Hydroalcoholic Extracts of Campomanesia lineatifolia R. & P. Seeds Inhibit the Germination of Rumex crispus and Amaranthus hybridus

Abstract

This study explores how hydroalcoholic extracts of Campomanesia lineatifolia R. and P. seeds affect the germination and viability of seeds from two weeds, Rumex crispus and Amaranthus hybridus. The phenolic compounds were initially characterized using ultra-high-performance liquid chromatography. In seeds sown in Petri dishes, four concentrations of hydroalcoholic extracts (0%, 3%, 6%, and 9% w/v) were used for single and continuous applications. The mean germination speed, mean germination time, germination percentage, and seed viability were evaluated. Epigallocatechin gallate, quercetin-3-glucoside, epicatechin gallate, ursolic acid, epigallocatechin, and trans-cinnamic acid were the main compounds discovered in that order. Notably, while the germination percentages of both species were reduced with a single application (38.5% for R. crispus and 52% for A. hybridus), they were reduced by 76.2% and 58.34% with a continuous application, respectively. A delay in germination was also observed, which produced changes in germination speed throughout the experiment. With a 9% application, the number of nonviable seeds of R. crispus increased by 40.53%, whereas in A. hybridus, viability decreased by at least 70.8%. Overall, the phenolic compounds in C. lineatifolia extract are thought to inhibit the germination process of the evaluated species.

1. Introduction

Rumex crispus is an herbaceous species in the Polygonaceae family with a wide distribution. It is also known as a problem weed in grasslands and arable lands around the world, being mainly a colonizer in disturbed areas [1]. Amaranthus hybridus, family Amaranthaceae, is a fast-growing annual herb [2] that reduces crop yields globally owing to its allelopathic activity, competition for light, and being a host for pest insects [3]. In addition, it has a high potential for herbicide resistance, making it one of the primary weed management objectives [4].

Plant interaction is generated by the synthesis of multiple chemical substances known as allelopathic or allelochemical compounds, which are released into the environment and can inhibit or stimulate the germination and development of various weeds and crops, depending on the case [1]. The effects of allele-specific chemicals on other plants are usually inhibitory, depending on the level of phytotoxicity and concentration of the secreted compounds, environmental conditions, and the plant’s susceptibility. Depending on the stage of its life cycle, it can cause germination and late emergence or embryo death [5]. Allelochemical compounds are grouped into three classes—phenols, terpenoids, and nitrogenous compounds [6]—and can be found in leaves, roots, flowers, fruits, and seeds [5]. These compounds can be extracted and used as bioherbicides in agricultural systems, making them a tool for sustainable weed management [7], as they maintain ecosystem balance by reducing the use of chemical herbicides [8,9], to which plants usually develop resistance and which also have negative effects on the environment and human health [10,11,12,13].

Campomanesia lineatifolia R. and P., also known as “chamba” or ”champa”, is a species of an Amazonian fruit tree that belongs to the Myrtaceae family [14]. Its fruits are edible berries with a pleasant flavor, contain six to eight seeds that are approximately 1 cm in diameter and have a high and diverse content of secondary metabolites [15]. Muñoz et al. [14] extracted phenolic compounds with antioxidant activity from this species and concluded that it has potential as a source of bioactive substances. In addition, Bonilla et al. [15] found that C. lineatifolia seeds contain compounds similar to ß-triketones, which are characterized by the presence of several methyl groups, a flavonoid ring, or chalcone, making them a relatively rare class of secondary metabolites with antimicrobial activity [16,17]. Some plants contain biologically active compounds, and their crude, metabolic, and allelopathic extracts inhibit weed seed germination and plant growth, resulting in necrosis or chlorosis [9].

Recently, an inhibitory allelopathic effect of C. lineatifolia extract on the germination and physiology of Taraxacum officinale [18] and on the germination of Sonchus oleraceus in concentrations of 3–9% of the ethanolic extract was found; in addition, with foliar application, an incidence of 100% with symptoms of chlorosis and necrosis was observed [19]. According to these findings, this extract has a high potential for use as a bioherbicide in agriculture. However, many aspects of its mode of action and impact on other weed species remain unknown.

Therefore, this study aimed to characterize the seed extract of Campomanesia lineatifolia and to evaluate its allelopathic activity on the germination of Rumex crispus and Amaranthus hybridus under laboratory conditions.

