In Situ Seedling Establishment and Performance of Cyperus esculentus Seedlings

Abstract

Cyperus esculentus seeds are often considered irrelevant for C. esculentus spread as their fragile seedlings would not establish or survive in agricultural soils. However, the ever-increasing spread and upsurge of genetically different clonal populations in NW-Europe raises questions about the establishment of C. esculentus seeds and the reproductive performance of seedlings. Indeed, little is known about the potential of C. esculentus seedlings to grow and propagate under outdoor conditions relative to plants grown from tubers. Seeds from different clonal populations were sown outdoors in various soil types and under different irrigation levels (rainfed, irrigated) to assess seed germination and seedling establishment. Additionally, two pot experiments were conducted with three different plant types (plants originating from mother tubers and from seeds harvested on open- or self-pollinated plants) obtained from eight clonal populations. Plant performance was investigated by measuring vegetative and generative parameters. Germination under outdoor conditions was significantly affected by clonal population and was highest in irrigated sand (5.3%). Germination in sand was 4.1 times higher in irrigated plots than in rainfed plots. In irrigated plots, germination was 3.8 and 4.7 times higher in sand than in sandy loam and clay, respectively. Depending on the year, three out of five to five out of six clonal populations produced more tubers when grown from mother tubers than from seeds. Maximal tuber reproduction factors of 1:965, 1:752, and 1:618 were achieved for plants from mother tubers and seeds from open- and self-pollinated flowers, respectively. Plants originating from open-pollinated seedlings have the potential to equal or exceed the vegetative reproductive capacity of plants originating from mother tubers. As a result of their ability to establish in situ and their substantial vegetative reproductive capacity, C. esculentus seedlings are highly relevant for agriculture and merit appropriate attention in any integrated weed management system targeting C. esculentus.

1. Introduction

Cyperus esculentus L. (yellow nutsedge) belongs to the world’s most troublesome and distributed weeds. It is classified by Holm et al. [1] as the 16th worst weed in the world and causes yield and quality losses and elevated production costs [2]. In Belgium, it was introduced in the early 1980s and has become one of the major weeds in recent years because of its high asexual reproduction capacity and ability to withstand different control measures [3,4]. Its primary mode of dispersal and reproduction is through tubers. The ability of these tubers to withstand low soil temperatures [5] has allowed it to spread in cold temperate regions. Several mechanisms favour the spread of C. esculentus: radial expansion through the formation of rhizomes, within-field dispersal through soil tillage and harvesting operations, and long-distance dispersal through infested soil carried from field to field by farm machinery [6].

In addition to tubers, plants can also produce viable seeds [7]. Although some authors have marginalised the importance of seeds for reproduction [8,9], seeds may play a vital role in the spread of C. esculentus to different continents or regions, even more so when they can survive cold winter months. A single C. esculentus plant can produce thousands of small and light seeds that, with the help of water, animals (e.g., mice, waterfowl), humans, and machinery, can be easily transported over far distances [9,10,11,12]. For example, they may end up in seed packages or dried fodder shipped to different areas [13].

A viable seed set has been reported in several European countries. Schmitt [14] reported germination rates ranging from 5 to 35% in Swiss C. esculentus populations, while, more recently, Keller et al. [15] and De Cauwer et al. [3] found germination rates up to 70% (Swiss populations) and 88% (Belgian populations) based on laboratory top-of-paper germination tests. Dodet [9] found germination rates ranging from 0 to 69% under controlled conditions and between 0 and 8% under outdoor conditions. Despite producing viable seeds, few to no seedlings are found in fields under outdoor conditions [16]. Perhaps partly because they are challenging to distinguish from grass seedlings (Poaceae) at early growth stages or because they establish better under irrigated conditions [17]. Even when seedlings grow under experimental conditions, they seem fragile and have a slow growth rate [15]. However, this does not mean they cannot develop into larger plants. Hill et al. [18] demonstrated that during one growing season under Northern American conditions, a single seedling of 2 cm planted in moist and fertile soil can develop into a stand of plants producing up to 90,000 seeds with a germination rate of up to 51%. Seed germination obtained under (semi-)controlled laboratory conditions is difficult to extrapolate to germination under outdoor conditions where soil moisture content, light conditions, and soil temperatures strongly fluctuate over time.

Cyperus esculentus is generally seen as an obligate outcrosser [2,16]. Nevertheless, more recently, Keller et al. [19] showed that self-pollinated plants can set seeds, albeit with lower germinability than seeds from cross-pollinated plants (up to 16% versus up to 30%, respectively). The seedlings developed from self-pollinated flowers produced less than or as many tubers per ten seeds as those from cross-pollinated flowers (up to 17 tubers per ten seeds) in one growing season. So, C. esculentus could rapidly spread through seeds, provided that the seeds can germinate (depending on germinability and germination requirements), seedlings can establish (depending on fitness, competitiveness, and growth rate), and reproduce (vegetative, generative) under natural conditions.

This paper evaluated the germination and in situ seedling establishment of C. esculentus seeds and aims to determine (1) if seeds from different clonal populations can germinate under Belgian outdoor climate conditions and if this germination depends on soil type and irrigation. Secondly, this paper also evaluated the reproductive performance (growth, development) of seedlings derived from different clonal populations under outdoor growing conditions and aims to (2) determine whether the shoot formation and (a)sexual reproductive capacity of seedlings are similar to that of C. esculentus plants developed from mother tubers. Plants grown from seeds derived from open-pollinated and self-pollinated inflorescences were hereby compared with plants grown from a single mother tuber, each within the same clonal population.

