A strain of myceliopthora roridum m3 and its application in producing single-cell protein feed from waste watermelon peel by fermentation

Abstract

The application belongs to the field of microbial technology, and provides a strain of Myrothecium roseum M3 and its application in the production of single-cell protein feed from fermented waste watermelon peel. The optimal fermentation conditions are determined through single factor and response surface optimization. The crude protein content of the obtained feed reaches 22.99% to 23.36%, which is increased by more than 16.82% compared with unfermented watermelon peel. The strain can grow well in the medium with watermelon peel as the only substrate, the crude protein content of the obtained single-cell protein feed is greater than or equal to 22.99%, the ratio of essential amino acids to non-essential amino acids EAA / NEAA is greater than or equal to 0.83, which is significantly better than the existing SCP production strains, and can be applied to the field of feed protein production, and also provides a new path for the high-value utilization of watermelon peel.

Description

Technical Field

[0001] This invention belongs to the field of microbial technology, and specifically relates to a strain of Geotrichumphurueaensis M3 and its application in the production of single-cell protein feed from fermented waste watermelon rinds. Background Technology

[0002] The ever-increasing global population has led to a surge in demand for protein, a crucial nutrient essential for cellular structure and metabolic function. However, protein production, especially animal protein, is the most expensive compared to other major components of the human diet (fat and carbohydrates). Currently, the primary source of animal protein is animal husbandry, whose costs depend on feed costs. Therefore, feed costs ultimately become the main factor influencing protein production costs. Low-cost, sustainable single-cell protein (SCP) feed has attracted widespread attention from industry and academia. SCP feed products can be produced using microorganisms from various sources utilizing agricultural waste and industrial byproducts. SCP feed production not only transforms low-cost organic materials into high-value products but also reduces the environmental burden caused by the indiscriminate disposal of agricultural waste and industrial byproducts.

[0003] Fruit by-products are an inexpensive and sustainable source of high-value bioactive substances in agricultural waste. Therefore, valuing fruit by-products is considered an important measure to reduce the harmful environmental impact of food waste, recover value-added compounds, and provide new income streams.

[0004] Watermelon (Citrullus lanatus), a widely cultivated cucurbitaceous crop globally, has a rind that accounts for approximately 30% of the fruit's total weight. A large amount of watermelon rind is directly discarded as a processing byproduct, causing serious waste of biomass resources and environmental pollution. Converting it into high-value-added SCP (Survival of the Fittest) feed would have significant economic and environmental benefits.

[0005] Currently, some studies have attempted to convert watermelon rinds into SCP feed using microbial fermentation technology. Chinese patent CN103695324A discloses a method for producing SCP feed using waste watermelon rinds as raw materials and Angel Yeast for solid-state fermentation. After fermentation, the crude protein content increased from 10.5% to 25.6%. In addition, the academic literature "Solid-state fermentation of waste watermelon rinds to produce single-cell protein" also reported a similar Angel Yeast solid-state fermentation process. The above-mentioned existing technologies show that it is technically feasible to produce SCP feed using watermelon rinds through fermentation, but there are still the following technical problems that need to be solved, such as (1) Although the Angel Yeast used in the existing technology (such as CN103695324A) is a commercial product, the genetic background of its specific strains is unclear and its performance parameters are not disclosed. There may be performance differences between different batches of products, and the stability and repeatability of the fermentation process are difficult to guarantee, which is not conducive to the standardization and quality control of industrial production. (2) The existing technology only focuses on the increase of crude protein content and does not involve a systematic evaluation of the amino acid composition and nutritional quality of the feed. However, the nutritional value of feed protein depends not only on the crude protein content, but also on the balance of its amino acid composition - especially the types and contents of essential amino acids, and whether the EAA / NEAA ratio meets the high-quality protein standards recommended by FAO / WHO. Using crude protein content as a single evaluation indicator alone is not enough to fully reflect the actual feed value of SCP feed. (3) There are no reports on the improvement of the amino acid profile of watermelon rind fermented SCP feed in the existing technology, especially the improvement of the content of essential amino acids such as lysine, valine, and leucine, which are crucial for animal growth and development. (4) There are no reports in the existing technology of using Geotrichum phurueaensis for the fermentation of watermelon rind to produce SCP feed. Although Geotrichum candidum, which belongs to the same genus, has been reported to be used for the production of SCP from various industrial and agricultural by-products such as orange peel extract, wastewater, and distiller's grains, Geotrichum candidum and Geotrichum phurueaensis are different species, and their substrate utilization spectrum, metabolic characteristics and fermentation characteristics are significantly different and cannot be simply compared. Whether *Geotrichum proteus* can effectively utilize the nutrients in watermelon rind for growth, reproduction, and protein accumulation, and the amino acid composition characteristics of its fermentation products, are not addressed in any existing technical information or teachings. (5) Even for known strains reported for use in SCP production, their fermentation effect on the specific substrate of watermelon rind is significantly uncertain. The fermentation performance of well-known SCP production strains such as *Geotrichum candida* and *Candida utilis* on watermelon rind substrates has not been systematically studied or compared. The differences in crude protein accumulation capacity and amino acid composition characteristics of different strains on watermelon rind substrates are a blank area in the existing technology.Therefore, screening out a specific strain from among many known microorganisms that can achieve efficient protein accumulation and excellent amino acid composition on watermelon rind substrate is a technical problem that urgently needs to be solved in this field.