2. Materials and Methods

2.1. Localization

Champa (Campomanesia lineatifolia R. and P.) seeds were obtained from ripening fruits (Figure 1) grown in the municipality of Miraflores in Boyacá, Colombia, located at 5°11′40″ N and 73°08′44″ W, at 1432 m a.s.l. Plant samples of Rumex crispus and Amaranthus hybridus weeds were collected in the municipality of Paipa, Boyacá, Colombia. Seeds were extracted from the plant samples using a stereoscope, dried, and stored in paper bags for 30 days. Experiments were conducted in the Biology Research Laboratory of El Bosque University (INBIBO, Bogotá, Colombia).

2.2. Hydroalcoholic Extract Preparation for Germination Tests

The dried champa seeds were crushed and equally distributed in 1000 mL glass containers with 96% ethanol and kept in the dark for seven days, constantly stirring. After this time, the heterogeneous solution was filtered with Sartorius Stedim grade 292 filter paper (Göttingen, Germany), and this filtrate was put into a 1000 mL flask inside the Heidolph HS digital rotary evaporator (Schwabach, Germany), where the hydroalcoholic solution was distilled at a temperature range of 58–68 °C and a pressure of 350 mm Hg. The resulting solution was left to dry at room temperature in a dark glass container with air inlets to remove excess ethanol and obtain a pasty and homogeneous texture.

Concentrations at 3%, 6%, and 9% necessary for the experiments were made using the ratio in percentage between the weight of the solute and the volume of the solvent, taking 15, 30, and 45 g of the extract and dissolving in 500 mL of distilled water.

Determination of Phenolic Compounds in C. lineatifolia Seeds

The compounds present in the C. lineatifolia seed extract were analyzed using ultra-high-performance liquid chromatography, employing a Dionex Ultimate 3000 chromatograph (Thermo Scientific, Sunnyvale, CA, USA). The LC–MS interface was electrosprayed (ESI), and the mass spectrometer was high resolution with an Obitrap ion current detector. Separation was achieved using a Hypersil GOLD aQ column (Thermo Scientific, Sunnyvale, CA, USA; 100 × 2.1 mm, 1.9 μm particle size) at 30 °C. The mobile phase was A: an aqueous solution (0.2% ammonium formate) and B: acetonitrile (0.2% ammonium formate). The initial gradient condition was 100% A, changing linearly to 100% B (8 min); it was kept for 4 min and returned to initial conditions in 1 min; the total running time was 13 min with 3 min for post-run. The Obitrap mass spectrometer (Exactive Plus, Thermo Scientific, Sunnyvale, CA, USA) was connected by an electrospray interface (HESI) and operated in positive mode with a capillary voltage of 4.5 kV. Nitrogen was used as the drying gas. Mass spectra were acquired in the mass range m/z 60–900. The compounds were identified through the full scan acquisition mode, and ion extraction corresponding to the [M + H]⁺ of the compounds of interest, the measurement of exact masses (accuracy of Δppm < 0.001) and corresponding comparison with standard certified compounds.

2.3. Experimental Design and Treatments

In the germination tests, a completely randomized design was used for each species and experiment. In all cases, four treatments and four repetitions were evaluated, which corresponded to concentrations of 0%, 3%, 6%, and 9% w/v of the hydroalcoholic extract, for a total of 16 experimental units per species and experiment. Each experimental unit comprised a Petri dish with 100 seeds on absorbent paper. For each species, two independent experiments were conducted in which the number of applications to the seeds varied, one with a single application of the extract at the beginning of the experiment, and the other experiment with application twice a week. In both cases, the extract was applied until the absorbent paper was moist. The experiments for the two species were conducted for 34 days.

For the control case, only the absorbent paper was kept moist, making constant applications with distilled water according to the methods of Cabeza–Cepeda et al. [18] and Martínez et al. [19]. The Petri dishes were kept in the laboratory under natural photoperiod conditions of 12–12 and an average temperature of 18 °C, which are favorable conditions for the germination of the evaluated species.

2.4. Germination Variables

Germination data and the time in which this occurred were recorded every two days from the germination of the first seed until constant germination was obtained. According to Niño-Hernández et al. [20], germination was defined as a seed with a visible radicle longer than 2 mm. Then, the germination percentage (GP), mean germination speed (MGS), and mean germination time (MGT) were calculated (

Table 1. Equations used to calculate GP, MGS, and MGT; adapted from [20].