2. Materials and Methods

2.1. Experiments

Clonal populations (tubers), used in all experiments, were collected in 2018 from heavily infested fields (>20% of the area infected) in Belgium and named after the location where they were collected (Figure 1).

2.1.1. Experiment 1: In Situ Seedling Establishment

To investigate the potential of outdoor establishment of C. esculentus seeds, an outdoor field experiment was executed in 2021. It was conducted as a split-split-plot design with irrigation as the main factor, soil type as the subplot factor, and clonal population as the subsubplot factor. The factor irrigation consisted of two levels: rainfed (i.e., only natural rainfall) and irrigated (natural rainfall plus irrigation). Irrigated plots received nine water gifts of 6.66 mm each (irrigation timings are provided in Figure 2B). Three soil types (clay, sand, and sandy loam) and three clonal populations (Aalter, Sinaai-Waas, and Blandain) were evaluated. The experiment consisted of four blocks.

Four plots measuring 250 cm long by 67.5 cm wide and 15 cm deep were dug out, one for each repetition. A bottomless box made from concrete plywood and divided into six subplots was placed in each plot. Each subplot was filled with either clay, sand, or sandy loam according to the split-split-plot design (Figure 3) and subdivided into three subsubplots of 30 by 27 cm each. On 1 July 2021, each subsubplot was sown with seeds (harvested from open-pollinated inflorescences in the preceding year) from one of three clonal populations at a density of approx. 120 germinable seeds. This density corresponded with 363, 190, and 218 seeds of the clonal populations Sinaai-Waas, Aalter, and Blandain, respectively. This was based on their seed germination percentages (33, 63, and 55%, respectively) determined on 4 × 100 seeds using the top-of-paper method on a Copenhagen germination table (‘Arec-cooling technology’, Oudenaarde, Belgium) under a 24/18 °C day/night temperature regime and 16/8 h day/night light cycle. For each clonal population, the seeds were mixed with 1.2 L of the corresponding soil type of the subplot. This mixture was evenly spread on the soil surface to a uniform layer thickness of 12 mm. This was in accordance with Leck [21], who found C. esculentus seeds in soil samples from 0–3 cm depth in wetlands. Plots were kept weed-free by removing the weeds without disrupting the soil.

In experiment 1, emerging seedlings were non-destructively counted weekly during the screening period (i.e., growth period of C. esculentus) from 14 July to 24 September 2021. To avoid double counting, emerged seedlings were marked with coloured ballhead picks once detected and counted. The germination percentage was calculated as the ratio of the number of emerged seedlings divided by the total number of sown seeds.

2.1.2. Experiments 2 and 3: Reproductive Performance of Seedlings

To evaluate the fitness of seedlings, we conducted two outdoor pot experiments with different clonal populations of C. esculentus to include variation between clones. Experiment 2, conducted in 2019, was a complete randomised design with five clonal populations and plants of three different origins, further called plant types. When relevant, ‘seedling type’ is used to refer to both plant types originating from seeds. The clonal populations used were Blandain, Geel, Poppel, Ternat, and Wielsbeke. These clones were chosen based on their ability to form inflorescences and seeds and based on their different genetic backgrounds, as reported by De Ryck et al. [20] (Figure 1). Clonal populations Geel, Poppel, Ternat, and Wielsbeke belong to a different genetic cluster than clonal population Blandain. In contrast with Geel, Poppel, Ternat, and Wielsbeke producing small tubers, Blandain forms large tubers [20]. The three different plant types were (1) plants grown from a mother tuber (MT), (2) plants grown from a seed originating from open-pollinated inflorescences (open pollination, OP), and (3) plants grown from a seed originating from isolated inflorescences (self-pollination, SP). Hereto, 25 pots (5 per clonal population) were grown in a well-ventilated rain shelter greenhouse in 2018. Pots were arranged 30 cm apart in a randomised 5 × 5 square grid to stimulate open pollination. The time window between the appearance (i.e., BBCH stage 51) of the first and last inflorescence was two weeks. Before flowering, one inflorescence per clonal population (from the central zone of the grid) was bagged with a nylon pollination bag (PBS 3d/55, PBS International, Eastfield, United Kingdom). All other inflorescences were left unbagged for open pollination. At full fruit maturity, all bagged (self-pollination) and unbagged inflorescences (open pollination) (from the same pot as the bagged inflorescence) were cut to the ground, dried in the glasshouse for 30 days, manually threshed, and cleaned (see Section 2.2). The tubers were harvested from the same pots used for seed collection. Both seeds and tubers were stored at five degrees until the start of the experiment.

On 23 April 2019, per clonal population, a random selection of seeds (50 per seedling type) and tubers (25 tubers) were placed on a Copenhagen germination table under the same conditions as in experiment 1. Seeds were considered germinated when the radicle penetrated the seed coat. On 14 May, germinated seeds and sprouted tubers were individually transplanted in pots (0.28 L) and placed outdoors in a nursery on a concrete floor. Pots were filled with a mixture of one part steamed sandy loam and one part peat. The steamed sandy loam contained 2.6% organic matter, 46.7% silt (2–50 µm), 43.4% sand (>50 µm), and 10.0% clay with a pH-KCl of 5.5. Irrigation was carried out by overhead sprinklers as needed. On 7 June, all plants were again individually transplanted to larger pots (7.6 L) to allow for unrestricted growth and prevent root-bound plants and fertilised with 200 mL of a 1:100 dilution of liquid fertiliser (3% N, 2% P₂O₅, and 5% K₂O) on 18 July 2019. There were nine repetitions for each combination of clone and seedling type, except for SP plants of Wielsbeke, Poppel, and Geel, for which only four, eight, and eight seedlings could be obtained from seeds of bagged inflorescences due to poor germination levels. For the MT plants, four repetitions were made for each clone. So, in total, there were 103 pots in experiment 2.

