Valorization of Vineyard By-Products Through Vermicomposting: A Comparative Pilot-Scale Study with Eisenia fetida and Eisenia andrei

Abstract

Vermicomposting aims to convert organic residues into valuable end products within a circular economy-based framework. Vineyards generate significant amounts of by-products, namely vine prunings (VPs), typically landfilled or incinerated, and rotten grape clusters (RGCs), which stay on the vines until removed by pruning. This pilot-scale study aimed to explore the role of two earthworm species (Eisenia fetida and Eisenia andrei) in transforming VP and RGC substrates by evaluating their physicochemical properties, phytotoxicity, and polyphenolic content before and after vermicomposting and the microbial activity at the end of the process. The substrates were vermicomposted in 2 L containers with coconut fiber (1:1 ratio) and 7.5 g of each earthworm species (clitellated and non-clitellated) per container for 100 days, with the earthworm biomass monitored every other week. Phytotoxicity was assessed using garden cress (Lepidium sativum L.) and lettuce (Lactuca sativa L.) seeds, and biological stability was assessed by microbial activity and polyphenolic content evaluation using the Folin–Ciocalteu method. The results showed that differences in the vermicompost properties were primarily substrate-dependent. The RGC-based vermicomposts exhibited higher electrical conductivity and P, K, S, and B levels, while the VP-based composts had higher C/N ratios. E. fetida produced vermicomposts with significantly higher K, Ca, and Mg contents and consistently lower phytotoxicity in germination assays with garden cress and lettuce, compared with E. andrei. Vermicomposting led to a decrease in polyphenolic content for both species. This study highlights the importance of earthworm species selection for vermicomposting vineyard residues. Further research should explore how these species perform with other residues to understand their suitability for producing high quality vermicomposts.

1. Introduction

Globally, the need for sustainable organic waste management has intensified alongside the massive generation of organic residues. In recent years, attention has shifted toward vermicomposting as a method to convert these residues into valuable end products within a sustainable and circular economy-based framework [1,2,3,4,5]. Vermicomposting is a biooxidative process carried out through the integrated activity of earthworms and microorganisms, which not only enhances the stability of the organic matter (OM) but also contributes to its hygienization. It offers a promising alternative for waste management by emphasizing resource recovery and promoting soil health and resilience [6]. When the mature vermicomposts produced meet the European Union Eco-label requirements for organic agricultural production, they are suitable for use as organic soil amendments in this sustainable practice [7]. Furthermore, vermicomposting can significantly increase bacterial diversity and enhance metabolic processes beneficial for plant growth, such as cellulose metabolism and plant hormone synthesis [8]. Given its capacity to valorize a broad range of domestic, industrial, and agricultural wastes, which can range from food scraps and paper to tannery sludge and crop residues, vermicomposting has been recognized for being a scalable, low-cost strategy for producing high-quality organic amendments while mitigating the environmental burden of unmanaged waste [9,10].

Most vermicomposting systems use epigeic earthworms of the genus Eisenia, particularly Eisenia fetida or Eisenia andrei, due to their fast growth, high reproductive performance, and efficiency in decomposing organic waste to produce high-quality and nutrient-rich vermicompost [11,12]. While these species were initially considered the same, their classification as distinct biological species [13] has led to comparative studies primarily focused on their use as ecotoxicological models [14,15]. However, research on their relative efficiency in vermicomposting different residue types remains limited. Despite their genetic differences, E. fetida and E. andrei are closely related phylogenetically [16]. Understanding which species is better suited for specific residues is crucial for optimizing the vermicomposting process and producing high-value end products. For instance, a study comparing E. fetida and E. andrei in sewage sludge-amended soils found that E. andrei demonstrates a superior ability to accumulate certain metals, making it more suitable for vermicomposting metal-rich residues (e.g., sewage sludge), where reducing the metal content in the final vermicompost is essential [17]. Conversely, E. fetida is known for its resilience and versatility across a wide range of organic residues while tolerating fluctuations in temperature and moisture [18]. Similarly, Gebrehana et al. [19] evaluated both species (along with Eudrilus eugeniae) in crop residue and manure mixtures and found that E. andrei performed better in biomass gain and cocoon production in nutrient-rich substrates like soybean husks, while E. fetida was more consistent across different mixtures and promoted more even nutrient release and microbial enzymatic activity. Additionally, Tajbakhsh et al. [20] assessed the efficiency of both species in recycling spent mushroom compost with various agricultural wastes and reported significant enhancements in total nitrogen, phosphorus, and micronutrients following vermicomposting with both species, though no species-specific trends were emphasized.