Variable	Equations
Germination percentage GP (%)	$\mathrm{GP}=\frac{\mathrm{No}. \mathrm{germinated} \mathrm{seeds}}{\mathrm{No}. \mathrm{sowed} \mathrm{seeds}}\times 100$
Mean germination speed (germinated seeds/day)	$\mathrm{MGS}=\frac{\mathrm{P}1}{\mathrm{T}1}+\frac{\mathrm{P}2}{\mathrm{T}2}+\frac{\mathrm{Pn}}{\mathrm{Tn}}$
Mean germination time (days)	$\mathrm{MGT}=\left(\frac{\mathrm{x}1}{\mathrm{d}1}\right)+\left(\frac{\mathrm{x}2}{\mathrm{d}2}\right)+\left(\frac{\mathrm{xn}}{\mathrm{dN}}\right)/\mathrm{Xn}$

P: number of germinated seeds, T: germination time, x1 and x2: germinated seeds on days 1 and 2. d1 and d2: days after sowing. Xn: number of germinated seeds until the last day.

).

2.5. Seed Viability (%)

The viability of seeds that did not germinate four weeks after starting the germination experiment with the two weed species was determined using a tetrazolium test. For this, a phosphate buffer solution was prepared by diluting 0.6 g of KH₂PO₄ and 0.7 g of K₂HPO₄ in 100 mL, and 1 g of 2,3,5-triphenyltetrazolim chloride was added. Using a stereoscope, each seed was cut longitudinally with a Minora^® blade and placed in a 60 × 15 mm glass Petri dish to be embedded in the mentioned tetrazolium solution. They were incubated at 35 °C ± 2 °C for 24 h, modifying the method of Paraiso et al. [21]. The seeds were then examined under a stereoscope to determine the degree of staining of the embryo and cotyledons because, according to Elías et al. [22], the families Amaranthaceae and Polygonaceae have a peripheral type embryo that surrounds the internal area of the storage tissue, which comprises dead tissue and is not stained by the tetrazolium solution.

A. hybridus viability was determined using the scales defined by Niño-Hernández et al. [20], whereas to determine R. crispus viability, 10 seeds were placed in imbibition for 8 h, after which five seeds were placed in microwaves for 1 min. Together with the other five, they were subjected to the tetrazolium buffer solution, providing a negative and positive control, respectively, and allowing for appropriate staining of the seeds with the solution.

Three categories of seeds were used: (1) viable, (2) doubtful, and (3) nonviable. This made it possible to determine whether the seed did not germinate because it was in a dormant state or because the embryo was dead.

2.6. Statistical Analysis

A fit was made to a logistic model in the R version 3.6.0 program to determine the behavior of germination (%) as a function of time; the first derivative of the model was considered to calculate the behavior of the germination speed (GS), following the methodology proposed by Carranza et al. [23]. In the SPSS v 19 program, tests for normality (Shapiro–Wilk test) and homogeneity of variances (Levene’s test) were performed, after which ANOVA and Tukey’s multiple-range test (p < 0.05) were conducted. The square root function was used to transform variables that did not meet the assumptions.

3. Results

3.1. Hydroalcoholic Extract of C. lineatifolia

A total of 154.76 g of C. lineatifolia seed extract was obtained, with a yield of 14.83%. The majority of the phenolic compounds (20) discovered were flavonols, flavanones, flavonol glycosides, hydroxycinnamic acids, and hydroxybenzoic acids (

Table 2. Phenolic compounds identified in C. lineatifolia seeds. Rt: retention time. [M + H]⁺: protonated molecular ion peak.

Classification	Compound	Rt, min	[M + H]+	Concentration (mg kg−1)
Methylxanthines	Caffeine	4.12	195.09	4.30
	Theobromine	3.57	181.07	15.60
Flavanol	Catechins	3.96	291.09	5.47
	Epicatechin (EC)	4.17	291.08	17.23
	Epigallocatechin (EGC)	3.80	307.08	53.50
	Epicatechin gallate (ECG)	4.58	443.10	83.30
	Epigallocatechin gallate (EGCG)	4.21	459.09	165.40
	Kaempferol	6.08	287.05	3.20
	Quercetin	5.59	303.05	40.00
Flavanone	Naringenin	5.87	273.07	0.50
	Pinocembrin	6.75	257.08	4.30
Flavone	Apigenin	6.01	271.06	3.10
Flavonol glycosides	Kaempferol-3-glucoside	4.84	449.11	15.50
	Quercetin-3-glucoside	4.66	465.10	138.40
Hydroxycinnamic acids	p-coumaric acid	4.70	165.05	12.50
	Trans-cinnamic acid	5.79	149.06	49.10
Hydroxybenzoic acids	p-hydroxybenzoic acid	4.19	139.04	100.70
	Vanilic acid	4.23	169.05	45.60
Anthocyanins	Cyanidin	4.86	287.05	18.90
Pentacyclic triterpenoid	Ursolic acid	9.79	457.37	76.40