Experiment 3 was conducted in 2021 to support the findings of experiment 2 and to test some additional clonal populations. Six clonal populations were selected: Ardooie, Blandain, Dessel, Poppel, Snellegem, and Ternat. Clonal populations Blandain, Poppel, and Ternat were the same as in experiment 2. The experiment consisted of the same three plant types as described above. The material used in experiment 3 was produced in 2020, like the material used in experiment 2. On 26 April 2021, seeds and tubers were germinated under the same conditions as in experiment 2. On 7 May, seedlings and sprouted tubers were individually transplanted in pots (0.28 L) and placed outdoors on a concrete floor. On 27 May, all plants were transplanted into larger pots (7.6 L). The fertiliser was added once on 6 July 2021. Ten repetitions were made for each combination of clone and plant type. The experiment comprised 180 pots (30 per clonal population × 3 plant types × 6 clonal populations). All pots were placed in the same outdoor nursery as in experiment 2.

2.2. Measurements

The nearby meteorological station measured daily global radiation (J cm⁻²), min. and max. daily temperatures (°C), and precipitation (mm) for each experimental year (Figure 2).

The performance of different plant types used in experiments 2 and 3 was investigated by determining four aboveground plant parameters: the number of shoots, inflorescences, seeds, and dry aboveground biomass, along with three belowground parameters: the number of tubers, dry tuber biomass, and dry individual tuber weight. In experiment 3, the germinability of newly produced seeds was also assessed. All measurements were performed per individual pot. Figure 4 shows the growth of a plant originating from a seed and a plant originating from a mother tuber throughout the growing season.

The number of shoots and inflorescences was determined by counting the shoots (>1 cm) and inflorescences visible at the end of the growing season (24 October 2019 for experiment 2 and 3 November 2021 for experiment 3).

To determine the dry aboveground biomass, the foliage was cut 0.5 cm above the ground level at the end of both growing seasons, dried at 75 °C for 16 h, and weighed. Tubers were extracted from the pot substrate to determine the fresh and dry tuber biomass, number of tubers, and individual tuber weight. Hereto, the content of each pot (without foliage) was individually emptied into a rotating metal basket and thoroughly rinsed with tap water until only the rhizomes and tubers remained. After that, tubers were manually separated from rhizomes, counted, and weighed. Afterwards, the tubers were dried in a drying oven for 20 h (16 h at 75 °C followed by four hours at 105 °C) and weighed to determine the dry tuber biomass and dry individual tuber weight.

To determine the number of seeds, the seeds were manually rubbed out of the fully air-dried inflorescences and cleaned with an air blast seed cleaner (Hearson blower, Hearson Laboratory equipment, London) that separates seeds from the chaff by adjusting the airflow rate. Afterwards, the seeds were weighed and counted with a seed counter (Contador, Pfeuffer GmbH, Kitzingen, Germany). The germinability of the cleaned C. esculentus seeds was determined with a top-of-paper germination test. Hereto, 50 seeds per pot were placed on the same Copenhagen germination table under the same conditions described in experiment 1. Germinated seeds were counted and removed. Germination was calculated as the total number of germinated seeds divided by 50.

2.3. Statistical Analysis

All data were analysed in R version 4.2.2 [22]. The statistical analysis of experiment 1 was executed on the germination percentage. The normality and homoscedasticity were checked using a Levene- and Kolmogorov-Smirnov test. The significance of the factors was checked following a split-split-plot analysis as described by Gomez and Gomez [23] with two irrigation levels (main plot), three soil types (subplot factor), and three clonal populations (subsubplot factor). The LSD method was used to determine significant (p < 0.05) differences in germination among groups.

Data of aboveground (number of shoots and inflorescences, dry aboveground biomass, and number of seeds) and belowground parameters (number of tubers, dry tuber biomass, and dry individual tuber weight) measured in experiments 2 and 3 were analysed with two-way Anova’s (five or six clonal populations × three plant types) to check for significances. A significant two-way interaction (p < 0.05) was present between the clonal population and plant type, irrespective of the measured parameter. Therefore, the datasets were split according to the clonal population, and homoscedasticity and normality were checked using the Levene- and Kolmogorov-Smirnov tests. The conditions were met (p > 0.05), and a one-way ANOVA (three plant types) was performed, followed by the LSD method to check for significant differences between means of plant types.

3. Results

The germination percentage of seeds in experiment 1 was significantly (p < 0.05) affected by clonal population. There was also a significant two-way interaction between irrigation and soil type (

Table 1. Anova and interaction between irrigation and soil type with mean and SE for the germination percentage in experiment 1. LSD1 = Least significant (p < 0.05) difference between irrigation levels within each level of soil type. LSD2 = Least significant difference between levels of soil type within each irrigation level. NS = Not significant; * = p < 0.05; ** = p < 0.01; *** = p < 0.001.

