Composting was performed by mixing organic waste from the University Restaurant (UR) and gardening waste from the Federal University of Lavras–UFLA, Lavras, Minas Gerais, Brazil. The research was conducted from 2017 to 2018 in Lavras, Minas Gerais, Brazil.

The initial physicochemical characterization of both residues was carried out so that the initial C: N ratio was calculated (Table 1) and, consequently, the ideal proportion (30:1) of each residue to be composted to start the process in an ideal and satisfactory way (Pisa et al., 2020). The waste from the restaurant is composed of leftover rice, beans, vegetables, lettuce leaves, purple cabbage, carrots, and zucchini pieces, among others in smaller quantities. Residues from the landscaping management of the university campus such as leaves from the trees that are swept and residue from tree and lawn pruning were also used to balance de C: N ratio.

The experiment was carried out using two different inoculants: a non-commercial (EM) and a commercial (CI) biological product composed of Lactobacillus plantarum 10⁴ CFU/mL, Saccharomyces cerevisiae 10³ CFU/mL, sugarcane molasses and water. EM was prepared according to Sharma et al. (2017) by cooking 700 g of unsalted rice in distilled water as a non-specific substrate for capturing different microorganisms in native forests, rich in beneficial microorganisms. After cooking, the rice was distributed in small amounts in a disposable plastic container (500 g), covered with perforated plastic, and placed under the burlap in the forest. Five disposable plastic bottles were distributed in the forest in a radius of 20 meters., for the growth of natural microorganisms, for 15 days, where the temperature and humidity were on average 19°C and 80%, respectively. Then, the microorganisms were deposited in a 20-liter bottle, homogenized with 1 liter of sterile sugar cane juice as a carbon source, and later, the volume of the bottle was filled with distilled water in a ratio of 20:1. The bottle was stored in a ventilated and cool place for 20 days, a period sufficient for multiplication and fermentation of the natural microbiota (Kwon et al., 2011) to obtain the inoculum EM. We did not identify the microbiota present in the inoculum since there was no growth in the culture medium.

The commercial inoculum was prepared and activated according to the manufacturer's instructions technical sheet. The recommended amount of distilled water and commercial sugar was added to the product, and the fermentation took place for 20 days in a conditioned place at room temperature (21–26°C). The compost without the addition of inoculants was used as a control treatment.

The waste was crushed separately into small particles (1–2.5 cm) with the help of a shredder. After being crushed, the organic waste from the UR was added directly to the gardening waste in a proportion of 1:2.5, respectively. Then the wastes were mixed with the help of a tractor, resulting in a homogeneous mixture. The present work used a bioreactor (1 m3) designed with an automatic air injection system and built with masonry blocks (Supplementary Figure 1). After homogenization, microbial inoculation was applied once, at the beginning of the experiment. In each experimental unit, 5 liters of inoculum (EM or CI) and 15 liters of water were used and properly identified. In the control treatment, only water was added. In the control treatment, only water was added.

Samples were collected based on the changes in the composting phases, which were monitored by temperature measurements at intervals of 0, 5, 10, 20, 40, 60, and 120 days, as seen in Table 2. To ensure a representative sample, subsamples were taken from nine distinct locations of each composter, being 3 surface samples, 3 interior samples, and 3 depth samples. The subsamples were mixed in equal amounts for each treatment to obtain a final sample weight of ~500 grams.

Twenty-five g of each sample was added to 225 ml of sterile peptone water on an orbital shaker at 120 revolutions per minute (rpm), for 30 min, at room temperature, on average of 25°C (López-González et al., 2013). The samples were mixed in a stomacher at an average speed of the 60's, and 10-fold dilutions were prepared. The microbial population was evaluated using four different culture media. YEPG (Yeast Extract-Peptone-Dextrose) and DRBC (Dichloran Rose Bengal Chloramphenicol) media at 28°C for 48 h for isolating yeasts (Nathaniel et al., 2020) and PDA (Potato Dextrose Agar) and DG18 (Dicloran Glycerol Agar) media at 25°C for 7 days for isolating filamentous fungi (Rempel et al., 2018).