The wine industry is one of the most important sectors of agricultural production, holding substantial socioeconomic significance, particularly in Portugal, where it contributes extensively to the national income [21]. Despite this, increasing wine production has brought challenges to the sector in terms of residue management [22]. The viticulture and wine production processes generate various by-products, namely vine shoots, grape stalks, grape pomace, vine prunings, and wine lees, which are traditionally managed through landfilling or incineration [22]. These have notable environmental consequences, including soil, water, and air pollution [23,24,25]. Rotten grape clusters, though not traditionally considered by-products, also contribute to residue production. Grape clusters often fall prematurely due to early ripening or disease, decomposing on the ground and fostering agricultural pests (e.g., flies and maggots). Instead of being left to decompose, they can be repurposed into value-added products through vermicomposting as organic amendments [26]. Although the efficacy of vermicomposting vineyard residues has been addressed in several studies [22,27,28], direct comparisons of E. fetida and E. andrei under the same conditions remain scarce. This gap is especially evident for certain under-studied residues, like those from the vine and wine industries. Such comparisons are essential to better understand species-specific nutrient transformations and optimize the production of high-quality vermicomposts.

This study aims to investigate the species-specific differences in vermicompost production by E. fetida and E. andrei, focusing on their performance with two different substrates: vine prunings (VPs) and rotten grape clusters (RGCs). The study evaluates the physicochemical properties of the vermicomposts and their phytotoxicity, microbial activity, and polyphenolic content to assess quality and suitability for agricultural use.

2. Materials and Methods

2.1. Earthworms and Substrates

Earthworms, namely E. andrei and E. fetida, were obtained from a culture maintained at the Universidade de Trás-os-Montes e Alto Douro (UTAD) in Vila Real, Portugal under optimal controlled moisture and temperature conditions and fed regularly with organic matter of a vegetal origin. The vineyard residues, namely vine prunings (VPs) and rotten grape clusters (RGCs), were collected from local vineyards in Vila Real, Portugal, with VPs collected during the 2022–2023 pruning season and RGCs obtained during the 2023 harvest. The VPs were roughly chopped to reduce the particle size and promote decomposition, whereas the RGCs were used whole without any physical processing. Coconut fiber (CocoPlus Aquarian Development Incorporated, San Pablo Laguna, Philippines) was used as the initial bedding for the earthworms based on its successful application in previous experiments by our team and its documented effectiveness in supporting earthworm activity in the literature [29].

The study involved four treatments: vermicomposting using E. fetida (VP-F and RGC-F) and E. andrei (VP-A and RGC-A). Each treatment was prepared in triplicate in 2 L (300 cm²) plastic containers. They consisted of a 150 g layer of coconut fiber placed at the bottom of each container and an upper layer of the different vineyard residues (150 g of VPs or RGCs) mixed with 10 g of fallen grape leaves, which were added to provide additional nutrients to the earthworms during the initial phase. To each container, 7.5 g of earthworms (clitellated and non-clitellated) of each species were added. The containers were maintained at controlled room temperature (22 ± 2 °C), and moisture conditions were assessed by touch (sponge test) and adjusted as needed with distilled water throughout the 14 week (100 days) experimental period while being kept in darkness. No mechanical homogenization was performed, as substrate mixing was ensured by earthworm movement. The earthworm biomass was frequently monitored during the experiment. At the end of the 100 day period, the entire content of each container was collected for physicochemical analysis, biological and phytotoxicity assays, and polyphenol content evaluation.

2.2. Physicochemical Analysis

The initial substrates (VPs and RGCs) were characterized at the Soil and Plant Laboratory of UTAD using certified analytical methods (

Table 1. Laboratory methods (ISO guidelines) and initial physicochemical characteristics of the substrate mixtures.

	Laboratory Methods	Substrate
		VP	RGC
Physical Properties
C/N	ISO 10694:1995 ISO 13878:1998	93.6	87.8
OM (g kg−1 DW)	ISO 23400:2021	439	533
Ph (H2O)	ISO 10390:2005	5.25	5.08
EC (dS m−1)	ISO 11265:1994	1.45	1.53
N-NH4+ (mg kg−1 FW)		n.d.	n.d.
N-NO3− (mg kg−1 FW)		n.d.	n.d.
WC (g kg−1 DW)	ISO 11465:1993	431.3	500.6
Macronutrients (g kg−1 DW)
N	ISO 13878:1998	2.72	3.52
P	ISO 11263:1994	0.37	0.74
K	ISO 23470:2018	6.61	10.80
Ca	ISO 11260:2018	2.53	1.68
Mg		1.05	0.93
S		0.36	0.51
Micronutrients (mg kg−1 DW)
B		12.4	16.9
Fe	ISO 11466:2005	9255	6671
Mn	ISO 11466:1995	132.0	99.3
Metals (mg kg−1 DW)
Cd	ISO 11466:1995	0.05	0.02
Cr		25.4	12.4
Cu		13.69	7.07
Pb		1.48	1.19
Hg		0.012	0.015
Ni		6.27	3.10
Zn		31.7	12.0

Abbreviations: DW = dry weight; EC = electrical conductivity; n.d. = not detectable.