			Irrigation		Soil Type		Mean ± SE
Irrigation × Soil type			No		Sand		1.25 ± 0.333
LSD1 Irrigation = 0.991					Sandy loam		0.81 ± 0.224
LSD2 Soil type = 0.866					Clay		0.19 ± 0.093
			Yes		Sand		5.13 ± 1.097
					Sandy loam		1.35 ± 0.227
					Clay		1.08 ± 0.220
Anova	Block	Irrigation	Soil type	Clonal population	Irrigation × Soil type	Irrigation × Clonal population	Soil type × Clonal population	Irrigation × Soil type × Clonal population
	NS	**	***	*	***	NS	NS	NS

). Blandain had a higher germination percentage than Aalter and Sinaai-Waas (2.39 ± 0.68% versus 1.29 ± 0.27% and 1.23 ± 0.35%, respectively). Germination in sand was 4.1 times higher in irrigated plots than in rainfed plots (Table 1). However, in heavier soil types (clay and sandy loam), there was no significant difference in germination between irrigation levels. In irrigated plots, the germination percentage was higher in sand than in sandy loam and clay, showing no significant differences in germination. On average, the germination percentage was 4.7 and 3.8 times higher in sand than in the clay and sandy loam, respectively (Table 1). In rainfed plots, germination was also higher in sand than in sandy loam and clay, but this was only significant for the comparison between sand and clay (1.25 and 0.19%, respectively), not for the comparison between sand and sandy loam (1.25 and 0.82%, respectively) (Table 1). There was no significant difference in germination between clay and sandy loam. The germination percentage was 4, 1.7, and 5.8 times higher when irrigation was added on top of the natural rainfall for sand, sandy loam, and clay, respectively.

Significant differences (p < 0.05) were present between the three plant types evaluated for all measured parameters in experiments 2 and 3, except for dry aboveground biomass in experiment 2 (

Table 2. Mean ± SE for the number of tubers and shoots, dry aboveground biomass, number of inflorescences, and number of seeds for the different clonal populations and plant types evaluated in experiment 2 (2019). The significant differences (p < 0.05, LSD-test) between the plant types within a clone are indicated with a different letter. n = number of repetitions, MT = plant grown from mother tuber, OP = plant grown from seed from open-pollinated flower, SP = plant grown from seed from self-pollinated flower, NS = not significant.

Clonal Population	Plant Type	n	Number of Tubers	Sign.	Number of Shoots	Sign.	Dry Aboveground Biomass (g)	Sign.	Number of Inflorescences	Sign.	Number of Seeds	Sign.
Blandain	MT	4	220.8 ± 7.74	b	26.0 ± 4.24	b	18.4 ± 1.43	NS	6.0 ± 1.08	a	10,455.2 ± 1825.08	a
	OP	9	343.9 ± 25.71	a	47.3 ± 3.62	a	22.7 ± 1.47	NS	0.2 ± 0.15	b	0.0 ± 0.00	b
	SP	9	160.4 ± 22.26	b	31.8 ± 4.41	b	17.7 ± 2.48	NS	0.2 ± 0.22	b	0.0 ± 0.00	b
Poppel	MT	4	954.2 ± 65.76	a	125.2 ± 5.12	a	19.5 ± 2.06	NS	1.8 ± 1.18	a	840.8 ± 493.54	a
	OP	9	614.1 ± 45.27	b	86.6 ± 7.58	b	24.0 ± 3.39	NS	0.0 ± 0.00	b	0.0 ± 0.00	b
	SP	8	435.0 ± 60.37	c	102.2 ± 9.50	ab	18.2 ± 2.32	NS	0.0 ± 0.00	b	0.0 ± 0.00	b
Ternat	MT	4	965.5 ± 112.97	a	78.2 ± 7.66	NS	14.3 ± 1.96	NS	9.0 ± 1.15	a	4644.8 ± 1831.87	a
	OP	9	751.3 ± 50.36	ab	80.2 ± 7.68	NS	21.8 ± 2.24	NS	0.0 ± 0.00	b	0.0 ± 0.00	b
	SP	9	586.9 ± 77.71	b	101.3 ± 16.25	NS	16.3 ± 2.99	NS	0.3 ± 0.17	b	0.0 ± 0.00	b
Geel	MT	4	563.8 ± 59.09	NS	47.8 ± 2.10	b	21.2 ± 0.29	NS	10.0 ± 1.78	a	8374.5 ± 2281.50	a
	OP	9	612.0 ± 42.83	NS	60.6 ± 4.04	a	24.4 ± 2.11	NS	0.0 ± 0.00	b	0.0 ± 0.00	b
	SP	8	476.0 ± 39.52	NS	48.5 ± 3.48	b	24.7 ± 2.12	NS	0.2 ± 0.16	b	0.0 ± 0.00	b
Wielsbeke	MT	4	956.0 ± 97.05	a	72.0 ± 9.42	a	19.8 ± 1.79	NS	1.2 ± 0.48	a	1118.0 ± 663.36	a
	OP	9	570.8 ± 31.93	b	51.4 ± 3.93	b	20.3 ± 0.91	NS	0.0 ± 0.00	b	0.0 ± 0.00	b
	SP	4	374.2 ± 121.37	b	48.5 ± 13.54	b	15.2 ± 4.62	NS	1.0 ± 1.00	a	0.0 ± 0.00	b

and

Table 3. Mean ± SE of ten repetitions for the number of tubers and shoots, dry aboveground biomass, number of inflorescences, and number of seeds for the different clonal populations and plant types evaluated in experiment 3 (2021). The significant differences (p < 0.05, LSD-test) between the plant types within a clone are indicated with a different letter. MT = plant grown from mother tuber, OP = plant grown from seed from open-pollinated flower, SP = plant grown from seed from self-pollinated flower, NS = not significant.