[0006] Therefore, there is an urgent need in this field to develop a method for producing fermented SCP feed from watermelon rinds that has a clear strain source, controllable performance, and can simultaneously improve the crude protein content and amino acid nutritional quality. Summary of the Invention

[0007] In view of the above, the present invention provides a strain of Geotrichum phurueaensis M3 and its application in the production of single-cell protein feed from fermented waste watermelon rinds.

[0008] This invention isolates and screens a strain of *Geotrichum prostrata* M3 from jackfruit peel, which was deposited on November 13, 2025, at the China General Microbiological Culture Collection Center (CGMCC), with accession number CGMCC No. 42318. This strain can grow and reproduce well in a culture medium with watermelon rind as the sole substrate, and converts the nutrients in the watermelon rind into microbial cells rich in protein and amino acids through semi-solid fermentation.

[0009] To achieve the objectives of this invention, this invention provides the application of Geotrichum phurueaensis strain M3 in the production of single-cell protein feed from fermented waste watermelon rinds. The Geotrichum phurueaensis strain M3 has the accession number CGMCC No. 42318; this strain is deposited at the China General Microbiological Culture Collection Center, address: No. 3, Courtyard 1, Beichen West Road, Chaoyang District, Beijing, on November 13, 2025.

[0010] The present invention also provides a method for producing single-cell protein feed by fermenting waste watermelon rinds using the above-mentioned Geotrichum phurueaensis M3 strain, comprising the following steps:

[0011] (1) The Geotrichum phurueaensis strain M3 was activated and expanded to obtain a seed culture, and then the seed culture was diluted with sterile water to obtain a bacterial suspension; the preservation number of Geotrichum phurueaensis M3 is CGMCCNo.42318;

[0012] (2) The waste watermelon rinds were pretreated to obtain a fermentation substrate;

[0013] (3) Inoculate the bacterial suspension obtained in step (1) into the fermentation substrate obtained in step (2) and carry out fermentation;

[0014] (4) After fermentation, the fermentation products are dried and crushed to obtain single-cell protein feed.

[0015] To further explain, the seed liquid preparation method in step (1) is as follows: the Proteus mirabilis M3 strain is inoculated into PDA liquid culture medium and cultured at 28-32 ℃ and 150-200 r / min for 48-72 h to obtain the seed liquid.

[0016] To further explain, the inoculation amount of the bacterial suspension in step (3) is 1.5% to 3.5% (V / W), the fermentation temperature is 25 to 37 ℃, the fermentation time is 48 h, and the shaking speed is 160 to 180 r / min.

[0017] To further explain, the drying temperature in step (4) is 50-60 °C, and the product is dried to constant weight.

[0018] To further clarify, no external carbon source or nitrogen source is added to the fermentation substrate.

[0019] The present invention also provides a single-cell protein feed, which is prepared by the method described above.

[0020] To further clarify, the crude protein content in the single-cell protein feed is ≥22.99%.

[0021] To further clarify, the ratio of essential amino acids to non-essential amino acids (EAA / NEAA) in the single-cell protein feed is ≥0.83.

[0022] To further clarify, the single-cell protein feed contains at least 16 amino acids, of which the content of lysine, valine, leucine, histidine, aspartic acid and glutamic acid is not less than 0.50%; and the content of the other 15 amino acids in the feed, except for arginine, is higher than that in the unfermented watermelon rind substrate.

[0023] The present invention has the following beneficial effects:

[0024] The *Dendrobium pruinii* M3 used in this invention is a strain isolated from jackfruit peel. Its taxonomic position was confirmed by morphological and 18S rDNA molecular identification, and it has been deposited in CGMCC (accession number CGMCC No. 42318). Compared with commonly used commercial microbial agents in the prior art (such as Angel Yeast), this strain has a clear origin, a well-defined genetic background, and more stable and controllable fermentation performance, which is undoubtedly beneficial for standardization and quality control in subsequent industrial production.

[0025] At the application level, while there are reports of using *Geotrichum candida* for fermenting SCPs from substrates such as orange peel and wastewater, the use of *Geotrichum prouecifolium*, as a different species, for fermenting watermelon rinds to produce SCP feed was not publicly disclosed prior to this invention. This research team's new application study can be said to fill a gap in the application of this species in the field of SCP feed production.

[0026] More noteworthy is the nutritional quality of the feed itself. This invention is the first to systematically analyze the composition of 16 amino acids in SCP feed. The results show that after fermentation, the total amount of essential amino acids increased from 1.93% to 3.61%, and the amount of non-essential amino acids increased from 2.77% to 4.35%. The EAA / NEAA ratio increased from 0.70 to 0.83, exceeding the FAO / WHO recommended standard for high-quality protein (>0.60). Specifically, aspartic acid increased by 0.4 percentage points, lysine by 0.36 percentage points, and glutamic acid and leucine by 0.31 percentage points respectively. The contents of lysine, valine, leucine, histidine, aspartic acid, and glutamic acid all exceeded 0.50%. These amino acids each have their own biological significance—aspartic acid helps maintain intestinal health and enhance immunity, lysine promotes animal growth and improves feed utilization, glutamic acid improves feed palatability, and leucine plays a role in maintaining nitrogen balance and normal growth. In contrast, existing technologies, such as Chinese Patent Publication No. CN103695324A, only focus on crude protein content and do not involve a systematic evaluation of amino acid composition. The research of this invention takes a step forward in the depth of evaluation of feed nutritional quality.