). Each of the compounds, their retention time (Rt), [M + H]⁺, and concentration are presented in Table 2. The main compounds were, in their order, epigallocatechin gallate (EGCG), quercetin-3-glucoside, p-hydroxybenzoic acid, epicatechin gallate (ECG), ursolic acid, epigallocatechin (EGC), and trans-cinnamic acid (Table 2).

3.2. Germination and Viability in Rumex crispus

3.2.1. Behavior of Germination as a Function of Time

The behavior of GP as a function of time was described by the logistic model (

Table 3. Equations of the logistic model for the germination percentage (%) and germination speed (seeds/day) of Rumex crispus. Meaning codes: 0.001 ‘***’ 0.01 ‘**’ RMSE: root mean square error.

Experiment and Treatm Ents	Equations	RMSE
[Extract] (% w/v) on R. crispus with one application
Control	Y = 60.93/1 + e−0.38×(days−9.20)	4.20 ***
3	Y = 40.77/1 + e−0.20×(days−16.29)	2.45 ***
6	Y = 50.91/1 + e−0.56×(days−8.39)	3.07 ***
9	Y = 25.13/1 + e−0.66×(days−17.87)	2.82 **
[Extract] (% w/v) on R. crispus with repeated application
Control	Y = 60.93/1 + e−0.38×(days−9.20)	4.20 ***
3	Y = 35.23/1 + e−0.23×(days−26.57)	0.70 ***
6	Y = 23.48/1 + e−0.23×(days−26.57)	1.33 ***
9	Y = 19.68/1 + e−0.28×(days−29.40)	0.55 ***

, Figure 2). Germination was higher in the control, with a significant increase observed from the first days of sowing until day 28 when it became constant and reached a maximum value of 60.92%. However, for the experiment with a single application (Figure 2A), the values obtained by the 6% concentration were similar to those obtained by the control until day 14, when it reached a maximum value of 50.90% germination, but had higher germination than the 3% concentration throughout the experiment. However, in the experiment with multiple applications (Figure 2C), the three treatments were lower than the control because the 3% concentration with the highest germination was still less than 30%. For both experiments, one application and from several, the highest concentration (9%) had the lowest germination rate, but in the first experiment, it achieved a slight increase to 25.13% and remained constant, for which the seeds stopped germinating; for the case of the second experiment, germination reached its maximum point (15.43%) on day 34 (Figure 2C).

The germination rate in the control began to increase from the first days of sowing, reaching a maximum value of 5.6 seeds/day on day 10 and then decreasing to 0 seeds/day on day 24 (Figure 2B). In the one-application experiment, the 6% concentration had a higher GS, and a later increase was observed compared to the control; however, it reached its maximum point before the control, with 6.8 seeds/day on day 9 and decreasing to 0 seeds/day on day 18. In contrast, the concentration of 3% had the lowest GS, with a maximum of 2.06 seeds/day, but even so, germination occurred throughout the experiment. In addition, a delay in seed germination was observed in the 9% concentration since its germination began to increase on day 14, reaching a maximum value of 4.11 seeds/day on day 18 and tending to 0 on day 25 (Figure 2B).

With repeated application of the extract, the maximum GS was reached by the control, and those of the treatments were very low since the 3% treatment presented its maximum point on day 27 with 2.05 seeds/day, that of 6% with 1.36 on the same day, and 9% with 1.36 on day 30. In general terms, germination was delayed 20 days compared to the control because, at this point, the treatments induced germination, which started in the first days regarding the control (Figure 2D).

3.2.2. Germination Indices

In the one-application experiment, GP showed statistically significant differences between the extracts; a reduction in the germination of 36.5% and 38.5% with extracts of 3% and 9%, respectively, was observed compared to the control (

Table 4. GP, MGS, MGT, and viability of R. crispus with different concentrations of C. lineatifolia extract. Means (± standard error) followed by different letters in each column and experiment indicate significant differences according to Tukey’s test (p < 0.05). GP: Percentage of germination; MGS: Mean germination speed; MGT: Mean germination time. Meaning codes: 0. 05 ‘*’, ns, No statistical differences.