). The initial physicochemical characteristics of both substrates, reflecting the complete substrate mixtures used in each treatment, including VPs or RGCs, coconut fiber, and fallen grape leaves, are also described in Table 1.

At the end of the experiment, each vermicompost obtained was randomly sampled, and their properties were determined following the same analytical procedures.

2.3. Phytotoxicity Analysis

The phytotoxicity of both initial substrates and the four vermicomposts was evaluated through germination and radicle elongation tests using garden cress (Lepidium sativum L.) [30] and lettuce (Lactuca sativa L.) [31,32] seeds.

Samples from the initial substrates and the final vermicomposts were dried at 80 °C for 24 h. The respective water suspensions were prepared by mixing the dried material with distilled water (20% w/v), shaking them at 180 rpm for 3 h using a linear reciprocating shaker (SH-200, Cole-Parmer Ltd., St. Neots, UK), and subsequently filtering them with Whatman no. 2 paper. Afterward, 20 seeds were placed in sterile 9 mm Petri dishes containing Whatman no. 42 filter paper embedded with 3 mL of either distilled water (control) or the filtered suspensions. Three replicates were prepared for each treatment. The Petri dishes were sealed with parafilm to prevent desiccation and incubated in the dark (E07086, Memmert GmbH + Co. KG, Schwabach, Germany) at 28 °C for 4 days. Following the incubation period, the number of germinated seeds was recorded and photographed to measure their radicle lengths using Digimizer 5.7.2 software (Digimizer, MedCalc Software, Belgium). Phytotoxicity responses were evaluated in terms of the germination index (GI), which was calculated with the following formula [33]:

$$GI (\%)=\left({\displaystyle \frac{G}{G0}}\right)\times \left({\displaystyle \frac{L}{L0}}\right)\times 100$$

(1)

where G and L are the germination percentage and radicle growth of the test suspensions, respectively, and G0 and L0 are the germination percentage and radicle growth of the control, respectively.

A germination index (GI) below 80% was considered indicative of phytotoxicity, while values below 50% were classified as highly phytotoxic, according to established thresholds [34,35].

2.4. Microbial Activity

Basal respiration, indicative of biological stability, was estimated from CO₂ production as described by Anderson [36]. Samples of the vermicompost (5 g fresh weight) were placed together with the CO₂ trap, a container with 30 mL of 1 M NaOH solution, in 1 L glass vessels, which were then hermetically sealed and incubated at 25 °C for 4 days. CO₂ production was quantified by NaOH titration with 1 M HCl to a phenolphthalein endpoint, after the addition of excess BaCl₂ (10%, m/v). Basal respiration was calculated using the following formula:

$$Basal respiration={\displaystyle \frac{\left[\frac{(Vc-Vs)\times N\times 22}{Z}\right]}{4}}$$

(2)

where basal respiration is expressed in mg CO₂ g⁻¹ OM d⁻¹, Vc is the volume of hydrochloric acid used for the control sample, Vs is the volume of hydrochloric acid used for each sample, N is the normality of the HCl solution (mol L⁻¹), 22 g is the equivalent weight of CO₂, and Z is the dry OM content in the sample.

2.5. Extraction and Determination of Total Polyphenols

Polyphenolic compounds were extracted from each initial substrate mixture and vermicompost sample, which were previously lyophilized and ground up, using 50% acidified ethanol as a solvent. Samples of approximately 0.1 g were mixed with 3.4 mL of solvent solution and stirred for 1 h. Following extraction, the samples were centrifuged at 1800× g for 10 min, and the pellets were re-extracted under the same conditions. Supernatants from both extractions were collected and combined. The extraction process was performed in triplicate for each sample.

The total polyphenol content was determined using the Folin–Ciocalteu reagent method [37]. Briefly, 100 µL of the supernatant was added to 1.5 mL of distilled water. After 3 min, 100 µL of the Folin–Ciocalteu reagent and 300 µL of Na₂CO₃ 20% solution were added. The mixture was incubated at room temperature for 2 h.

Blank and calibration solutions of gallic acid were prepared using the same procedure. The absorbance of the samples and calibration solutions was measured against the blank solution at 765 nm using a spectrophotometer (Evolution 201, Thermo Scientific, Watham, MA, USA). The total polyphenol content was expressed in milligrams of gallic acid equivalents (GAEs) per gram of dry mass [37].

2.6. Statistical Analysis

Statistical analyses were performed using GraphPad Prism Version 10.4.1 (GraphPad Software, San Diego, CA, USA). Data were tested for normality using the Shapiro–Wilk test. Two-way analysis of variance (ANOVA) was employed for analysis of the results, followed by Tukey’s post hoc test. Repeated-measures two-way ANOVA was employed for analysis of the biomass results throughout the experimental study. Data are presented as the mean ± standard deviation (SD) unless otherwise specified. Significant differences were considered when p < 0.05.