[0027] Furthermore, this invention further validated the fermentation advantage of *Geotrichum pullulans* M3 on watermelon rind substrate by conducting parallel comparative experiments with known SCP-producing strains (Angel Yeast, *Geotrichum candida*, and *Candida utilis*) under identical fermentation conditions. Experimental results showed that the crude protein content of the feed fermented with *Geotrichum pullulans* M3 was 23.36%, significantly higher than that of *Angel Yeast* (16.38%), *Geotrichum candida* (15.34%), and *Candida utilis* (16.81%). Regarding the total essential amino acid content, M3 was 3.61%, which was 0.80, 0.13, and 0.74 percentage points higher than that of *Angel Yeast* (2.81%), *Geotrichum candida* (3.48%), and *Candida utilis* (2.87%), respectively. Of particular note is that, even compared to the related species *Geotrichum candida*, M3 exhibited 8.02 percentage points higher crude protein content on watermelon rind substrate, and its EAA / NEAA ratio (0.83) was significantly superior to *Geotrichum candida* (0.64). These results indicate that not all *Geotrichum* strains or known SCP strains can achieve the same results in watermelon rind fermentation; strain M3 demonstrates unique fermentation performance on this specific substrate and produces unexpected technical effects.

[0028] From the perspective of process optimization, this invention investigated the effects of fermentation time, inoculum size, temperature, and rotation speed on crude protein content through single-factor experiments, and established a four-factor, three-level regression model using Box-Behnken response surface methodology. The optimal fermentation conditions were determined to be 2.5 days, 1.3% inoculum size, 31 ℃ temperature, and 154 rpm rotation speed. Under these conditions, the crude protein content reached 23.36 ± 0.73%, an increase of over 16.82% compared to unfermented watermelon rind. Analysis of variance showed an F-value of 4.49, a P-value less than 0.0040, and no significant lack-of-fit term (P = 0.4999), indicating good model fitting accuracy and providing reliable guidance for process optimization.

[0029] It is also worth mentioning that the isolation source of strain M3 is quite unique. The known *Geotrichum prostrata* strain LYSM5T was isolated from soil in a Thai terrestrial forest, while the strain of this invention was isolated from jackfruit peel. This different isolation source may suggest that M3 has some unique metabolic characteristics, which may be related to its good performance in watermelon peel fermentation. Attached Figure Description

[0030] Figure 1 Colony diagram (A) and microscopic observation (B) of Geotrichum phurueaensis strain M3.

[0031] Figure 2 Phylogenetic tree diagram of Geotrichum phurueaensis strain M3 based on 18S rDNA sequence.

[0032] Figure 3 The figures show the effects of single-factor experiments on the crude protein content in SCP feed. In the figure, A is the effect of fermentation time, B is the effect of inoculum size, C is the effect of fermentation temperature, and D is the effect of fermentation revolutions.

[0033] Figure 4 The diagram shows the normal probability distribution of the residuals in the response surface optimization experiment (A), the distribution of the residuals and the predicted values ​​of the equation (B), and the distribution of the predicted values ​​and the actual values ​​(C).

[0034] Figure 5 The response surface diagrams represent the interactions between any two factors in the response surface optimization experiment. Among them, A is the response surface diagram of the interaction between inoculum amount and time, B is the response surface diagram of the interaction between temperature and time, C is the response surface diagram of the interaction between revolutions and time, D is the response surface diagram of the interaction between temperature and inoculum amount, E is the response surface diagram of the interaction between revolutions and inoculum amount, and F is the response surface diagram of the interaction between revolutions and temperature.

[0035] The Geotrichum phurueaensis M3 isolated and screened from jackfruit peel in this invention has been deposited at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 42318, located at No. 3, Courtyard 1, Beichen West Road, Chaoyang District, Beijing, on November 13, 2025. Detailed Implementation

[0036] All features disclosed in this specification, or all steps in all disclosed methods or processes, may be combined in any way, except for mutually exclusive features and / or steps.

[0037] Unless otherwise stated, each feature disclosed in this specification (including any appended claims and abstract) is merely one example of a series of equivalent or similar features.

[0038] Example 1: Isolation and identification of Dietaria proueta strain M3

[0039] In this embodiment, a pure culture strain was isolated and purified from jackfruit peel using the dilution plating method. The strain was identified as Geotrichum phurueaensis M3 by morphological observation and 18S rDNA molecular identification.

[0040] This example demonstrates the isolation and identification of *Geotrichum prouetum* strain M3.

[0041] 1. Preparation of culture medium

[0042] 1.1 PDA liquid culture medium

[0043] Weigh 2.6 g of PDB (composition (g / L): potato extract powder 6.0; glucose 20.0) and place it in a 250 mL Erlenmeyer flask. Add 100 mL of distilled water and sterilize at 121 °C for 20 min.