For morphological analysis and grouping of filamentous fungi, the following macroscopic characteristics were observed: (1) color of the colony; (2) mycelium color; (3) color of the back of the petri dish, and (4) absence or presence of streaks. After grouping based on macroscopic characteristics, the fungi were identified based on microscopic characteristics, such as (1) septation of hyphae; (2) shape and size of vesicles and conidia, and (3) length and texture of the stipe observed on fresh slides stained with methylene blue and observed under a microscope. Subsequently, these fungi were picked on specific media according to the genera found. The fungi identified as belonging to the genus Cladosporium were picked in MEA medium (in g L-1: 20 g malt extract, 1 g bacteriological peptone, 20 g glucose, 20 g agar) and incubated at 25°C for 7 days. The fungi identified as the genera Aspergillus and Penicillium were picked in CYA medium (in g L-1: 1g K₂HPO₄, 10mL, Czapek concentrate, 5 g yeast extract, 1 mL metallic solution, 20 g agar) and incubated with MEA at 25°C and 37°C for 7 days.

The identification based on the microscopic characteristics of the filamentous fungi of the Aspergillus, Penicillium, and Cladosporium species was based on the keys proposed by Klich (2002), Bensch et al. (2012), and Visagie et al. (2014), respectively, and the analyzes were conducted in the laboratory of mycology and mycotoxins of the Department of Food Science of UFLA.

After incubation, the isolates were further classified according to their macroscopic and microscopic characteristics. For taxonomic identification at the species level, identification keys for the genera Aspergillus and Penicillium and the descriptions of the genus Cladosporium by Bensch et al. (2012) were used.

For the extraction of yeast DNA, colonies were suspended in 500 μL of sterile ultrapure water, and the DNA was extracted using the QIAamp DNA Micro Kit (Qiagen, Germany). For amplification, PCR was performed in a final reaction volume of 25 μL, which contained: 17.2 μL of sterile MilliQ water, 5 μL of MasterMix 5X buffer (Promega), 0.5 μL of dNTP, 0.5 μL of each primer: ITS1 (5' TCCGTAGGTGAACCTGCGG 3') and ITS4 (5' TCCTCCGCTTATTGATATGC 3'), 0.1 μL Taq DNA polymerase (Invitrogen), 0.25 μL MgCl2 and 1 μL DNA. The thermal cycler was programmed for an initial denaturation of 5 min at 95°C, followed by 35 steps of denaturation at 95°C for 1 minute, primer annealing at 55°C for 1 min, and DNA strand extension at 72°C for 1 min. The final extension took place at 72°C for 7 min and the storage temperature was 4°C.

PCR products were analyzed for purity and subjected to denaturing gradient gel electrophoresis (DGGE) using the vertical DCode system (Bio-Rad). The PCR products were applied in 10 μL aliquots, added with 10 μL of running dye (2% bromophenol blue, 2% xylene cyanol, glycerol, and MilliQ water) on a polyacrylamide gel (8% w/v) in TAE buffer (0.5X) with denaturation gradient ranging from 40 to 60% (100% corresponding to 7 M urea and 40% (v/v) formamide). Electrophoresis was carried out at a constant voltage of 150 V for 4 hours, with a constant temperature of 60°C. After electrophoresis, the gels were stained with SYBR-Green I (Molecular Probes1:10,000 v/v), and the image was captured using a transilluminator. Sequencing was performed based on the DGGE bands, where the different and constant bands were excised with the aid of a sterile blade, and the DNA was eluted in sterile ultrapure water. The DNA was reamplified in PCR under the same conditions as above. The products of the second PCR were sequenced by Macrogen and the sequences obtained were aligned with the GenBank database, with the aid of the BLAST software (National Center for Biotechnology Information, Maryland, USA). Multiple sequence alignments were performed using the MEGA-X software with the Clustal-W package. The Hasegawa-Kisinho-Yano nucleotide substitution method was used with a gamma distribution. The maximum Likelihood Tree test and the bootstrap method with 500 bootstrap replications were used for the phylogenetic analysis.