Spearman correlations and principal component analysis (PCA) were conducted with all variables. Spearman correlations were conducted with SPSS 29.0 (IBM Corp., Armonk, NY, USA), and PCA was performed using STATISTICA v. 12 software (StatSoft, Tulsa, OK, USA).

3. Results

3.1. Evolution of Earthworm Population Dynamics

This study revealed a non-significant effect from the substrate, species, and time on the earthworm fresh biomass (Figure 1). Except for RGC-F, all treatments presented a similar time-dependent decrease in earthworm biomass. In RGC-F, an initial and steady increase was observed until day 35, and thereafter, the earthworm fresh biomass showed similar behavior to the other treatments. For all of the treatments, the earthworms’ biomass decreased along the vermicomposting process, reaching the lowest values at the 100th day (VP-F: 2.99 ± 1.23 g; VP-A: 2.58 ± 0.52 g; RGC-F: 5.07 ± 0.72 g; RGC-A: 2.34 ± 1.25 g).

3.2. Physicochemical Compositions of the Substrates

The physicochemical compositions of the vermicomposts are presented in Figure 2 (ANOVA results in

Table 2. Effects of substrate, earthworm species, and their interaction on the physical properties of vermicomposts based on two-way analysis of variance (ANOVA).

Source of Variation	Physical Properties
	C/N	EC	OM	pH	N-NH4+	WC
Substrate	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001
Species	0.288	0.040	0.004	<0.001	<0.001	0.004
Substrate × Species	<0.001	0.325	<0.001	<0.001	<0.001	0.001

), focusing on the carbon-to-nitrogen (C/N) ratio, electric conductivity (EC), OM content, pH (H₂O), ammonium (N-NH₄⁺) and nitrate (N-NO₃⁻) concentrations, and water content (WC), which are important parameters of vermicompost quality and agricultural suitability. Differences were found between the substrates regarding all parameters. The VP-based vermicomposts exhibited higher C/N ratios and OM contents but lower pH levels and EC compared with the RGC-based ones.

The C/N ratio and OM analysis highlight the differences between species in the decomposition process, particularly for VPs, where VP-F resulted in a significantly higher C/N ratio compared with VP-A (p < 0.001). On the contrary, RGC-A yielded a higher C/N ratio and OM content than RGC-F (p < 0.001). The pH levels varied significantly with both the species and substrate (p < 0.001); E. fetida led to more alkaline conditions in both the VPs and RGCs compared with E. andrei, while RGCs yielded vermicomposts with significantly higher pH levels (p < 0.001). Vermicomposting significantly reduced the EC of the VP substrates. The N-NH₄⁺ content also differed significantly between species and substrates (p < 0.001), with E. andrei presenting higher values for both substrates (p < 0.001). Regarding N-NO₃⁻, it was below the detection limits for VP-F, preventing statistical comparisons, but followed the same trend as N-NH₄⁺, with E. andrei presenting the highest values for both substrates. Lastly, the WC was similar between species for the VPs but higher for RGC-F compared with RGC-A (p < 0.001).

The contents of macroelements in the vermicomposts are summarized in Figure 3 (ANOVA results in

Table 3. Effects of substrate, earthworm species, and their interaction on the contents of macronutrients present in the vermicomposts based on two-way analysis of variance (ANOVA).

Source of Variation	Macronutrients
	N	P	K	Ca	Mg	S
Substrate	<0.001	<0.001	<0.001	<0.001	0.211	<0.001
Species	0.811	0.017	<0.001	<0.001	0.016	0.001
Substrate × Species	<0.001	0.001	0.018	<0.001	0.050	0.001

). The RGC vermicomposts presented higher P, K, and S contents than the VP vermicomposts, particularly for E. fetida (p < 0.001). In contrast, the VP vermicomposts yielded higher Ca concentrations for both E. fetida and E. andrei, with significant differences between substrates and species (p < 0.001). E. andrei consistently showed lower Ca contents than E. fetida across both materials (p < 0.001). The magnesium content was significantly higher in the vermicompost produced by E. fetida, using the RGC substrate (p < 0.01).

Figure 4 presents the microelement contents of vermicomposts (ANOVA results in

Table 4. Effects of substrate, earthworm species, and their interaction on the contents of micronutrients present in the final vermicomposts based on two-way analysis of variance (ANOVA).