[0044] 1.2 PDA solid culture medium

[0045] Weigh 11.5 g of PDA nutrient agar (composition (g / L): potato extract 6.0; glucose 20.0; agar 20.0; pH 5.6±0.2, 25 ℃) and place it in a 500 mL Erlenmeyer flask. Add 250 mL of distilled water and sterilize at 121 ℃ for 20 min.

[0046] 1.3 YPD liquid culture medium

[0047] Weigh 1 g of Saccharomyces cerevisiae extract, 2 g of tryptone, and 2 g of glucose into a 250 mL Erlenmeyer flask, add 100 mL of distilled water, and sterilize at 121 °C for 20 min.

[0048] 2. Isolation and Identification of Strains

[0049] 2.1 Strain Isolation

[0050] Take a small amount of jackfruit peel and place it in a 20 mL sterile bottle. Add 5 mL of sterile water and shake until the solution becomes cloudy. Then, take 100 μL and spread it onto PDA solid medium and incubate at 28 ℃ for 48 h. Next, pick a single colony and streak it onto PDA solid medium, and incubate at 28 ℃ for 48 h. Continue to pick single colonies and streak them until all colonies on the plate have the same morphological characteristics.

[0051] 2.2 Strain Identification

[0052] 2.2.1 Morphological identification

[0053] The colonies of this strain appear white, round, with irregular edges, a rough, opaque surface, and wrinkles in the center on PDA solid medium plates. They are easy to pick up and do not produce pigment. Figure 1 A). Under an optical microscope (40x), the conidia of this strain appear as short rods ( Figure 1 B).

[0054] 2.2.2 18S rDNA Identification

[0055] First, the genome of this strain was extracted using a genome extraction kit. Then, 18S rDNA was amplified by PCR using universal fungal primers: NS1 and NS6 (NS1: 5′-GTAGTCATATGCTTGTCTC-3′; NS6: 5′-GCATCACAGACCTGTTATTGCCTC-3′). The amplified products were then verified by agarose gel electrophoresis. Finally, the PCR products were sent to Sangon Biotech (Shanghai) Co., Ltd. for sequencing. The sequences were then compared and analyzed using the BLAST tool in the NCBI database.

[0056] The 18S rDNA sequence of this strain is shown in SEQ ID NO.1:

[0057] TAGCTAAGTATGCTTAGTACGGCGAGTGAGCGGCAAAAGCTCAAATTTGAAATCGGCCAACAGGTCGAGTTGTAATTTGTAGATTGTACCTTGAGAGCGGATCAAAGTCTGTTGGAACACAGCGCCTTAGAGGGTGACA GCCCCGTAGGATCTATTCTCATTGTAAGGTGCTTTCGAAGAGTCGAGTTGTTTGGGAATGCAGCTCTAAGTGGGAGGTAAATTCCTTCTAAAGCTAAATATTGACGAGAGACCGATAGCGAACAAGTACTGTGAAGGAAA GATGAAAAGCACTTTGAAAAGAGAGTGAAAAAGTACGTGAAATTGTTAAAAGGGAAGGGTATTGAATCAGACGTGGTGCTGTTGTTCAACTATGTTTTGGCATGGTGTACTCAGCAGTACTAGGCCAAGGTGGGGTGTT TGGGAGTGAAAAAGAGATGAGAATGTAGCTCTTCGGAGTGTTATAGCTCATTTTCATAGCTCCTCAGGCGCCTCAGGACTGCGCTTCGGCAAGGACTTTGGCATAATGATTCTATACCGCCCGTCTTGAACCACGGACCA

[0058] The sequencing results were compared and analyzed using the BLAST tool in the NCBI database. The results showed that the 18S rDNA sequence of this strain had high similarity to several *Geotrichum phurueaensis* strains, and its NS sequence was also similar to that of *Geotrichum phurueaensis* in the phylogenetic tree (e.g.,...). Figure 2 As shown, the strain clusters at the same node. Based on the morphological characteristics of the strain, it was identified as Geotrichum phurueaensis and named Geotrichum phurueaensis M3.

[0059] This strain was deposited on November 13, 2025, at the China General Microbiological Culture Collection Center (CGMCC) with accession number CGMCC No. 42318. The deposit address is No. 3, No. 1 Beichen West Road, Chaoyang District, Beijing.

[0060] Example 2: Preparation of Ditheria prouetes M3 suspension

[0061] This embodiment provides a method for preparing a suspension of Ditomyces purulentus M3, providing a stable liquid strain for subsequent fermentation experiments.

[0062] Five to six isolated and purified *Geotrichum proteus* M3 single colonies were inoculated into 100 mL of PDA liquid medium and cultured at 30 ℃ and 160 rpm for 3 days. Then, a certain amount of the culture medium was taken into a sterile tube and diluted with sterile water to prepare the bacterial suspension for fermentation.

[0063] Five to six isolated and purified Proteus mirabilis M3 strains were inoculated into 100 mL of PDA liquid medium and cultured at 28–32 °C and 150–200 r / min for 48–72 h to obtain seed culture. Then, a certain amount of culture medium was taken into a sterile tube and diluted with sterile water to serve as a bacterial suspension for fermentation.

[0064] Example 3: Effect of different fermentation times on crude protein content in SCP feed

[0065] This embodiment investigated the effect of different fermentation times on the crude protein content in SCP feed through single-factor experiments, and determined the optimal fermentation time to be 2 days.