Clonal Population	Plant Type	Number of Tubers	Sign.	Number of Shoots	Sign.	Dry Aboveground Biomass (g)	Sign.	Number of Inflorescences	Sign.	Number of Seeds	Sign.
Blandain	MT	222.6 ± 8.87	b	25.4 ± 2.02	b	26.7 ± 1.35	a	6.9 ± 1.66	a	7538.7 ± 1457.21	a
	OP	473.7 ± 52.72	a	54.4 ± 5.32	a	22.8 ± 1.81	a	0.1 ± 0.10	b	23.0 ± 23.00	b
	SP	190.7 ± 63.75	b	27.3 ± 7.87	b	12.1 ± 2.60	b	0.0 ± 0.00	b	0.0 ± 0.00	b
Poppel	MT	879.3 ± 51.06	a	116.7 ± 10.69	a	36.1 ± 2.44	a	7.5 ± 1.41	a	5104.8 ± 498.15	a
	OP	605.9 ± 40.75	b	85.9 ± 6.26	b	23.7 ± 1.30	b	0.0 ± 0.00	b	0.0 ± 0.00	b
	SP	516.7 ± 83.50	b	90.8 ± 7.92	ab	18.3 ± 2.36	b	0.1 ± 0.10	b	0.0 ± 0.00	b
Ternat	MT	836.5 ± 47.25	a	87.4 ± 5.01	NS	26.6 ± 1.78	a	11.4 ± 1.31	a	15,557.0 ± 1931.43	a
	OP	611.7 ± 60.97	b	71.5 ± 6.62	NS	20.6 ± 1.47	b	0.1 ± 0.10	b	23.2 ± 23.20	b
	SP	567.1 ± 63.41	b	75.2 ± 7.04	NS	14.1 ± 1.53	c	0.0 ± 0.00	b	0.0 ± 0.00	b
Ardooie	MT	697.6 ± 24.10	a	80.1 ± 4.22	a	35.8 ± 1.81	a	19.6 ± 1.74	a	14,095.4 ± 1557.41	a
	OP	540.0 ± 33.51	b	63.5 ± 4.17	b	14.7 ± 1.10	b	0.0 ± 0.00	b	0.0 ± 0.00	b
	SP	485.3 ± 62.44	b	54.2 ± 5.14	b	16.7 ± 0.89	b	0.0 ± 0.00	b	0.0 ± 0.00	b
Dessel	MT	866.5 ± 35.22	a	97.3 ± 2.16	a	46.3 ± 1.48	a	12.8 ± 0.79	a	14,559.5 ± 1101.39	a
	OP	540.1 ± 43.41	b	56.8 ± 4.11	b	21.9 ± 1.70	b	0.1 ± 0.10	b	117.8 ± 117.80	b
	SP	314.2 ± 82.81	c	48.0 ± 9.92	b	34.9 ± 5.67	a	0.2 ± 0.20	b	49.0 ± 49.00	b
Snellegem	MT	780.4 ± 25.86	a	98.8 ± 2.57	a	40.9 ± 1.23	a	14.9 ± 1.55	a	11,975.6 ± 1820.85	a
	OP	553.5 ± 53.14	b	63.5 ± 4.31	b	27.1 ± 1.91	b	0.1 ± 0.10	b	19.1 ± 19.10	b
	SP	618.4 ± 49.03	b	72.6 ± 2.99	b	24.8 ± 1.08	b	0.0 ± 0.00	b	0.0 ± 0.00	b

). Differences mentioned in the results described hereafter are significant unless stated otherwise. In experiment 2, up to 61% fewer tubers were produced by OP and SP plants, except for Blandain, where OP plants produced 36 and 53% more tubers than MT and SP plants, respectively (Table 2). Comparable results were found in experiment 3, where plants originating from seeds produced 23–38% (OP plants) and 21–64% (SP plants) fewer tubers than MT plants, except for Blandain (Table 3). For Blandain, OP plants produced the most tubers, with MT and SP plants producing 53 and 60% less tubers than OP plants, respectively. In experiment 2, SP plants produced 22 (Ternat and Geel) to 53% (Blandain) fewer tubers than OP plants, irrespective of the clonal population (Table 2). This was confirmed in experiment 3, except for Snellegem, for which the number of tubers of SP plants was 12% higher than in OP plants but not significantly different (Table 3).

Ternat, unlike the others, showed no significant differences in the number of shoots between plant types in both experiments. For Blandain and Geel in experiment 2, OP plants produced more shoots than SP plants (33 and 20% more, respectively), while for Poppel, MT plants produced more shoots than OP plants (Table 2). Comparable results were obtained in experiment 3, where the number of shoots was 18–42% and 14–51% lower than the number of shoots produced by MT plants for OP and SP plants, respectively, except for Blandain (Table 3). For Blandain, MT and SP plants produced 53 and 50% fewer shoots than OP plants, respectively.

Differences between plant types were found in experiment 3 for the dry aboveground biomass but not in experiment 2. The dry aboveground biomass of OP and SP plants was 14–59% and 25–55% lower than that of MT plants, irrespective of clonal population. The difference was significant for four of the six clonal populations. For Dessel and Blandain, the dry aboveground biomass of MT plants differed only from that of OP and SP plants, respectively (Table 3).