Source of Variation	Micronutrients
	B	Fe	Mn
Substrate	<0.001	<0.001	0.907
Species	0.001	<0.001	<0.001
Substrate × Species	0.004	<0.001	0.015

). The boron and Fe contents were significantly higher in the vermicompost derived from RGCs (p < 0.001). For the VPs, the B and Fe contents were significantly higher in the vermicompost produced by E. fetida (VP-F) (p < 0.01). Similar levels of Mn were observed between both substrates, but species-specific differences were noted, being substantially higher in the vermicomposts from E. andrei for both substrates (p < 0.001).

The vermicomposts’ metal contents (Cd, Cr, Cu, Pb, Hg, Ni, and Zn) are detailed in Figure 5 (ANOVA results in

Table 5. Effects of substrate, earthworm species, and their interaction on the contents of metals present in the vermicomposts based on two-way analysis of variance (ANOVA).

Source of Variation	Metals
	Cd	Cr	Cu	Pb	Hg	Ni	Zn
Substrate	0.122	0.582	<0.001	0.054	1.000	0.110	<0.001
Species	0.002	0.005	0.811	0.902	0.448	0.005	<0.001
Substrate × Species	0.282	0.007	<0.001	0.011	0.923	0.944	<0.001

). Except for the Hg content, which presented no differences among the substrates and species, the other metals showed significant effects either from the substrate or species. Among the significant differences, it is noteworthy that the vermicomposts using VPs as a substrate presented superior Cu and Zn contents, particularly when E. andrei conducted the process. The remaining metal content differences were random, with the species, the substrates, or the interaction of both factors being responsible for the observed results.

3.3. Germination Index

Germination index (GI) tests using garden cress and lettuce seeds were used to evaluate the phytotoxicity of VPs and RGCs before and after vermicomposting. The phytotoxicity analysis of the initial material revealed that VPs and RGCs were highly toxic for the selected species, as indicated by the extremely low GI values obtained. The garden cress seeds presented a GI of 16.64 ± 2.00% for VPs and 9.58 ± 1.19% for RGCs, while for lettuce seeds, the GIs were even lower, being 0.64 ± 1.11% for VPs and 0% for RGCs.

Regarding the vermicomposts, the GI values for the garden cress and lettuce seeds are presented in

Table 6. Germination indexes (GIs) of garden cress and lettuce seeds, with basal respiration rates and total polyphenols of the vermicomposts. Data are presented as mean ± SD. Statistical analysis was performed using two-way ANOVA. Here, * represents significant differences between the substrates within the same species (p < 0.05).

	GI (%)		Basal Respiration (mg CO2 g−1 OM d−1)	Total Polyphenols (mg GAE g−1 DW)
	Garden Cress	Lettuce
Vine Prunings (VPs)
E. Fetida (VP-F)	250.04 ± 30.43 *	125.54 ± 26.78 *	4.930 ± 3.360	0.85 ± 0.09
E. Andrei (VP-A)	212.06 ± 39.78 *	112.06 ± 6.22 *	6.020 ± 0.095	0.81 ± 0.16
Rotten Grape Clusters (RGCs)
E. Fetida (RGC-F)	48.52 ± 18.53	34.05 ± 8.01	7.863 ± 1.438	0.95 ± 0.24
E. Andrei (RGC-A)	39.82 ± 13.83	29.80 ± 16.59	8.093 ± 4.953	0.96 ± 0.15
Two-Way ANOVA (p Values)
Substrate	<0.001	<0.001	0.197	0.235
Species	0.181	0.381	0.720	0.851
Substrate × Species	0.385	0.642	0.815	0.773

Abbreviations: DW = dry weight; GAE = gallic acid equivalent.

. The results revealed that the GI values were significantly affected by the substrate origin (p < 0.001), while the earthworm species used had no significant impact for either plant species. After vermicomposting, the toxicity of the substrates was depleted, and the VP vermicompost even presented germination- and root elongation-stimulating properties, particularly for garden cress, which showed GI values exceeding double that of the control (distilled water). The RGC-based vermicomposts, although presenting lower toxicity than the initial substrates, did not reach the control values regarding the GI. Although no significance was observed between the vermicomposts produced by the two species, E. fetida seemed to be better at reducing the initial substrates’ toxicity.

3.4. Basal Respiration

The basal respiration rates of the vermicomposts were evaluated to determine biological stability based on microbial activity, as presented in Table 6. Overall, no significant differences in basal respiration rates were observed between the earthworm species or vermicompost types in the final products. However, the lowest values were found consistently in the vermicomposts produced by E. fetida (p > 0.05).

3.5. Total Polyphenols

The total polyphenol concentrations in the final vermicomposts, as shown in Table 6, also presented no significant differences between those derived from VPs and RGCs, regardless of the earthworm species used. However, these values were considerably lower when compared with the initial materials (VP: 3.87 ± 0.43 mg GAE g⁻¹ DW; RGC: 4.93 ± 0.72 mg GAE g⁻¹ DW).