[0066] 1. Watermelon rind culture medium

[0067] Clean, fresh watermelon rinds are directly crushed into a paste, and 200-300 g are dispensed into 500 mL Erlenmeyer flasks. The flasks are then sterilized at 121 °C for 20 min to obtain the fermentation substrate for later use.

[0068] 2. Effects of different fermentation times on crude protein content in SCP feed

[0069] Add 5 mL of *Geotrichum prouetum* M3 suspension (OD) to 200 g of watermelon rind culture medium. 600 The concentration of crude protein was 0.845. Fermentation was carried out at 30℃ and 160 rpm for 2, 3, 4, 5, and 6 days, respectively. The resulting fermentation broth was then dried at 50–60℃ to constant weight, pulverized, and the SCP feed product was obtained. The crude protein content in the SCP feed was determined according to GB / T 6432-2018 (7.2) Determination of Crude Protein in Feeds by Kjeldahl Method.

[0070] The results are as follows Figure 3 As shown in Figure A, the crude protein content gradually decreased with the extension of fermentation time of *Dendrobium pullulans* M3, indicating that the fermentation time of *Dendrobium pullulans* M3 has a significant impact on the crude protein content. This may be because the longer the fermentation time, the fewer usable nutrients are available in the watermelon rind, and the fungi gradually die due to nutrient deficiency. Therefore, the suitable fermentation time is 2 days.

[0071] Example 4: Effect of different inoculation amounts on crude protein content in SCP feed

[0072] Based on the optimal fermentation time, this embodiment investigated the effect of different inoculum amounts on the crude protein content in SCP feed through single-factor experiments, and determined the optimal inoculum amount to be 1.5%.

[0073] Based on the optimal fermentation time (2 days), 3, 4, 5, 6, and 7 mL of *Dendrobium prostrata* M3 suspension (OD200) were added to 200 g of watermelon rind culture medium, respectively. 600 The inoculum amounts were 1.5%, 2.0%, 2.5%, 3.0%, and 3.5%, respectively, and the mixture was fermented at 30 ℃ and 160 rpm for 2 days. The resulting fermentation broth was then dried at 55 ℃ for 48 h, pulverized, and the SCP feed product was obtained. The crude protein content in the SCP feed was determined according to GB / T 6432-2018 (7.2) "Determination of Crude Protein in Feed - Kjeldahl Method".

[0074] The results are as follows Figure 3 As shown in Figure B, the crude protein content was highest (30.17%) when the inoculum amount of *Geotrichum prunella* M3 was 1.5%. Increasing the inoculum amount of *Geotrichum prunella* M3 decreased the crude protein content, but the crude protein content did not change significantly across the four higher inoculum amounts. This may be because at lower inoculum amounts, *Geotrichum prunella* M3 could utilize the limited nutrients in the watermelon rind for rapid growth and reproduction in a shorter time. However, as the inoculum amount increased, *Geotrichum prunella* M3 could only maintain basic survival with the limited nutrients, and could no longer reproduce rapidly in large quantities, resulting in a decrease in crude protein content that remained relatively stable across the four higher inoculum amounts. Therefore, the optimal inoculum amount for *Geotrichum prunella* M3 is 1.5%.

[0075] Example 5: Effect of different fermentation temperatures on crude protein content in SCP feed

[0076] Based on the optimal fermentation time and inoculum amount, this embodiment investigated the effect of different fermentation temperatures on the crude protein content in SCP feed through single-factor experiments, and determined the optimal fermentation temperature to be 34 ℃.

[0077] Based on the optimal fermentation time (2 days) and inoculum size (1.5%), 3 mL of *Dendrobium prostrata* M3 suspension (OD500) was added to 200 g watermelon rind culture medium. 600The concentration of crude protein was 0.879. Fermentation was carried out for 2 days at temperatures of 25, 28, 31, 34, and 37 °C and a rotation speed of 160 rpm. The resulting fermentation broth was then dried at 55 °C for 48 h, pulverized, and the SCP feed product was obtained. The crude protein content in the SCP feed was determined according to GB / T 6432-2018 (7.2) "Determination of Crude Protein in Feed - Kjeldahl Method".

[0078] The results are as follows Figure 3 As shown in Figure C, the crude protein content reached its highest value (28.87%) at a fermentation temperature of 34 ℃. This indicates that 34 ℃ is the optimal temperature for the growth and reproduction of *Dendrobium pullulans* M3, while temperatures below or above 34 ℃ are detrimental to its growth and reproduction. This may be because excessively low or high temperatures lead to the inactivation of enzymes in *Dendrobium pullulans* M3, thus affecting its absorption of nutrients. Therefore, 34 ℃ was selected as the optimal fermentation temperature for *Dendrobium pullulans* M3.

[0079] Example 6: Effect of different fermentation revolutions on crude protein content in SCP feed

[0080] Based on the optimal fermentation time, inoculum size, and temperature, this embodiment investigated the effect of different fermentation speeds on the crude protein content in SCP feed through single-factor experiments, and determined the optimal fermentation speed to be 160 rpm.