Extract Application	Concentration (w/v)	Germination			Viability (%)
		GP (%)	MGS (Germinated Seeds/Days)	MGT (Days)	Viable	Doubtful	Nonviable
One application	Control	65.0 ± 5.5 a	6.4 ± 0.7 a	10.5 ± 1.1 a	3.5 ± 0.5 b	11.2 ± 4.0 ab	20.3 ± 4.3 b
	3%	41.3 ± 3.1 b	2.7 ± 0.3 b	9.7 ± 1.6 a	10.2 ± 1.4 a	18.5 ± 3.1 a	30.0 ± 2.8 ab
	6%	54.3 ± 2.9 ab	5.3 ± 0.5 a	8.2 ± 0.7 a	9.5 ± 1.9 a	8.0 ± 2.3 b	27.5 ± 2.9 ab
	9%	40.0 ± 8.5 b	1.7 ± 0.3 b	11.8 ± 3.0 a	10.8 ± 3.5 a	17.5 ± 3.8 a	37.5 ± 5.7 a
Significance		*	*	ns	*	*	*
Repeated application	Control	65.0 ± 5.4 a	6.4 ± 0.7 a	10.5 ± 1.1 a	3.5 ± 0.5 b	11.2 ± 4.09 b	20.3 ± 4.3 b
	3%	30.0 ± 2.5 b	1.2 ± 0.2 b	9.9 ± 0.6 a	30.8 ± 2.2 a	14.0 ± 2.3 ab	25.3 ± 4.1 ab
	6%	22.3 ± 4.6 b	0.9 ± 0.2 b	6.9 ± 1.4 a	21.3 ± 5.7 a	20.8 ± 2.4 ab	35.8 ± 3.2 a
	9%	15.5 ± 4.6 b	0.6 ± 0.1 b	5.6 ± 1.8 a	17.0 ± 3.2 ab	31.5 ± 2.8 a	36.0 ± 3.6 a
Significance		*	*	ns	*	*	*

).

In the multiple-application experiment, all treatments showed statistically significant differences when compared to the control, with the 3% concentration having a 53.8% GP reduction and the 9% concentration having a 76.2% GP reduction (Table 4).

The MGS in the one-application experiment was shown to be statistically significant between the treatments of 3% and 9% with values of 2.67 and 1.73 seeds/day compared to the control, which was 6.34 seeds/day, for which they were defined as optimal concentrations to reduce GS, with special emphasis on 9%, which reduced it by 88.5%. With various applications, there were differences between the control and concentrations; a reduction of at least 80.5% in the MGS was observed with the 3% concentration (1.24 ± 0.17) seeds/day (Table 4).

In the one-application experiment, the MGT did not show any significant differences (p < 0.05) between the extracts and the control, nor in repeated application, where the extract had no effect (Table 4).

3.2.3. Seed Viability

Figure 3 shows the three visual scales defined as determinants in R. crispus seed viability: (1) viable seed (Figure 3A), whose staining was a homogeneous vivid red or pink; (2) doubtful (Figure 3B), in which the embryo was stained pale pink or only certain areas of it and finally (3) nonviable (Figure 3C), with no embryo staining.

With one application of the extract, viable, doubtful and non-viable seeds presented significant differences (p < 0.05) between extract concentrations. In viable seeds with any concentration of the extract, a higher percentage was obtained compared to the control. The highest percentage of doubtful seeds was presented with 3% and 9% of the extract. Likewise in nonviable seeds, there were differences (p < 0.05) between the control and the 9% concentration, with 20.3 ± 4.3 and 37.5 ± 5.7 seeds, respectively. In the multiple-application experiment, 3% and 6% concentrations yielded the highest number of viable seeds when compared to the control. In addition, the 9% concentration had the highest number (p < 0.05) of seeds classified as doubtful with 31.5 ± 2.8%. For nonviable seeds, with 6% and 9% of the extract the highest percentage (p < 0.05) was generated (Table 4).

3.3. Amaranthus hybridus

3.3.1. Germination Behavior as a Function of Time

In the one-application experiment, the seed with 6% concentration started its germination after day 10, whereas the seed with the 3% treatment started on day 19; however, by day 34, they converged at the same point with germination close to 49% and 50.5%, respectively. The 9% extract only reached a maximum GP of 31.4% until day 34, whereas in the control, the behavior of germination from the first moment showed a constant increase up to its maximum, which was 60% in both experiments (Figure 4A).