3.6. Correlation Analysis

Figure 6 presents the Spearman correlations between the various parameters analyzed in this study, highlighting the impact of the different earthworm species on these parameters. The seed germination indexes for cress and lettuce were found to be influenced by several parameters, namely the physicochemical properties [C/N ratio (ρcress = 0.94, ρlettuce = 0.92), OM (ρcress = 0.87, ρlettuce = 0.84), pH (ρcress = −0.74, ρlettuce = −0.74), EC (ρcress = −0.95, ρlettuce = −0.93), WC (ρcress = −0.77, ρlettuce = −0.78)], macronutrients [N (ρcress = 0.92, ρlettuce = 0.91), P (ρcress = −0.95, ρlettuce = −0.92), K (ρcress = −0.95, ρlettuce = −0.93), Ca (ρcress = 0.59), and S (ρcress = −0.93, ρlettuce = −0.91)], as well as the micronutrients [B (ρcress = −0.94, ρlettuce = −0.71)] and metals [Cu (ρcress = 0.92, ρlettuce = 0.91) and Zn (ρcress = 0.90, ρlettuce = 0.90)]. The final earthworm biomass was found to be positively correlated with the WC (ρ = 0.74) and Mg (ρ = 0.62), and negatively correlated with Cr (ρ = −0.85).

3.7. Principal Component Analysis

Principal component analysis (PCA), based on the physicochemical properties, macro- and micronutrient contents, metal concentrations, and biological parameters, revealed that the first two factors explained 85% of the total variance (Figure 7). Factor 1, which was the most significant one, accounted for 58% of the variance and differentiated the residues (RGCs and VPs) used in the vermicomposting process. The clear separation between these groups highlights significant structural and functional differences between the two types of residues. Factor 2 explained an additional 27% of the variance and separation of the earthworm species: E. fetida (F) and E. andrei (A). Despite this differentiation, a high degree of similarity was observed between the samples treated with E. fetida and E. andrei within each residue type, suggesting that the earthworm species had a lesser influence compared with the nature of the organic material itself. These findings highlight the key role of the initial composition of the organic substrate in determining the final characteristics of the vermicompost. Conversely, the impact of the earthworm species appeared to be more limited within the range of variables analyzed, emphasizing the importance of selecting appropriate raw materials for vermicomposting.

4. Discussion

This study assessed the physicochemical transformation of vineyard-derived residues, namely VPs and RGCs, during vermicomposting with E. fetida and E. andrei. Vermicomposting is an efficient bioprocess for valorizing and transforming agro-industrial by-products into stabilized, nutrient-enriched amendments suitable for agricultural application [22,38,39]. To the best of our knowledge, this study represents the first direct comparison of E. fetida and E. andrei on vineyard by-products under the same controlled conditions, focused on determining the effects of each species on the agronomic attributes of the vermicomposts.

4.1. General Effects of the Vermicomposting Process

Vermicomposting caused characteristic physicochemical changes in the substrate across all treatments, including reduced C/N ratios, elevated pH levels, a decreased total polyphenol content, and an improved GI [40,41]. These changes reflect a sequence of biochemical events, both microbial- and earthworm-driven, leading to OM mineralization [42]. The initial substrates, particularly VPs, exhibited high C/N ratios, indicative of recalcitrant carbon pools (e.g., lignin and cellulose), which were progressively degraded through oxidative enzymatic pathways and microbial respiration [43]. However, none of the final vermicomposts reached the commonly accepted maturity threshold of a C/N ratio below 20, reflecting poor mineralization due to the high lignocellulosic content of the substrate. Seeing as a C/N ratio below 12 is preferred for agricultural application [44,45], the introduction of N-rich materials is required to improve agronomic quality, especially for VP-derived vermicomposts. The use of coconut fiber as the initial bedding, though common to ease earthworm acclimation, may have contributed as an additional lignocellulosic material, increasing the C/N ratios in the VP and RGC mixture treatments. As a potential strategy for future studies, co-composting with N-rich inputs like manure could help balance the C/N ratios and improve the efficacy of the vermicomposting process [46,47], though this may also introduce challenges such as increased salinity or pathogen risk that would need to be addressed [48]. Although Spearman’s correlation indicated a positive association between the total N and C/N ratio across the samples, this reflects treatment-specific patterns rather than a biological relationship. During vermicomposting, the total N typically increases while the C/N ratio decreases due to carbon mineralization [49]. In this study, the C content decreased moderately, while N increased substantially, leading to overall lower C/N ratios in the final vermicomposts. Additionally, the apparent increases in nutrient and metal concentrations, including N, Ni, and Zn, in the final vermicomposts can be attributed to organic matter degradation, leading to a concentration effect. In addition, gut passage, cast deposition, and the stimulation of microbial activity by earthworms may enhance the solubilization and redistribution of these elements. Notably, earthworms have been shown to significantly influence the diversity and abundance of functional microbial genes such as amoA and nirS, thereby affecting, for example, nitrogen biotransformation pathways during vermicomposting [50].