[0081] Based on the optimal fermentation time (2 days), inoculum size (1.5%), and temperature (34 ℃), 3 mL of *Dendrobium prostrata* M3 suspension (OD) was added to 200 g of watermelon rind culture medium. 600 The concentration of crude protein was 0.865. Fermentation was carried out at 34 ℃ and rotation speeds of 120, 140, 160, 180, and 200 rpm for 2 days. The resulting fermentation broth was then dried at 55 ℃ for 48 h, pulverized, and the SCP feed product was obtained. The crude protein content in the sample was determined according to GB / T 6432-2018 (7.2) Determination of Crude Protein in Feed - Kjeldahl Method.

[0082] The results are as follows Figure 3As shown in Figure D, as the fermentation speed increased from 120 rpm to 200 rpm, the crude protein content first increased and then decreased, reaching its highest values ​​of 21.61% and 21.64% at 160 rpm and 180 rpm, respectively. This result indicates that when the speed is below 160 rpm, the cells do not have sufficient contact with nutrients, resulting in slow cell growth; while when the speed is above 180 rpm, the cells may rupture due to shear force, leading to cell death. Therefore, the suitable fermentation speed is 160–180 rpm, but considering energy costs, choosing 160 rpm is relatively appropriate.

[0083] Example 7: Response surface methodology optimization experiment for fermentation of watermelon rind by Diplostomum prouetum M3

[0084] This embodiment employs the Box-Behnken response surface methodology, using fermentation time, inoculum size, temperature, and rotational speed as factors, and crude protein content as the response value, to establish a four-factor, three-level regression model. The optimal fermentation conditions were determined through analysis of variance and response surface methodology, and a verification experiment was conducted.

[0085] A four-factor, three-level response surface methodology (RSM) experiment was designed using the Box-Behnken (BBD) design method in Design-Expert 13 software. The RSM factors were fermentation time (A), inoculum size (B), temperature (C), and rotational speed (D), with crude protein content (Y) as the response value. The experimental levels and coefficients are shown in Table 1, and the experimental design and results are shown in Table 2. The experimental results were analyzed using Design-Expert 13 software to confirm the optimal fermentation conditions and predict the optimal values.

[0086]

[0087]

[0088] Using Design-Expert 13 software, a multiple regression analysis was performed on Table 2, yielding the following regression equation: Y = 21.01 + 1.22A + 0.3408B - 2.26C - 0.7183D - 0.3425AB - 0.3800AC - 0.5550AD + 0.0250BC - 0.4300BD + 0.6000CD - 0.4833A 2 -0.1658B 2 +0.1054C 2 -0.5721D 2

[0089] The results of the ANOVA of the BBD response surface regression model are shown in Table 3. The model's F-value is 4.49, and the P-value is <0.0040, indicating that the response surface regression model reached a significant level. This suggests good fitting accuracy, and the approximate response surface model can be used for subsequent optimization design. The lack-of-fit term is 0.4999 > 0.05, indicating it is not significant. Among them, the p-values ​​of the linear terms A (fermentation time) and C (fermentation temperature) are <0.05, indicating a highly significant impact on crude protein content; other factors are not significant. Therefore, based on the F-values, the order of factors affecting crude protein content is: fermentation temperature > fermentation time > fermentation revolutions > inoculum size.

[0090]

[0091] The statistical analysis results of the BBD response surface regression equation error are shown in Table 4. 2 =0.8180, indicating that the model has good correlation; the CV is 5.98% <10%, indicating that the experiment has high reliability and accuracy. Therefore, the results in Table 4 show that the fitted regression equation has good adaptability.

[0092]

[0093] The distribution plots of the uneven normal probability, residuals and equation predicted values, and predicted and actual values ​​are shown below. Figure 4 As shown. By Figure 4 It can be seen that the model fitted using the response surface methodology has good adaptability. Figure 5 A response surface plot reflects the interaction between any two factors. The steeper the surface plot, the more pronounced the interaction between the factors. Figure 5 It can be seen that the interaction term BC has the most significant effect on crude protein content.

[0094] Through response surface methodology (RSM) screening and prediction, the optimal fermentation conditions for increasing crude protein content were determined to be: fermentation time 2.48 days, inoculum size 1.334%, temperature 31.306 ℃, and rotation speed 153.967 rpm, with a predicted crude protein content of 24.058%. Considering practical feasibility, the fermentation conditions were adjusted to: fermentation time 2.5 days, inoculum size 1.3%, temperature 31 ℃, and rotation speed 154 rpm. Three parallel experiments yielded a crude protein content of 23.36 ± 0.73%, which is close to the predicted value. This indicates that the optimal fermentation conditions obtained through RSM optimization have high accuracy, and this method can be used to determine the optimal conditions for increasing crude protein content.

[0095] Example 8: Nutritional Composition Analysis of SCP Feed

[0096] In this embodiment, SCP feed was prepared under response surface optimization conditions. The contents of crude protein, crude fat, crude fiber, neutral detergent fiber, acid detergent fiber, crude ash, moisture, calcium, and total phosphorus were determined according to national standard methods. Unfermented watermelon rind was used as a blank control to comprehensively evaluate the nutritional value of SCP feed.

[0097] SCP feed was prepared under the optimized conditions determined in Example 7 (fermentation time 2.5 days, inoculum size 1.3%, temperature 31 ℃, rotation speed 154 rpm). The content of various nutrients was determined according to the corresponding national standard methods, with unfermented watermelon rind as a blank control. The results are shown in Table 5.