In experiment 2, MT plants had a higher number of inflorescences (up to 10) than the OP (up to 0.2) and SP plants (up to 1.0), except for Wielsbeke, for which no significant difference was observed between the MT and SP plants (Table 2). Again, in experiment 3, hardly any inflorescences were produced by OP (0 to 0.1) and SP plants (0 to 0.2), in contrast to MT plants (6.9 to 19.6) (Table 3). In contrast with MT plants, which all produced seeds in experiment 2 (841 to 10 455), none of the OP and SP inflorescences produced seeds (Table 2). However, in experiment 3, seed production was observed for the OP (0 to 118) and SP plants (0 to 49). MT plants produced more seeds (5 105 to 15 557) than the OP and SP plants (Table 3).

Dry tuber biomass decreased in the following order: MT > OP > SP for two of five and five of six clonal populations in experiments 2 and 3, respectively (

Table 4. Mean ± SE for the dry tuber biomass and dry individual (ind.) tuber weight for the different clonal populations and plant types evaluated in experiment 2 (2019). The significant differences (p < 0.05, LSD-test) between the plant types within a clone are indicated with a different letter. n = number of repetitions, MT = plant grown from mother tuber, OP = plant grown from seed from open-pollinated flower, SP = plant grown from seed from self-pollinated flower, NS = not significant.

Clonal Population	Plant Type	n	Dry Tuber Biomass (g)	Sign.	Dry ind. Tuber Weight (g)	Sign.
Blandain	MT	4	56.4 ± 4.58	a	0.26 ± 0.023	a
	OP	9	45.6 ± 5.18	a	0.13 ± 0.010	b
	SP	9	17.9 ± 4.84	b	0.10 ± 0.018	c
Poppel	MT	4	81.7 ± 5.58	a	0.09 ± 0.004	NS
	OP	9	47.5 ± 4.06	b	0.08 ± 0.004	NS
	SP	8	30.6 ± 6.46	c	0.06 ± 0.007	NS
Ternat	MT	4	58.1 ± 7.09	a	0.06 ± 0.004	b
	OP	9	61.3 ± 3.38	a	0.08 ± 0.002	a
	SP	9	28.6 ± 5.68	b	0.04 ± 0.007	b
Geel	MT	4	54.1 ± 2.25	a	0.10 ± 0.006	NS
	OP	9	54.9 ± 2.78	a	0.09 ± 0.004	NS
	SP	8	36.2 ± 3.14	b	0.08 ± 0.007	NS
Wielsbeke	MT	4	79.2 ± 5.67	a	0.08 ± 0.003	ab
	OP	9	47.1 ± 1.87	b	0.08 ± 0.004	a
	SP	4	24.0 ± 8.49	c	0.06 ± 0.015	b

and

Table 5. Mean ± SE of ten repetitions for the dry tuber biomass and dry individual (ind.) tuber weight for the different clonal populations and plant types evaluated and of ten or fewer repetitions (n) for the germination percentage (mean ± SE) of seeds harvested from inflorescences from the three plant types in experiment 3 (2021). The significant differences (p < 0.05, LSD-test) between the plant types within a clonal population are indicated with a different letter. MT = plant grown from mother tuber, OP = plant grown from seed from open-pollinated flower, SP = plant grown from seed from self-pollinated flower, NS = not significant. Combinations of clonal populations and plant types without seed production are indicated by ‘/’.

Clonal Population	Plant Type	Dry Tuber Biomass (g)	Sign.	Dry ind. Tuber Weight (g)	Sign.	n	Germination (%)
Blandain	MT	59.4 ± 1.92	a	0.27 ± 0.010	a	9	74.0 ± 2.77
	OP	69.8 ± 6.82	a	0.16 ± 0.020	b	1	38.0 ± 38.00
	SP	19.6 ± 7.22	b	0.09 ± 0.015	c	/
Poppel	MT	103.2 ± 5.97	a	0.12 ± 0.006	a	10	27.4 ± 2.72
	OP	62.5 ± 2.85	b	0.10 ± 0.003	ab	/
	SP	45.3 ± 7.98	b	0.08 ± 0.013	b	/
Ternat	MT	69.6 ± 5.19	a	0.08 ± 0.004	a	10	23.2 ± 2.97
	OP	61.4 ± 5.92	a	0.10 ± 0.010	a	1	10.0 ± 10.00
	SP	29.8 ± 4.33	b	0.05 ± 0.003	b	/
Ardooie	MT	88.6 ± 3.10	a	0.13 ± 0.007	a	10	7.0 ± 1.09
	OP	58.3 ± 3.10	b	0.11 ± 0.008	a	/
	SP	40.1 ± 4.82	c	0.08 ± 0.006	b	/
Dessel	MT	79.8 ± 2.62	a	0.09 ± 0.003	NS	10	26.6 ± 3.06
	OP	58.8 ± 5.26	b	0.12 ± 0.014	NS	1	28.0 ± 28.00
	SP	28.8 ± 8.34	c	0.07 ± 0.011	NS	1	14.0 ± 14.00
Snellegem	MT	106.0 ± 4.17	a	0.14 ± 0.005	NS	10	22.6 ± 3.11
	OP	79.9 ± 3.76	b	0.16 ± 0.016	NS	1	0.0 ± 0.00
	SP	70.4 ± 3.08	b	0.12 ± 0.008	NS	/

). In experiment 2, up to 68% (Blandain) more dry tuber biomass was produced by the MT and OP plants than by the SP plants (Table 4). Also, in experiment 2, OP seedlings produced 34 (Geel) to 61% (Blandain) more dry tuber biomass than SP seedlings, irrespective of the clonal population. In experiment 3, the dry tuber biomass of OP and SP plants was 12−39% and 34−67% lower than MT plants, except for Blandain (Table 5). For Blandain, the dry tuber biomass of OP plants was 15% higher than that of MT plants.