Overall, with repeated application, there was a lower maximum GP because 3% concentration had barely 36.4% and started on day 10, whereas with 6%, it only reached 28% from day 26 and started its germination on day 22, and finally, 9% concentration started germination on day 26 and its maximum was 25.2% (Figure 4C).

Table 5. Equations of the logistic model for the germination percentage (%) and the germination speed (seeds/day) of A. hybridus with different concentrations of C. lineatifolia extract. Meaning codes: ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05. RMSE: root mean square error.

Experiment and Treatments	Equations	RMSE
[Extract] (% w/v) on A. hybridus with one application
0	Y = 301.65/1 + e−0.03×(days−83.23)	2.31 *
3	Y = 51.98/1 + e−0.45×(days−26.23)	1.00 ***
6	Y = 58.61/1 + e−0.24×(days−27.27)	2.92 **
9	Y = 34.88/1 + e−0.39×(days−28.32)	1.28 ***
[Extract] (% w/v) on A. hybridus with repeated application
0	Y = 301.65/1 + e−0.03×(days−83.23)	2.31 *
3	Y = 38.14/1 + e−0.30×(days−23.77)	1.16 ***
6	Y = 28.50/1 + e−3.13×(days−22.71)	4.54 **
9	Y = 27.92/1 + e−0.53×(days−29.79)	0.05 ***

shows the logistic model fit equations.

Regarding GS, with one application, the 6% concentration started to increase this variable, exceeding that of the 3% concentration from day 4 and 18, respectively; however, the 3% concentration reached a maximum GS exceeding all treatments on day 26 with 5.8 seeds/day, but it decreased rapidly (Figure 4B). In contrast, with repeated application, the 6% concentration reached a maximum GS on day 23 with 18.3 seeds/day, but it was drastically reduced, but in the 3% concentration, it began to increase from day 10 to day 24 with a maximum of 2.8 seeds/day. Finally, the concentration of 9% increased its germination on day 23; however, its maximum speed was present until day 30, with 3.4 seeds/day (Figure 4D).

3.3.2. Germination Indices

With the one-application experiment, GP showed significant differences (p < 0.05) between the control and the 9% concentration with 60.5 ± 6.5% and 31.0 ± 1.0%, respectively, indicating that it was the only treatment with which it was sufficient to add one application to reduce GP; however, only a reduction of 48.8% was observed. Regarding the experiment with repeated application, there were differences between the control and the other concentrations, for which, even with the lowest concentration (3%), it was possible to reduce GP up to 42.1%, and with the 9% concentration, there was a reduction of 58.34% (

Table 6. Germination percentage (GP), mean germination speed (MGS), mean germination time (MGT), and viability of A. hybridus seeds with different concentrations of C. lineatifolia seed extract. Means (±standard error) followed by different letters in each column and experiment indicate significant differences according to Tukey’s test (p < 0.05). Meaning codes: 0. 05 ‘*’, ns, No statistical differences.

Extract Application	Concentration (w/v)	Germination			Viability (%)
		GP (%)	MGS (Germinated Seeds/Days)	MGT (Days)	Viable	Doubtful	Nonviable
One application	Control	60.5 ± 6.5 a	9.0 ± 2.8 a	11.2 ± 3.3 b	36.0 ± 5.8 a	1.3 ± 0.5 b	1.8 ± 0.7 c
	3%	50.0 ± 1.8 a	1.8 ± 0.4 b	18.8 ± 0.7 a	10.5 ± 1.9 b	24.8 ± 2.2 a	12.0 ± 3.7 bc
	6%	47.5 ± 3.4 a	2.1 ± 0.4 b	16.7 ± 0.6 ab	7.3 ± 2.2 b	19.0 ± 3.7 a	23.8 ± 3.4 ab
	9%	31.0 ± 1.0 b	1.1 ± 0.1 b	12.1 ± 0.6 ab	15.5 ± 1.6 b	25.8 ± 0.8 a	28.8 ± 3.1 a
Significance		*	*	*	*	*	*
Several application	Control	60.5 ± 6.5 a	9.0 ± 2.853 a	11.2 ± 3.4 a	36.0 ± 5.8 a	1.3 ± 0.5 b	1.8 ± 0.7 c
	3%	35.0 ± 3.1 b	1.5 ± 0.2 ab	12.4 ± 1.5 a	5.5 ± 0.7 b	25.5 ± 0.7 a	34.0 ± 2.9 b
	6%	27.0 ± 0.9 b	0.9 ± 0.6 bc	10.4 ± 0.3 a	6.5 ± 0.7 b	20.5 ± 2.4 a	46.0 ± 2.1 a
	9%	25.2 ± 1.7 b	0.8 ± 0.1 c	10.6 ± 0.8 a	5.8 ± 0.9 b	22.5 ± 1.9 a	46.5 ± 2.8 a
Significance		*	*	ns	*	*	*

).