[0098]

[0099] Table 5 shows that compared with unfermented watermelon rind, the SCP feed increased crude protein content by 16.82%, crude fat content by 25.00%, crude fiber content by 41.81%, neutral detergent fiber by 51.60%, acid detergent fiber by 13.27%, crude ash content decreased by 8.14%, moisture content decreased by 24.14%, calcium content increased by 11.43%, and total phosphorus content increased by 7.02%. The change in crude protein content indicates that adding *Dendrobium prusti* M3 can increase the crude protein content of the SCP feed. Based on the changes in crude fiber, neutral detergent fiber, and acid detergent fiber content, it can be inferred that *Dendrobium prusti* M3 has a low utilization rate of cellulose in watermelon rind. This further reminds our team to improve cellulose utilization efficiency through strain domestication or co-fermentation with other cellulose-degrading bacteria.

[0100] Example 9: Amino acid composition analysis of SCP feed

[0101] This embodiment systematically analyzed 16 amino acids in the SCP feed prepared under optimized conditions, and used unfermented watermelon rind as a control to comprehensively evaluate the amino acid nutritional quality of the SCP feed. The results showed that the SCP feed obtained by this invention has excellent amino acid balance, with an EAA / NEAA ratio of 0.83, which is higher than the FAO / WHO recommended standard for high-quality protein.

[0102] According to the methods of GB / T 18246-2019 (3) and GB / T 15400-2018, amino acid profile analysis was performed on SCP feed and unfermented watermelon rind. The results are shown in Table 6.

[0103]

[0104] As shown in Table 6, consistent with unfermented watermelon rind, the amino acid profile of SCP feed contains 16 amino acids, including 7 essential amino acids (EAAs) and 9 non-essential amino acids (NEAAs). The EAAs are isoleucine, valine, lysine, leucine, phenylalanine, threonine, and histidine; the NEAAs are aspartic acid, arginine, glutamic acid, proline, alanine, serine, glycine, tyrosine, and tryptophan.

[0105] Compared to unfermented watermelon rind, SCP feed showed increased levels of 15 out of 16 amino acids (with a slight decrease in arginine). The total essential amino acid (EAA) content increased from 1.93% to 3.61%, and the total non-essential amino acid (NEAA) content increased from 2.77% to 4.35%. The EAA / NEAA ratio increased from 0.70 to 0.83, a ratio higher than the FAO / WHO recommended standard for high-quality protein (EAA / NEAA > 0.60). The EAA / TAA ratio increased from 0.41 to 0.45.

[0106] Among the specific amino acids, aspartic acid saw the largest absolute increase, rising by 0.4%; followed by lysine, rising by 0.36%; and then glutamic acid and leucine, both rising by 0.31%. In terms of absolute content after fermentation, the contents of lysine (0.64%), valine (0.56%), leucine (0.58%), histidine (0.60%), aspartic acid (0.88%), and glutamic acid (1.02%) were all above 0.50%. These amino acids all play a significant role in animal growth and development. Aspartic acid helps maintain intestinal health and enhances immunity; lysine improves growth rate and feed utilization; glutamic acid enhances feed palatability; and leucine maintains normal animal growth and nitrogen balance.

[0107] Example 10: Comparison of the fermentation effects of Ditheria pluvialis M3 and different SCP production strains on watermelon rind

[0108] In this embodiment, under identical fermentation conditions, Proteus mirabilis M3 was compared with known SCP-producing strains in the prior art (Angel Yeast Saccharomyces cerevisiae type I, Geotrichum candida, and Candida utilis) in parallel fermentation to verify the fermentation advantage of strain M3 on watermelon rind substrate.

[0109] 1. Test strains

[0110] The *Dendrobium prouetum* M3 of this application,

[0111] CK1: Angel Yeast Feed Yeast, commercially available (Angel Yeast Co., Ltd.)

[0112] CK2: Geotrichum candidum, commercially available (Shanghai Center for Preservation of Tibetan Medicine).

[0113] CK3: Candida utilis, commercially available (Shanghai Center for Preservation of Biotechnology)

[0114] 2. Preparation of bacterial suspension

[0115] The tested strains *Geotrichum candida*, *Angelica pubescens*, and *Candida utilis* were inoculated into 100 mL of PDA, YPD, and YPD liquid media, respectively, and cultured at 30 ℃ and 160 rpm for 3 days. Then, a certain amount of culture solution was taken into a sterile tube, diluted with sterile water, and the OD of the bacterial suspension was adjusted uniformly. 600 Up to 0.85±0.03.

[0116] 3. Fermentation conditions

[0117] All strains were fermented under identical conditions: fresh watermelon rind was crushed into a paste, and 200 g of the paste was dispensed into 500 mL Erlenmeyer flasks and sterilized at 121 °C for 20 min. After cooling, each strain suspension was inoculated at an inoculum rate of 1.3% (V / W) and fermented at 31 °C and 154 rpm for 2.5 days. After fermentation, each fermentation broth was dried at 55 °C to constant weight and then crushed to obtain the SCP feed sample.

[0118] 4. Detection Indicators

[0119] The crude protein content was determined according to GB / T 6432-2018; the composition of 16 amino acids was determined according to GB / T 18246-2019 and GB / T 15400-2018.