Polyphenolic compounds, which are known to exert antimicrobial and phytotoxic effects [51,52], also influence the microbial activity and decomposition dynamics during vermicomposting. These compounds, however, underwent significant degradation during the process, as previously observed in a study on grape marc vermicomposting [53], with no significant differences between earthworm species. This suggests that phenolic breakdown is largely microbially mediated or possibly facilitated by conserved earthworm detoxification systems (e.g., glutathione S-transferases or drilodefensins) [53,54]. Additionally, as vermicomposting progresses, and the easily degradable OM is consumed, the process enters a stabilization phase characterized by reduced microbial activity, as evidenced by the lower basal respiration observed in the final vermicomposts, indicating depletion of labile carbon substrates and a shift toward a more stable product [55].

4.2. Substrate-Specific Differences

Substrate origin exerted a strong influence on the vermicompost physicochemical profiles. The RGC-derived vermicomposts consistently exhibited lower C/N ratios and higher electrical conductivity (EC), paralleled by higher contents of P, K, S, and B. These traits are consistent with the intrinsic composition of decomposed grape clusters, being rich in organic acids and mineral ions [56,57]. The rapid microbial turnover of these substrates likely accelerated nutrient mineralization but also contributed to salt accumulation, as evidenced by higher EC values. Additionally, the B levels in the RGC vermicomposts (>30 mg kg⁻¹ DW observed in RGCs) fell within a range that has been associated with potential phytotoxicity in plant species [58,59,60]. Regarding S, excessive levels of this macronutrient may induce phytotoxicity and chlorosis or necrosis, ultimately leading to plant death [61]. However, the S levels detected in this study should be interpreted with caution, given that they are not excessively high but may ultimately contribute to increased osmotic stress. The VP substrate contains higher amounts of lignocellulosic materials, which decompose more slowly [62] and thus do not contribute as rapidly to salt release. In contrast, RGCs decompose more rapidly, contributing to a higher salt release. Although RGC-based vermicomposts are apparently richer in nutrients, their elevated salinity may pose challenges in sensitive applications like seed germination or peat substitution [63,64], in order to not compromise early plant development [65,66]. To mitigate this, strategies such as mixing with other materials like zeolite have been recommended [67]. This is particularly important in early plant development stages, as elevated osmotic potentials are known to impair seed imbibition and germination [68].

VP-based vermicomposts, despite having higher C/N ratios and being similar to other woody residue composts [69], indicative of lower maturity, had lower EC values compared with the initial materials. This reduction in EC, unlike what was observed with the RGCs, may have contributed to the observed decrease in phytotoxicity. These results may also be attributed to the presence of bioactive compounds such as lignin-derived fragments, which have been reported to promote plant growth [70,71]. Principal component analysis strongly supported this substrate-driven divergence, indicating that initial material composition was the principal determinant of the vermicompost physicochemical properties.

Phytotoxicity assays provided further insights into the agronomic suitability of the final vermicomposts. The GI values were mainly determined by the substrate, with the VP-based vermicomposts exhibiting markedly higher GIs compared with those derived from RGCs. For garden cress, the GI exceeded 200% in both the VP-F and VP-A treatments, indicating strong biostimulant properties rather than toxicity. The lettuce seeds showed a similar trend, although the GI values were generally lower, reflecting their greater sensitivity [72]. In contrast, the RGC-based vermicomposts yielded GI values below 50% for both species, suggesting residual phytotoxicity, which is likely linked to elevated EC and B concentrations [60,73]. These findings align with previous studies where mature grape marc or vine pruning composts exhibited higher GI values, while phytotoxicity was commonly reported in less stabilized residues, especially those rich in sugars and phenolics, such as grape skins or pomace [74,75,76,77]. These findings further reinforce the conclusion that the substrate characteristics, especially the salinity and nutrient contents, are primary determinants of phytotoxicity outcomes in vermicompost.

4.3. Earthworm Species-Specific Influences

Earthworm biomass dynamics further illustrated the interaction between species and substrate. Both earthworm species showed biomass gains during the initial 7 days in most treatments. However, in RGC-F, the biomass continued increasing until day 35, suggesting sustained availability of labile organic compounds [78]. This trend aligns with E. fetida’s known preference for nutrient-rich, rapidly degradable substrates such as fruit residues [79]. In contrast, E. andrei did not exhibit extended growth in the same substrate, indicating a possible limitation in substrate processing or nutrient assimilation. These results contrast with prior studies that showed superior E. andrei biomass gains in other organic residues, such as animal manure [80] and grape marc [53]. From day 91 onward, the biomass declined across all treatments, consistent with the exhaustion of accessible OM in closed systems, coupled with the appearance of the transformed substrates and the expectation that further biological activity would cease beyond this point [78]. However, this decline may not solely indicate reduced earthworm performance or mortality. Cocoon production was visually observed throughout the trial, suggesting active reproduction. Therefore, the decrease in biomass may also reflect population turnover, with biomass redistributed among a higher number of younger, less developed individuals (i.e., hatchlings and early-stage juveniles).