There were no significant differences in the dry weight of individual tubers between plant types for two of five and two of six clonal populations in experiments 2 and 3, respectively (Table 4 and Table 5). In experiment 2, the MT plants of Blandain produced 50% heavier tubers than the OP plants, and these, in turn, produced 23% heavier tubers than the SP plants (Table 4). While the OP plants of Ternat produced heavier tubers than the MT and SP plants (25 and 50% heavier, respectively). In experiment 3, SP plants of Ternat and Ardooie produced lighter tubers than MT and OP plants (Table 5). For Blandain, as in experiment 2, dry individual tuber weight was highest for MT plants (1.7 and 3 times higher than OP and SP plants, respectively).

The seeds harvested from the inflorescences in experiment 3 were able to germinate, apart from the seeds of OP plants from Snellegem, which had a germination of 0%. The seeds produced on MT plants of all six clonal populations varied from 7–74% in germination. For Blandain, Dessel, and Ternat, the germination of seeds from OP plants varied between 10 and 38%. No seeds were obtained from SP plants, except for Dessel, showing seed germination of 14% (Table 5).

4. Discussion

Although the germination percentage was low, seedlings could germinate under Belgian climate conditions. Seedings were detected in every combination of irrigation level, soil type, and clonal population. This provides evidence for successful seed germination and seedling establishment under Belgian outdoor field conditions and has not been reported previously in Belgium. The results align with the results from Switzerland, where germination was also higher at irrigated sites [19]. These results indicate that germination and establishment are possible elsewhere in North-Western Europe, e.g., the Netherlands, Germany, and Northern France, where the climatological conditions are similar [24]. However, this might not be of relevance for all clonal populations as not every clonal population is able to produce flowers and seeds, either because of their genetic background [20] or because the necessary conditions for flower induction are not reached, i.e., intermediate photoperiods between 12–14 h [25,26]. As seed germination is influenced by temperature, as shown by Fort [27], where a 24/18 °C day and night temperature led to significantly more seed germination than a 15/10 °C day and night temperature, it can also be relevant in warmer regions, e.g., Central and Southern Europe. In these warmer regions, C. esculentus might find better growing conditions given its C4 photosynthetic pathway. Germination is also stimulated under elevated soil moisture content, as shown in these results, making it highly relevant for any irrigated cropland infested with C. esculentus seeds. There were significant differences between irrigation levels as well as soil types. However, the differences were not equal for each of the factor levels, as a significant interaction was present between the factor irrigation and soil type. The critical role of water in seed germination has been studied broadly [28]. The results from the in situ germination experiment show that the germination of C. esculentus seeds is positively influenced by the soil moisture content. The germination percentage was highest in the lighter soil type, sand. Compared to the heavier soil types of sandy loam and clay, germination was up to 4.7 and 6.6 times higher in irrigated plots than in rainfed plots, respectively. The higher germination in the coarse-grained sand may be explained by its higher light penetration depth or higher drainage level. Light transmission in the soil increases with increasing particle size indeed [29]. However, Bell et al. [30] stated that full darkness was not inhibitive for germination (46%) under a favourable day/night-time temperature of 35/20 °C. However, in full darkness and under a day/night-time temperature of 30/20 °C, about 8/7 °C warmer than under our soil temperature regime, seed germination was only 1.5%. Bell et al. [30] also found that C. esculentus seeds positioned at 3.8 cm or lower were not able to emerge in pots of sandy loam. However, it was not determined whether the seeds remained dormant or died after germination at lower depths.

Cyperus esculentus has the potential to reproduce from seed and contribute to expanding the tuber stock in the soil. Overall, after approx. 170 growing days in pots, plants from most clonal populations (three out of five and five out of six in experiments 2 and 3, respectively) produced more tubers when grown from mother tubers than from seeds. Across both experiments, the plants grown from mother tubers exceeded a tuber reproduction factor (i.e., number of newly produced tubers by a single mother tuber) of 1:800 (Dessel, Poppel, Ternat, and Wielsbeke) with a maximum of 1:965 (Ternat, experiment 2). Similar results for plants grown from a mother tuber were obtained in pot experiments by De Cauwer et al. [3], who found a tuber reproduction factor of 1:638 (Ternat), and by Bohren and Wirth [31], who reported a tuber reproduction factor of 1:746 for a Swiss clone. The maximum tuber reproduction factor was 1:752 (Ternat, experiment 2) and 1:612 (Ternat, experiment 3) for plants grown from open-pollinated seeds and 1:587 (Ternat, experiment 2) and 1:618 (Wielsbeke, experiment 3) for plants grown from self-pollinated seeds. A distinct response was observed for clonal population Blandain, belonging to a different genetic cluster. Here, plants from open-pollinated seeds produced 1.6 to 2.1 times more tubers than plants from mother tubers in experiments 2 and 3, respectively. Still, their dry individual tuber weight was up to two times lower. Eppler et al. [32] also found that seeds collected from Swiss fields could germinate, develop into vigorous plants, and produce new tubers in the same growing season.

