1. Introduction
Teff (
Eragrostis tef (Zucc.) Trotter) is among the oldest cereal crops in the tropical African region and is originated and more diversified in Ethiopia than in any other part of the world [
1]. It is a C
4 plant classified under the family Poaceae, and genus Eragrostis [
1]. It is the major cereal crop in Ethiopia in both production and consumption. It is consumed by more than 60% of the country’s population on a daily basis [
2]. The crop is not only adaptable to a wide diversity of agroclimatic conditions, but also has wide genetic variability encoding for a variety of agronomic traits [
1,
3]. Such a wide adaptation to different agroecologies exposes it to various weed species [
1,
3,
4,
5,
6,
7]. It faces challenges from weeds throughout its growing period that lead to decreased yields [
8,
9,
10]. There are different cultural methods of weed control, most of which are included in agronomic practices with teff [
11]. The most common include frequent tillage before sowing the crop, hand weeding, and to some extent, the use of post- and pre-emergence herbicides. Planting design, which includes planting density, row spacing, and orientation, can also be used as a method of weed management because it can enhance the crop’s competitive advantage over weeds [
12]. The evolution of herbicide-resistant weeds has become a bottleneck and thus calls for alternatives [
13,
14].
The use of competitive varieties is an interesting option for the nonchemical weed control method because it does not add any additional costs and does not create agronomic and environmental concerns. In contrast to the use of weed competitive potential of wheat (
Triticum aestivum L.), barley (
Hordeum vulgare L.), and oat (
Avena sativa L.) varieties [
15], knowledge on the potential use of competitive teff varieties is very limited. Haftamu et al. [
11] studied the competitive ability of teff varieties as well as their traits, e.g., canopy heights and tillering potential of different teff varieties. However, a potential link between the traits of teff varieties and their weed competitive ability was not investigated in that study. There have been a huge number of studies that illustrate the variation in competitive ability between varieties of different cereal crops such as wheat, oat, and barley [
15]. They give many examples of traits, e.g., canopy height, early vigor, and release of different allelochemicals that contribute to increased suppressive ability against weeds.
There are many definitions for the word “competition”, see e.g., in [
16]. In this study, we include a specific kind of competition, form of interaction among plants, called allelopathy, which refers to plant released toxins. Allelopathy is an interference interaction among organisms (specifically plant species) through the release of allelochemicals that hinder the growth and development of another organism (plant species) living near to them [
17,
18]. This study mainly focused on the suppressive effect of one plant species (teff) on other plant species (weeds). Past studies have stressed the impact of allelopathy on various plant species on overall growth and performance of teff [
19,
20]. Phytotoxins are released by most plant species into the soil environment during four important processes—residue decomposition, volatilization, leaching, and root exudates, which may or may not be of benefit to other receptor plants (plants that receive allelochemicals during nutrient and water uptake from the soil) [
18,
21]. The presence of such phytochemicals may inhibit the proliferation of the roots of susceptible receptor plants and interrupts their life cycle through prevention of nutrient and water uptake. Allelochemicals can also hinder weed growth and development during crop production [
18,
22,
23]. Cereal crops such as oat, corn, barley, and wheat have shown suppressive or allelopathic effects on weeds [
22,
23,
24,
25,
26], but many of these have been transient or temporal. However, the use of different crop varieties that interfere with the root growth of weeds through exudation of allelochemicals [
22,
23,
25] shows promising results. Bertholdsson found that early crop biomass and allelopathic activity of barley and wheat were the most important traits significantly contributing to their competitiveness against weeds [
24] and also found that older landraces were most weed suppressive in laboratory and field assays.
Therefore, it is of paramount importance to identify teff varieties with weed suppressive potential through laboratory assays and field experiments and clearly define the most effective traits associated with weed suppression, which can be incorporated into integrated weed management systems.
The objectives of this study were (i) to uncover new knowledge about the allelopathic activity of teff varieties and (ii) to identify the most important agronomic traits of teff contributing to the competitive ability of the crop. We raised the following hypotheses:
4. Discussion
Plants exhibit multiple interactions when growing among each other and such interactions include competition for soil resources such as water, nutrients, solar radiation and space. This competition includes an interference through chemical exudates emitted from plants to the soil. Crop plants and weeds can exhibit such interactions during their life cycle under field or laboratory conditions. Allelopathy is defined as the adverse effect of one plant on another through the production of Phytotoxins or allelochemicals, thereby impacting growth and development of neighboring plants [
30]. Crop species show differences in their responses to weeds.
Many studies have shown that the use of competitive varieties can be an important tool for integrated weed management [
15,
31]. As already mentioned, there is little knowledge on the differences in competitiveness between varieties of teff and which traits of the different varieties might explain the differences in weed competitiveness. Apart from the study by Haftamu et al. [
11], we have not found any studies showing the influence of different varieties of teff on weed growth. In this study, the weed dry-weight ranged from 150.11 g to 356.37 g per m
2 in unweeded plots between the most and the least (plus 137%) competitive teff varieties, representing two fold difference in weed suppression [
11].
