. Nevertheless, we should be aware of the possible presence of selenium in commercially available vegetable oils. The selenium levels in those oils should be monitored to ensure that they fall within a safe range. In addition, vegetable oils might prove to be an additional method of supplementing human diets with levels of selenium that have proven effective at lowering the risk of certain types of cancer. Research on the presence of selenium in vegetable oil and other food products warrants further study.

Effects of Selenium on Vegetative Development — Overall, selenium had clear and distinct effects on the growth and development of canola. Selenium caused plants to grow shorter (Figure 1) and produce fewer leaves than normal (Figure 2). Similar effects on vegetative growth were reported previously by Bañuelos et al., (1997a). They reported delayed emergence and stunted growth of canola and kenaf (Hibiscus cannabinus) when grown in seleniferous Turlock soil. A generalized reduced growth rate may be a general phenomenon of plants grown in high-selenium conditions.

We did not measure total leaf surface area and therefore it is not clear whether there were significantly different surface areas for transpiration between the two treatments. In fact, transpiration rates were comparable between the two treatments for about the first 45 days. After that time, plants in the control group showed distinctly lower transpiration rates than the selenium-treated plants (Figure 3). These results seem contradictory when compared with the number of leaves produced per plant, as the control group had more leaves following the onset of flowering. However, the leaves produced with the onset of flowering were small and probably did not drastically increase the overall surface area available for transpiration. The selenium in the leaves may have also inhibited transpiration by interfering with normal metabolic processes in the plant. Thus, our results reveal that the main effect on the transpiration rate was the stage of development, which is affected by the presence of the selenium.

Except for leaves, selenium had no significant effect on the DWs of all tissues assayed (Table 2). Our data reveal that selenium-treated plants had higher leaf DW than control plants. This was expected since the selenium treated plants had a longer period of time to invest resources in leaf tissue while the control plants were developing fruits and seeds. There is a trend of higher DW for fruits and seeds in the control plants. However, the lack of statistical difference is likely explained by the small sample size for the selenium-treated plants.

The difference in leaf DW is intriguing when one considers that these organs had such high concentrations of selenium. These results suggest that canola plants might be able to tolerate a certain concentration of selenium within their tissues without any observable effects. Selenium concentrations in stems and roots were likely below this threshold. Once a certain threshold is surpassed, as it likely was in the leaves, the plants respond with modified rates of growth and development. These results also indicate that although the plants were stressed, they still produced more biomass per plant than the controls. This would be important if using canola to phytoextract the selenium in the soil as biomass in addition to concentration are the two driving factors in how much selenium can be extracted per growing season.

Effects of Selenium on Reproductive Development — Perhaps the most dramatic effect that selenium had on canola plants was the delay in the onset of flowering (Figure 4). Only six plants in the selenium-treated group had flowered by the time all 24 plants in the control group were flowering. This delayed response was unexpected since no previous reports indicated such a phenomenon. However, the majority of published studies have been done with soil amended with selenium that would have lower bioavailability than the nutrient solution used in this study. The delayed flowering may be a result of the continuous availability of selenium within the liquid growth medium under experimental conditions. It is unlikely that such a phenomenon would be encountered in crops growing in natural settings in selenium-rich soil. However, if grown in extremely high-selenium concentrations such as that seen in the soils of Kesterson reservoir, this may become a consideration.

Although there was a delay in flowering, those plants that ultimately did produce flowers were also able to produce fruits (Figure 5). Interestingly, although the selenium-treated plants took longer to set fruit, those plants actually tended to form more fruits per plant than those of the control group (Figure 6). Increased reproductive efforts, in the form of increased fruit production, have been reported for plants grown under low nitrogen conditions (Davenport and Vorsa, 1999). It is plausible that low nitrogen levels and selenoproteins might both have the same effect on fruit set by effectively reducing the levels of functional proteins. This could be because selenium and sulfur are similar in structure and the plant often takes up selenium in the sulfur uptake pathways. This leads to competition between selenium and sulfur, which may result in inhibition of disulfide bonds in proteins, amino acids such as cysteine, co-enzymes sulfur is a part of, and metallothions. Disruption of metallothions would make the plant unable to deal with high levels of other metals that may be present in the tissue of the plant. Selenium appeared to have no effect on the number of seeds formed per plant and both treatments produced about seven seeds per fruit (Figure 7). Thus, plants in both treatments were able to successfully set seed despite the fact that hand pollinations were not done.

Another drastic effect that selenium appeared to have on the reproductive capacity of canola was the germinability of seeds collected from plants in both treatments (Figure 8). Only 60% of the selenium seeds germinated while 100% of the control seeds were viable. This was correlated with high amount of selenium within the seeds and may be due, in part, to inadequate production of fully functional enzymes needed for normal germination events. Indeed, one of the primary mechanisms of selenium toxicity is the incorporation of selenium cysteine and selenium methionine into proteins (Brown and Shrift, 1981). Germination enzymes such as amylase have been shown to have cystein components (Tiedemann et al. 2001; Taneyama et al. 2001; Asano et al. 1999).

In conclusion, when selenium is at high levels in the solution surrounding the roots it can have negative effects on canola reproduction. Also notable is that selenium caused a higher leaf biomass and thus overall biomass of the canola per seed planted. This could be important for phytoextraction applications. The most significant result obtained from this study is that the selenium can get into the oil of plants grown in high-selenium soil. This has potential positive effects on getting selenium into the diets of humans at moderate levels and thus effectively lowering the rates of certain types of cancers. This finding could potentially also give canola farmers in selenium-rich regions a higher selling price for their “anti-cancer” vegetable oil.