[0120] 5. Experimental Results

[0121] The crude protein content and amino acid composition of the SCP feed obtained from the fermentation of each strain are shown in Table 7.

[0122]

[0123] Table 7 shows that the crude protein content of the feed fermented by *Geotrichum protozoa* M3 was 23.36%, significantly higher than that of feeds fermented by Angel Yeast (16.38%), *Geotrichum candida* (15.34%), and *Candida utilis* (16.81%). Among the 16 amino acids tested, the feed fermented by strain M3 had higher levels of 15, 14, and 8 amino acids, respectively, than those fermented by Angel Yeast, *Candida utilis*, and *Geotrichum candida*. Specifically, the lysine content (0.64%) was 0.10, 0.03, and 0.08 percentage points higher than that of Angel Yeast (0.54%), *Geotrichum candida* (0.61%), and *Candida utilis* (0.56%), respectively; the valine content (0.56%) was 0.12, 0.08, and 0.13 percentage points higher, respectively. Regarding the total essential amino acids, M3 had 3.61%, which was 0.80, 0.13, and 0.74 percentage points higher than Angel Yeast (2.81%), Geotrichum candida (3.48%), and Candida utilis (2.87%), respectively. Lysine is the first limiting amino acid in pig feed and the second limiting amino acid in poultry feed, and its content is a key indicator for evaluating feed protein quality. The above results indicate that under identical fermentation conditions, strain M3 has a significant advantage in lysine enrichment efficiency. Furthermore, the EAA / NEAA ratios for feeds fermented by *Dendrobium pullulans* M3, Angel Yeast, Geotrichum candida, and Candida utilis were 0.83, 0.54, 0.64, and 0.57, respectively. Therefore, the crude protein content and amino acid nutritional quality of feed fermented by *Dendrobium pullulans* M3 are superior to those of the three known SCP-producing strains: Angel Yeast, Geotrichum candida, and Candida utilis. Of particular note is that, even compared to other members of the same genus, *Geotrichum candida*, M3 exhibits higher crude protein and lysine content on watermelon rind substrates. This clearly demonstrates that not all *Geotrichum candida* strains or known SCP strains can achieve the same results in watermelon rind fermentation; the fermentation performance of strain M3 on this specific substrate is unique and cannot be simply inferred from existing technologies.

[0124] In summary, the *Dendrobium prussima* M3 strain and its method for producing single-cell protein feed from fermented waste watermelon rinds provided by this invention have broad industrial application prospects. my country generates a large amount of waste watermelon rinds annually, and this invention provides a new approach for the resource utilization of this agricultural waste. The method is simple, low-cost, and highly operable. The resulting SCP feed has a high crude protein content and a balanced amino acid composition, and can be used as a protein supplement in conjunction with other feeds, demonstrating good economic and social benefits.

[0125] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. The application of Geotrichum phurueaensis strain M3 in the production of single-cell protein feed from fermented waste watermelon rinds, characterized in that... The aforementioned Geotrichum phurueaensis M3, with accession number CGMCC No. 42318, is deposited at the China General Microbiological Culture Collection Center, located at No. 3, Courtyard 1, Beichen West Road, Chaoyang District, Beijing, on November 13, 2025.

2. A method for producing single-cell protein feed by fermenting waste watermelon rinds using Geotrichum phurueaensis strain M3 as described in claim 1, characterized in that, Includes the following steps: (1) The Geotrichum phurueaensis strain M3 was activated and expanded to obtain a seed culture, and then the seed culture was diluted with sterile water to obtain a bacterial suspension; the preservation number of Geotrichum phurueaensis M3 is CGMCCNo.42318; (2) The waste watermelon rinds were pretreated to obtain a fermentation substrate; (3) Inoculate the bacterial suspension obtained in step (1) into the fermentation substrate obtained in step (2) and carry out fermentation; (4) After fermentation, the fermentation products are dried and crushed to obtain single-cell protein feed.

3. The method according to claim 2, characterized in that, The method for preparing the seed liquid in step (1) is as follows: inoculate the Proteus mirabilis M3 strain into PDA liquid culture medium and culture it at 28-32 ℃ and 150-200 r / min for 48-72 h to obtain the seed liquid.

4. The method according to claim 2, characterized in that, The inoculation amount of the bacterial suspension in step (3) is 1.5% to 3.5% (V / W), the fermentation temperature is 25 to 37 °C, the fermentation time is 48 h, and the shaking speed is 160 to 180 r / min.

5. The method according to claim 2, characterized in that, The drying temperature in step (4) is 50-60 °C, and the product is dried to constant weight.

6. The method according to claim 2, characterized in that, No external carbon or nitrogen source is added to the fermentation substrate.

7. A single-cell protein feed, characterized in that, Prepared by the method according to any one of claims 2 to 6.

8. The single-cell protein feed according to claim 7, characterized in that, The crude protein content of the single-cell protein feed is ≥22.99%.

9. The single-cell protein feed according to claim 7, characterized in that, The ratio of essential amino acids to non-essential amino acids (EAA / NEAA) in the single-cell protein feed is ≥0.

83.

10. The single-cell protein feed according to claim 7, characterized in that, The single-cell protein feed contains at least 16 amino acids, of which the content of lysine, valine, leucine, histidine, aspartic acid and glutamic acid is not less than 0.50%.

Images