Studies have shown that cultivars within a cereal species emanate difference in allelopathic activity and that this trait can contribute to reducing weed interference [
31]. The genotypic effect of a single crop species and its environment has also been shown to affect the release of allelochemicals and their effects over time [
18]. Inhibition of root growth of weeds and other plant species is among the indicators of the allelopathic effects of crop plants [
32,
33,
34]. At this time, we believe there are no research reports on the allelopathic effect of teff on different weed species. Most reports have focused on the allelopathic effect of certain plant species on teff [
19,
20,
35,
36,
37,
38,
39,
40,
41,
42,
43]. For example, leaf extracts from different eucalyptus species and
Parthenium hysterophorus had an effect on the germination and early growth of teff [
19,
20]. In our experiment, where we used a method that did not potentially allow competition for resources between teff and model weeds, we showed that the different teff varieties caused differential root growth of ryegrass and radish. Our first hypothesis, that there are differences in allelopathic activity between teff varieties, was therefore supported. This is consistent with the idea that genotypic variation can impact the allelopathic effect of crop species [
18]. Among the tested teff varieties, the local landrace and the variety DZ-01-2675 were successful in reducing the root length, area, and dry weight of both ryegrass and radish. The PAA and SPAA can be used as parameters to express the allelopathic effect of crops on weeds [
24,
26]. Among all the varieties, the local landrace and DZ-01-2675 had relatively higher allelopathic effects on ryegrass because they had the highest PAA and SPAA values. On the other hand, the variety Boset had the highest allelopathic effect on radish followed by DZ-01-2675 and Kora. These are among the high-yielding teff varieties in Ethiopia [
1,
44]. However, the dominantly produced and high yielding variety widely grown in Ethiopia, DZ-Cr-387 (Quncho) [
45], exhibited the least growth suppression on both ryegrass and radish model weeds.
From the field experiments conducted in Axum and Mekelle in 2015 and 2016, the most important agronomic traits affecting weed growth were days to emergence, days to heading, days to maturity, plant height, tiller number per plant, and crop biomass yield. As already described, the allelopathic potential of the teff varieties obtained from our laboratory experiments was among the important traits of the crop. Most of these traits showed a strong and significant correlation with weed biomass, cover, and density, and this is consistent with many other studies showing that crops with “early vigor”, including early emergence, early flowering, and early maturity, often cause significant growth reductions of weeds [
24,
46]. Furthermore, many other studies have shown that taller plants [
46,
47] with higher tillering potential [
48], higher crop biomass yield, and higher allelopathic effect [
15,
24,
49,
50] have higher competitive ability and cause reasonable reductions in weed biomass, cover, and density. Days to crop emergence is probably the main factor for crop competitive ability and the release of root exudates, i.e., allelochemicals [
24]. Days to emergence was an important trait for explaining differences in weed interference between cultivars in our study. Therefore, our second hypothesis, that emergence and potential allelopathic activity are the two most important traits for teff weed competitive ability, was partly supported. However, not only the early vigor, but also the later growth stages of the teff varieties affected the biomass, cover, and density of weeds during teff production. Teff varieties that exhibit early emergence have the potential to compete with weeds for water and nutrients. Most of the early emerging varieties have early maturing habits and are mostly adapted to low to mid-altitude environments of Ethiopia characterized by high weed intensity [
1,
3,
7,
44,
51]. As already stated above, the PAA of the teff varieties contributed on average from 21.5% to 28.2% of the variance in weed biomass, cover, and density. This suggests that this trait is particularly important in driving weed suppression and it is interesting to note that this mean is comparable to values found in spring barley in which the PAA in the model explained 12%–26% of the variance in weed biomass, although PAA was a less important trait in wheat compared to barley [
24]. Another interesting finding in our study was that the local landrace has a high allelopathic potential. This is similar to the findings on barley observed in experiments conducted in Sweden [
31]. In this study, Bertholdsson concluded that “the allelopathic activity of barley probably originated from different landraces, and in most cases from a specific landrace from the Swedish island of Gotland. We suspect that more than 100 years of selection and breeding have resulted in a dilution of the genes from landraces and consequently a declining allelopathic activity”. Although we believe that the method we have used in the present study, for estimating the allelopathic potential, is proven and reliable [
29], it is also crucial to be aware that this part of the study was performed in laboratory petri dishes and not under field conditions. Such factors as osmotic potential, as discussed by Haugland and Brandsaeter [
29] who used radish and ryegrass as test species, and nutrient or microbial interactions may influence results of laboratory petri dishes experiments.
In this study, teff phenology significantly explained the variance in weed biomass, cover, and density, thus confirming these traits’ contribution to reducing of the overall performance of weeds under field conditions during production. The other agronomic traits, i.e., plant height, tiller no./plant, crop biomass, and allelopathic effect of teff contributed to the competitive ability of the crop as observed during the study period and explained more than 90% of the variance in the weed responses. We found temporal and spatial variability of the allelopathic effect of teff varieties, which might indicate the impacts of environmental factors on the PAA of crops [
18]. Generally, the strong correlation between the weed responses and the agronomic traits of the different teff varieties indicated that these are the important traits explaining the potential competitive ability of teff during production and explaining more than 98% of the variance in weed performance under field conditions.