A total of 318 yeast and 41 filamentous fungus isolates were used in the assay for the activity of eight different enzymes, specifically ligninolytic (Widiastuti and Wulaningtyas, 2008), cellulolytic and hemicellulolytic (Kasana et al., 2008), amylolytic (Rempel et al., 2018), and pectinolytic (Ferreira et al., 2013), proteolytic (Scatamburlo et al., 2015), lipolytic (Fuciños et al., 2005) and ammonification (Jana and Roy, 1985) activities. The tests were performed on Petri dishes with solid media containing the appropriate substrate for each enzyme. Each yeast isolate (incubated in YEPG medium at 28°C for 48 hours) was normalized to a concentration of 10⁷ cells/mL in 3 mL of sterile saline (0.9% NaCl), according to the McFarland scale. Petri dishes for each enzyme to be evaluated were inoculated with 10 μL aliquots of the yeast cells at 10⁷. For the filamentous fungi, 6-mm diameter disks from a colony grown for 3 days on MA medium were placed in the center of the Petri dishes. The inoculated media were then incubated at 28 and 25°C for the yeasts and filamentous fungi.

Iodine solution was used to confirm the cellulolytic, hemicellulolytic amylolytic, and the results were classified based on the presence or absence of a clear halo around the yeast and filamentous fungi colonies. Discoloration around the colonies in the presence of Azure-B was used for assessing ligninolytic activity. A solution of 1% hexadecyltrimethylammonium bromide was used for the pectinolytic activity. The lipolytic activity was evaluated by forming a clear, crystallized halo around the colonies, and ammonification activity was assessed based on the formation of a pink halo. Extracellular protease production (Strauss et al., 2001) was determined after 5 days at 30°C in a YPD medium by the presence of a clear zone around the colony.

The enzymatic index (EI) was determined using Equation 1 (Eq 1), which is the ratio of the mean diameter of the degradation halo to that of the colony (Hankin and Anagnostakis, 1975).

The isolates exhibiting the highest EI (>2.0) had the highest extracellular enzymatic activity (Oliveira et al., 2012).

The experimental design was completely randomized; the entire experiment was conducted in triplicate. The efficiency of the composting process was statistically proven and conducted according to an already published article with physicochemical data, by Gaspar et al. (2022).

The population of yeasts and filamentous fungi in the inoculum was evaluated before being added to the compost. The number of yeasts in the EM inoculum was approximately 5.53 log CFU g⁻¹, smaller than the 8.09 log CFU g⁻¹ population in the CI inoculum (Figure 1A). However, filamentous fungi were not observed in any inoculum evaluated (Figure 1B). The yeasts found in each inoculum were sequenced, identified, and presented in Table 3, together with their respective quantifications in each inoculum.

Monitoring the different microbial populations present throughout the composting process is extremely important to determine the rate of particle biodegradation and the quality of the final product (Jurado M. et al., 2014; Oviedo-Ocaña et al., 2019). After the addition of inoculants in the composters and over the time of the composting process, it was observed that the total yeast population decreased during the process and became null in the cooling and maturation phases, with the highest values of diversity and sizes obtained during the thermophilic phase (Table 5). In the thermophilic phase, the yeast population reached 6.4, 6.56, and 6.63 log CFU g⁻¹ for the control, EM, and CI treatments, respectively (Figure 1C). Different yeast species are typically associated with the beginning of the composting process due to the environmental factors favorable to these microorganisms, such as the high availability of O₂ and low pH (Langarica-Fuentes et al., 2014; Nakasaki and Hirai, 2017). The population of filamentous fungi was higher than the yeasts at the beginning of the process, mainly during the thermophilic phase. During this phase, a larger filamentous fungi population with 6.38, 6.39, and 6.50 log CFU g⁻¹ in control, EM, and CI treatments, respectively (Figure 1D), may be present because these microorganisms produce spores, and consequently are more resistant to high temperatures (Nakasaki and Hirai, 2017; Wang Y. et al., 2022). During the cooling and maturation phases, the population of filamentous fungi increased and remained constant, which may result from favorable environmental factors and waste degradation (Meng et al., 2018; Huzairi et al., 2022).

Several yeasts and filamentous fungi were identified throughout the composting process in all treatments and are shown in Table 4 and Figure 2.

At the beginning of the composting process, a substantial number of isolates and species diversity were observed in both yeast and filamentous fungi in the CI treatment. However, throughout the process, this number was higher and constant in the EM treatment, showing higher microbial activity and efficiency in the transformation of organic matter in the EM and IC treatments compared to the control treatment (Figure 3).