While the substrate type dominated the transformation profile, the earthworm species modulated specific trends. The lowest C/N ratios were recorded for VP-A and RGC-F, suggesting a substrate-specific advantage for each earthworm species. Overall, E. andrei was more effective in promoting nitrogen transformations through N mineralization and nitrification [81], resulting in increased mineral (ammoniacal and nitrate) N contents across substrates. This indicates both improved decomposition [82] and enhanced N availability for plant uptake [83]. Overall, these results conflict with those reported previously when comparing the vermicomposting of sewage sludge by E. fetida and E. andrei, which highlighted E. fetida as promoting more N transformations [17]. However, the lower C/N ratio in RGC-F suggests a greater reduction in C levels associated with E. fetida, which in turn is supported by this same study [17]. These results, accompanied by those reported by the literature, highlight the evident differences between earthworm species in terms of bioconversion of different input materials. Similarly, a study by Devi and Khwairakpam [84] concluded that when vermicomposting Lantana camara, E. fetida consistently led to greater organic carbon degradation and reductions in CO₂ evolution and nitrogen transformations in comparison with Eudrilus eugeniae, implying that these differences extend to a wide variety of earthworm species. Furthermore, E. fetida produced vermicomposts with consistently higher P, K, Ca, and Mg contents, elements which are central to plant nutrition [85], likely due to species-specific differences in digestion efficiency and gut microbiota. A previous study found that the use of E. andrei in vermicomposting cow manure led to a decrease in bacterial and fungal biomass, as well as reduced bacterial growth rates and total microbial activity, in contrast to E. fetida [86], which may reflect the observed differences in nutrient mineralization and solubilization [87]. Conversely, the E. andrei vermicomposts showed relatively elevated levels of Mn and Ni, suggesting an alternative metal accumulation capacity, as previously reported [17]. Interestingly, the E. fetida vermicomposts were enriched with Cu and Zn, both of which showed positive correlations with the GI in garden cress and lettuce. These findings suggest that at the concentrations observed, Cu and Zn functioned within their optimal physiological ranges, supporting enzymatic and hormonal activity during seed germination [88,89] rather than inducing phytotoxicity, as previously observed [90,91]. The final vermicomposts from both species remained within the permissible limits for heavy metals under Portuguese legislation (Portaria 185/2022 of 21 July), ensuring their suitability for agricultural application. Despite minor differences between earthworm species, both were effective at degrading phenolic compounds, as evidenced by comparable reductions in polyphenol content and basal respiration. Nevertheless, stabilization of microbial activity should be interpreted with caution, given that the basal respiration values were not as low as expected according to the stability thresholds reported in the literature [92].

Taken together, the results support the use of E. fetida for enhanced nutrient contents and reduced phytotoxicity, particularly in substrates with higher salt and nutrient contents. However, the dominant factor in vermicompost quality remains the physicochemical nature of the input substrate. Substrate-specific strategies (e.g., introduction of animal manure) may be required to enhance vermicomposting efficacy and improve its characteristics for agronomic use. Additionally, combining VPs and RGCs could also mitigate substrate-specific drawbacks (high EC in RGCs and high C/N ratio in VPs), but it is important to consider that they are typically obtained at different vineyard management stages, complicating synchronous co-vermicomposting. Given the close phylogenetic and functional similarities between E. fetida and E. andrei, future studies should explore more distantly related species, namely E. eugeniae, Perionyx excavatus, and Dendrobaena veneta, with potentially distinct decomposition pathways and microbial associations that could further optimize residue transformation. Measuring the elemental composition of earthworms pre- and post-vermicomposting would also help determine nutrient assimilation and should be included in future studies.

5. Conclusions

This pilot-scale comparative study highlights the importance of carefully considering earthworm species for vermicomposting specific residues. Although E. fetida and E. andrei are similar, and most changes were observed between substrates. Their differences can also significantly impact vermicompost quality, namely through nutrient profiles. Overall, E. fetida showed superior performance in several key parameters essential to producing high-quality vermicompost from these by-products, namely favorable physicochemical properties, better nutrient profiles, and lower phytotoxicity. Further research should explore how these species influence other parameters such as humic acids, phytohormones (including auxins, gibberellins, and cytokinins), enzymes (such as dehydrogenase, phosphatase, and urease), and species-specific gut microbiota, while also exploring different types of residues to better understand their suitability and optimize the final product quality.