In contrast with plants grown from mother tubers, seedlings produced almost no inflorescences in the year of germination. This could be attributed to the mode of the flower stem genesis; flower stems usually arise from the basal tuber of primary or secondary aerial shoots [8]. For MT plants, the inflorescence number varied between 1.2 (Wielsbeke) and 10 (Geel) in 2019 and between 6.9 (Blandain) and 19.6 (Ardooie) in 2021. De Cauwer et al. [3] tested 25 Belgian C. esculentus clonal populations and found 0 to 13 inflorescences per pot (5.3 L) on plants grown from mother tubers. The three clonal populations that were used in both experiments 2 and 3 produced 1.2 to 4.2 times more inflorescences in 2021 compared to 2019. Although a photoperiod of 12–14 h is generally seen as the most important trigger for the formation of inflorescences [25], the higher rainfall during the experimental period in 2021 (495 mm in experiment 3 vs. 327 mm in experiment 2) may have led to more humid circumstances, which could have stimulated inflorescence formation [33]. The competition between flowering and vegetative growth may have favoured vegetative growth in 2019 more than in 2021 [25]. Although the plants produced inflorescences, seedlings with inflorescences did not set seeds, except in experiment 3, where seedlings produced 8 to 122 times fewer seeds than plants grown from mother tubers. Inflorescences on plants from seeds were formed more than a month later than on plants of mother tubers, leaving less time to form seeds before the senescence of the plants or the first frost. The seed production of plants from mother tubers varied across both experiments between 841 and 15,557 seeds. Similar results for plants grown from mother tuber were found by Eppler et al. [32], who found that seed production was 10 times higher after cross-pollination (about 400 seeds) than after self-pollination (about 48 seeds).

Given that growth rate and biomass provide an indication of the overall performance of a plant [34], the results of the pot trials showed that the performance of seedlings grown from open-pollinated seeds can be as high or higher than plants grown from mother tubers. This indicates that genetic recombination may result in more aggressive C. esculentus clones relative to the parent clones. As the diversity of clonal populations present in and around Belgian fields increases [20], the relevance of seed production and genetic recombination through seeds increases (as open pollination typically results in higher seed production than self-pollination). This may eventually lead to the introduction of more aggressive clones, provided that seedlings can successfully establish themselves in situ [3,4]. Moreover, the high seed production will undoubtedly bring genetic variation that will enable the population to adapt to the pedohydrological, climatic, and cultural conditions prevailing in situ. Seed-setting plants have the potential to easily leave thousands of vital seeds in the soil that can potentially grow into full-fledged plants [3,18]. Even when self-pollinated, vital seeds can be produced, and this can be important for regions where C. esculentus is currently less widespread or abundant or where less genetic variability is present. The extent of seedling establishment in situ depends on several factors affecting in situ emergence, such as seed longevity, germination rate, burial depth, and temperature [17] and seedling survival, such as competitiveness against other weeds and crops [35] and sensitivity towards control methods. Plants grown from seeds formed no or hardly any seeds in the year of germination but produced hundreds of tubers that, if they survive the winter, can grow into plants that are able to reproduce sexually. Hence, quick detection and early eradication before seedlings reach reproductive stages are key. Fortunately, young seedlings are probably more sensitive towards mechanical, thermal, and chemical control methods than plants grown from mother tubers [12,36].

Multiplying the average seed production (11,472 seeds) and seed germination (30%) found in experiment 3 gives an average production of 3457 viable seeds per clone in pots. Combining this with the germination percentage of 1.25% on rainfed sand gives an average of 43 seeds that germinate and can establish under outdoor field conditions. For our experimental area, the 1.25% germination multiplied by the 120 germinable seeds gave rise to 1.5 plants per 0.28 m⁻², or 5.3 plants per square meter. Using an average maximal tuber reproduction factor of 1:615 (experiment 3) for plants originating from OP and SP seeds, these 5.3 plants can then make approx. 3260 new tubers per square meter, adding to the already huge reproduction capacity of C. esculentus plants originating from mother tubers.

To conclude, any seed set on C. esculentus plants should be prevented given the (1) successful establishment under Belgian outdoor conditions, particularly under wet conditions and in lighter soils, the (2) wide genetic variation, and (3) the surprisingly effective vegetative reproduction of seedlings. Most C. esculentus seedlings form fewer shoots, tubers, and inflorescences than plants originating from mother tubers. However, for some clonal populations, the tuber production, tuber biomass, individual tuber weight, shoot production, or dry aboveground biomass of seedlings, particularly the ones originating from open-pollinated seeds, equalled or exceeded those of C. esculentus plants grown from a mother tuber, suggesting that C. esculentus seedlings can also be prolific tuber producers and even constitute more aggressive C. esculentus clonal populations. Hence, when given optimal growing conditions, seedlings are not de facto inferior to plants grown from mother tubers. Given the species’ broad temperature window for seed germination and successful in situ seedling establishment under widely differing pedohydrological conditions, the abovementioned findings apply to more than the Belgian context. Nevertheless, similar experiments could be conducted in other climates and soil conditions and in multiple years to generalise these research findings further. To develop more targeted and effective weed management strategies, future research should also determine the long-term effects of seed propagation within C. esculentus as well as the driving factors of seed germination and seedling growth, e.g., by investigating their physiological and molecular mechanisms under different environmental conditions.