Inoculation of compost with starter cultures has several benefits related to microbial diversity, a reduction in composting time, and the highest and fastest rate of particle degradation (Manu et al., 2017; Jia et al., 2021; Huzairi et al., 2022). Yeasts perform essential roles in the composting process, P. kudriavzevii, for example, acts on the degradation of organic acids, which consequently causes an increase in pH and promotes bacterial growth that contributes significantly to the waste degradation (Ngoc et al., 2015). In general, filamentous fungi are considered fundamental in the degradation of recalcitrant fractions of agricultural residues to be composted (Sundberg et al., 2011; Huzairi et al., 2022).

Although some studies have shown that microbial inoculation can improve composting, mainly by reducing the processing time, increasing the enzymatic activity, and several other factors can affect the process (Gou et al., 2017; Singh et al., 2021), some microorganisms naturally present in the initial residues can be competitive or resistant to different environmental changes (López-González et al., 2015b; Huzairi et al., 2022). Nakasaki and Hirai (2017) observed that, despite stimulating the naturally occurring microbiota, the use of P. kudriavzevii as an organic waste compost inoculant did not positively influence the quality of the final compost.

The criteria used to select for persistent microorganisms in each treatment was the detection of the same specie at 2 or more sampling times during the composting process because it shows that he appeared more than once and in different phases. Table 5 shows the yeast and filamentous fungi species identified from each treatment.

Among the yeast species isolated from the compound inoculated with efficient microorganisms (EM) and commercial inoculum (CI), the genus Pichia corresponded to 66.6 and 40%, respectively. While the genera Candida and Blastobotrys were exclusive to treatments inoculated with EM, with 25.9 and 7.5% respectively. On the other hand, the genera Hanseniaspora and Wickerhanomyces corresponded to 33.3 and 26% of the compost with commercial inoculum (CI). Among the species of filamentous fungi, those of the genus Aspergillus were the most frequent, also occurring in the presence of the genera Penicillium, Rhizopus, and Cladosporium. In the control compost, the appearance of yeasts I. orientalis and P. kudriavzevii was constant between the composting phases, as well as filamentous fungi of the genus Aspergillus and Penicillium.

Each composting process has unique characteristics (Niu et al., 2019), and some yeast species, such as Blastobotrys adeninivorans, Blastobotrys raffinosifermentans, and Bipolaris sorokiniana, have not been previously reported in other composting processes. As an organic compost, several microorganisms, even phytopathogenic organisms, may be present in the early stages, such as B. sorokiniana (Song et al., 2018; Puspitaloka et al., 2022). However, this yeast was found only in the control treatment and during the thermophilic phase (Table 5), after which it was eliminated in the final composting stages.

The identified filamentous fungi mainly represented the Aspergillus, were present in both treatments and predominated at all sampling times (Table 5). This genus of filamentous fungi produces spores that are resistant to high temperatures (Wang S.-K. et al., 2022), and although it was widely detected throughout the composting process, it may not have participated in all stages since what could be present were the spores. In a previous study on composting organic waste, species of the genus Aspergillus were also detected (Zhao et al., 2017; Roca-Barcelo et al., 2020). Aspergillus and Penicillium species were detected during the composting of wheat straw for mushroom production (Zhang et al., 2014). Penicillium species were detected in composting from white wine distillers' grains (Yu et al., 2021). In composting organic waste, Di Piazza et al. (2020) reported having identified species of Penicillium spp., Rhizopus spp., and Aspergillus spp. Different fungi species have been isolated from different composting methods, including A. fumigatus, Chaetomonium thermophilum, Humicola grisea, Humicola insolens, Mucor pusilus, Talaromyces duponti, Scytalidium thermophilium (Yim et al., 2019). Species of the genus Rhizopus spp. were also reported by Huzairi et al. (2022). Additionally, filamentous fungi belonging to the genus Aspergillus have been detected during composting with bovine manure and construction wastes (Holman et al., 2016).

The genera Candida and the species A. fumigatus have also been detected during lignocellulosic material compost (Silva et al., 2009; Jurado M. et al., 2014). P. farinosa was the most representative yeast, as it was identified during the thermophilic phase of all the treatments (Table 5) and in the EM inoculum. When analyzing the lignocellulosic composting process, López-González et al. (2015a) identified the yeast species Candida railenenses, Candida sequanensis, Pichia fermentans and the filamentous fungi species A. fumigatus, P. solitum, and P. citrinum during the early composting phases.