With one application, there was a higher MGS in the control (p < 0.05). No statistically significant differences (p > 0.05) between the 3%, 6%, and 9% treatments were observed. In the experiment with repeated application, a gradual reduction was observed as the extract concentration increased (Table 6).

Significant differences were found between the MGT of the control and the 3% concentration with one application of the extract, which was greater with the lowest concentration; however, it also tended to be low with the other concentrations. Conversely, when applying the extract several times, no significant differences (p < 0.05) were observed between the applied concentrations (Table 6).

3.3.3. Seed Viability

Figure 5 shows the three visual scales determined according to the degree of tetrazolium staining: viable, doubtful, and nonviable. This scale was adapted from that reported by Niño-Hernández et al. [20]. The percentage of viable seeds was significantly higher in the control than in the three concentrations for one application, where there was a reduction in viability of up to 70.8% with the 3% concentration; for those that were classified as doubtful, there was a lower quantity in the control than in the other treatments, as well as an increase in the count of nonviable seeds where the applied concentration was increased. With repeated application, the viable seeds had a similar behavior, being higher than those of the control and achieving a reduction in viability of 84.7%; the doubtful ones were lower in the control than in the treatments, as well as the nonviable ones, where the viability was reduced by increasing the extract concentration (Table 6).

4. Discussion

4.1. Hydroalcoholic Extract of Campomanesia lineatifolia

Following the characterization of C. lineatifolia seeds, a large number of phenolic compounds were found (Table 2) which have direct effects on germination. While Munoz et al. [14] reported diphenols and polyphenols in C. lineatifolia, Madalosso et al. [24] found flavonoids (catechin and quercetin) and tannins. Meanwhile, Oliveira et al. [25] revealed that the seeds of another Myrtaceae, C. xanthocarpa, contain phenolic compounds with different biological activities. The same authors also indicated that the species of the Myrtaceae family are recognized for having high levels of bioactive compounds, mainly phenolics and even carotenoids.

4.2. Germination and Viability in R. crispus Seeds

The control seeds showed the highest germination (65.0 ± 5.5%) and a low percentage of viable seeds, indicating the presence of a low degree of dormancy. The seeds treated with the extract from C. lineatifolia significantly decreased germination and GS. Treatment with a 9% concentration appeared to inhibit germination up to 76.2%. However, there was a high percentage of nonviable seeds and a higher percentage of viable seeds in relation to the control (Table 4), indicating that the extract used caused the death of a large part of the seeds while inducing dormancy in others. This is an interesting result that partially clarifies the concern raised by Martínez et al. [19], who recommended using the tetrazolium viability test to understand the mechanism underlying S. oleraceus seed inability to germinate under C. lineatifolia extract application. These authors also found that a single application was sufficient to totally inhibit germination. They found that, with two weekly applications, germination was inhibited to a greater degree, but it was not totally inhibited. These findings demonstrate the species’ differential response to C. lineatifolia extract, but the causes of this response are unknown.

The phenolic compounds found in the extract, such as flavanols, phenolic acids (Trans-cinnamic acid), anthocyanins, other flavonoids and glycoside forms are primarily responsible for the extract-induced seed death (Table 2). Notably, secondary metabolites can inhibit germination by affecting important processes, such as respiration and cell division [26]. Allelochemicals can affect germination by inhibiting the cell cycle in meristematic cells [27]. In addition, C. lineatifolia extract could inhibit germination and induce dormancy due to phenol accumulation in the seed coats that can prevent the entry of oxygen into the seed; consequently, germination cannot be induced. Notably, allelochemicals can affect the permeability of seed coats and gibberellin synthesis [27]. According to Stross [28], nondormant seeds can experience biochemical reactions that deepen dormancy under conditions that affect the germination process and put seedling survival at risk, as occurred in our study with the C. lineatifolia extract. Some plants contain biologically active compounds, and their crude and methanolic extract inhibit weed seed germination [29], as observed in this study.

Notably, allelochemicals can also cause germination rate inhibition [30], which supports the results found in this study (Table 4) and can reduce initial growth. This is also a desirable effect because C. lineatifolia extract, by delaying germination, allows the crop of interest to establish itself without initial interference from weeds, and later, due to its own growth, it can generate competition for weeds. The germination of R. crispus has also been affected by the application of extracts of Swinglea glutinosa and Lantana camara [31]. This last extract inhibited the germination rate of Bidens pilosa [32], supporting the evidence that allelopathic extracts affect the rate at which germination occurs, as observed in our study. The MGT did not show statistical differences between the extracts and the control because the seeds treated with C. lineatifolia extract germinated less but faster; in the control, many germinated at a higher speed, resulting in similar MGT in different treatments.

4.3. Germination and Viability in A. hybridus Seeds

In this study, applying the extract at different concentrations (Figure 4) delayed A. hybridus germination and even inhibited it up to 58.34% in the 9% extract (Table 6). Concurrently, the viability test analysis revealed that the viable and doubtful seeds were significantly lower, with the highest concentrations of the extract (Table 6), for which the C. lineatifolia extract was shown to induce embryo death [33]. However, we do not know for sure why A. hybridus is more susceptible than R. crispus. It is possible that it is due to aspects such as the thickness and composition of the coat, the imbibition speed, among others. The C. lineatifolia extract reduced both GP and GS, most likely because plants with phenolic compositions have phytotoxic activity toward other organisms [29]. In this regard, Fan et al. [34] identified and isolated various allelochemicals from Polygonum cuspidatum (Polygonaceae); catechins and epicatechins were found, inhibiting germination by approximately 40% each, and these compounds were also found in the seeds of C. lineatifolia; therefore, it is possible that these compounds influenced the germination process of A. hybridus. Interestingly, it has been reported that the EGCG compound (present in the C. lineatifolia extract) decreased germination in tomato seeds, by affecting the balance between ABA and GA, because EGCG treatment increased the ABA content and decreased the GA₃ content [35]. In addition, Pereira et al. [36] reported that these two compounds, plus quercetin extracted from Cecropia pachystachya, independently induced a reduction in the germination of Lepidium sativum, by 10.3%, 14.5%, and 7.7%, respectively. Ximenez et al. [37] found that the flavonoid myricetin isolated from Miconia spp. had the greatest phytotoxic effect on Ipomoea triloba and Digitaria insularis.

The control seeds exhibited dormancy due to a high percentage of non-germinated viable seeds (Table 6), most likely because the seeds have different mechanisms that limit germination only when favorable conditions exist that do not jeopardize the seeds’ survival. It is possible that the total free amino acids after germination were delayed as a result of protein degradation by proteases and the synthesis of new enzymes that participate in the release of free amino acids [38]. However, according to El-Shora et al. [39], the increase in phenolic acids, such as vanillic acid, a compound characterized in this study, reduces the incorporation of specific amino acids in proteins, thus reducing the amount of proteins present in the seeds. In a comparative study between germinated and raw Amaranthus, an increase in the amount of non-essential amino acids, such as Glu, Ala, Ser, and Asp, in the germinated one was observed. The total content of free amino acids in seeds is determined as a physiological factor of the herbicide effect [40] because high concentrations indicate a high activity of proteases for reserve use.

In R. crispus, germination was inhibited by up to 38.4% with one application and 76.9% with repeated application, and in A. hybridus, by 48.73% and 58.4%, respectively. However, Cabeza–Cepeda et al. [18] and Martínez [19] obtained 100% inhibition in the germination of Taraxacum officinale and Sonchus oleraceus by applying C. lineatifolia extract in the same concentrations. These inconsistencies are most likely due to allelopathic compounds’ specificity on the amino acids that prevent their incorporation for protein formation, and these amino acids tend to differ depending on the species in which they are treated; it is also well known that resistance to traditional chemical herbicides is caused by some populations developing the ability to substitute the amino acids required for this assembly [38].

5. Conclusions

Although a single application of the hydroalcoholic extract of C. lineatifolia was sufficient to reduce the GP of both species, inhibiting R. crispus by 38.5% and A. hybridus by 48.8%, it was not as effective as constant applications that induced 76.2% and 58.34%, respectively. The application of a hydroalcoholic extract of C. lineatifolia reduced the viability of R. crispus with a concentration of 9%, increasing the number of nonviable seeds by 40.53%, whereas in A. hybridus, there was a reduction in viability of at least 70.8%. The phenolic compounds present in the hydroalcoholic extract of C. lineatifolia probably contributed to the inhibition of the germination of the two species.