In metaproteomic studies with two types of raw materials prepared by mixing commercial rabbit food with cooked rice at ratios of 7:3 and 5:5, the most abundant genera were Saccharomyces, Candida, and Schizosaccharomyces; species of the genera Kluyveromyces, Pichia, and Aspergillus were also detected (Nakasaki and Hirai, 2017).

Alternaria, Aspergillus, Candida, Cladospodium, Fusarium, Gibellulopsis, Ochrocladosporium, Penicillium, Plectosphaerella, and Scopulariopsis were also isolated previously by Jurado M. et al. (2014).

Microbial enzymes are essential in the composting process (Huzairi et al., 2022), and according to Di Piazza et al. (2020), the mesophilic phase of the composting process is characterized by the most intense metabolic activity and the highest enzyme synthesis.

All microorganisms in the study were tested for enzyme activities and those with EI >2.0 were described and correlated with the taxonomy in Table 5.

The isolated yeasts showed a higher diversity of enzymatic activity compared to filamentous fungi among the inoculated treatments during the initial 40 days of the composting process. Among the yeasts, 7.84% have the ligninolytic capacity, 5.88% cellulolytic, 11.80% hemicellulolytic, 9.80% amylolytic, 5.88% proteolytic, 35.3% lipolytic, and 23.50% ammonification capacity.

At the end of composting, filamentous fungi predominated and yeasts were found to be absent in the culture medium. None of the isolated filamentous fungi exhibited the lipolytic, hemicellulolytic, amylolytic, or lipolytic activities tested. The filamentous fungi showed 33.0% cellulolytic, 31.90% pectinolytic, 28.60% ammonification, and 6.50% proteolytic capacity.

Many species showed some enzymatic activity, but only a few were potential enzyme producers, such as yeast species of the Pichia and Issatchenkia genera, with an enzymatic index of lignolytic, cellulolytic and proteolytic activities >4.0, and filamentous fungi of the genus Aspergillus, with an enzymatic index for cellulolytic activity reaching 5.0. Among isolates of the same species, differences in enzyme production were found, showing different biotechnological potentials and differences in the same species. The isolates with high potential for enzyme production (i.e., EI > 2.0) are presented in Table 6. Issatchenkia orientalis was the yeast species with the highest enzymatic potential for ligninolytic activity (EI = 4.2). Regarding the cellulolytic and amylolytic activities, C. tropicalis stood out, with EIs of 4.7 and 2.8, respectively. P. farinosa presented EIs of 2.0, 4.5, and 3.3 for hemicellulolytic, proteolytic, and lipolytic activities, respectively. Aspergillus spp. presented high enzymatic potentials for cellulolytic activity with EIs = 5.0. Thus, the results obtained allow us to identify the I. orientalis, C. tropicalis, P. farinosa, and Aspergillus spp. isolates as promising enzyme producers (Table 7).

In composting, enzymes are important because the biotransformation of organic materials occurs under the action of the enzymatic activity of the microorganisms that can improve and accelerate the degradation processes, directly reflecting on the final product (Zhang and Sun, 2016; Chen et al., 2021). Applying compost with a microorganism with good enzymatic activity can help improve soil fertility (Huzairi et al., 2022). In this sense, the presence of enzyme-producing microorganisms in composting is essential for the degradation of solid waste and, subsequently, when applied in the form of compost, for maintaining soil dynamics and in cycles of nitrogen, carbon, or phosphorus (Deus et al., 2017; Zhang et al., 2021). I. orientalis strains can hydrolyze phytates in the soil, which improves the absorption of phosphorus by plants and, consequently, reduces the concentration of fertilizers added to the soil (Mezeli et al., 2017). Species of the Aspergillus genus can degrade residual herbicides that are extremely dangerous to the environment (Sondhia et al., 2016).

Composting has stood out among the most viable alternatives for the use of waste because it is a sustainable technology that adds value to organic waste (Zhang et al., 2021). Besides, the final product allows for the recycling of organic waste—however, the type of waste in composting changes the microbiota present in each process (Bohacz, 2016; Torrijos et al., 2021; Huzairi et al., 2022).

In general, isolates that present some biotechnologically relevant enzymatic activity has a range of metabolic activities and physiological capabilities (i.e., such as the production of enzymes and improvements in the composting process, including shorter processing time) that make them potentially suitable for a wide variety of applications (Jurado M. M. et al., 2014; Sajid et al., 2022).