Drying treatment for sludges of the Chilean salmon farming industry and its potential as an agricultural soil amendment

ABSTRACT

Organic waste generation in aquaculture worldwide requires circular economy approaches. The revalorization of wastes in the Chilean salmon industry could be a potential soil amendment. We investigated the thermo-drying process on salmon sludge, evaluating it as an alternative to reusing in horticulture. The chemical characterization of untreated sludge (wet sludges: WS) and dried sludges (DS) were evaluated with thermogravimetry, IR spectroscopy, and fluorescence. Additionally, the germination index of the DS was determined in L. sativa, R. sativus, and S. lycopersicum. The results show that DS reduces salinity, ammonium content, and pathogens compared to WS. During the thermic process, there were no significant decreases in organic matter, phosphorus, potassium, and humic and fulvic substances. The high salinity in DS inhibited seed germination. This research constituted a novel contribution to studying the DS as a strategy for horticultural production. In the same way, the drying technology for salmon sludge favored the management of wastes, contributing to the circular economy in both the agricultural and aquaculture sectors. However, we recommend considering that excessive incorporation of DS into germination substrates or direct application into the soil could affect the crop’s performance due to its high salinity.

KEYWORDS:

• Chilean salmon farming
• dried sludge
• circular economy
• thermogravimetry
• spectroscopy

Introduction

Industrial waste generation continues to increase and is considered one of the main concerns worldwide. If not properly treated, they can become substances of environmental risk affecting aquatic systems and soil quality, directly impacting food production (Lin et al. 2018). Thus, the industry faces a considerable challenge to supply an increasing food demand sustainably, incorporating the concept of circular economy in its process, where the generation of new added value to organic waste and industrial waste resources is enhanced (Lozano-Muñoz et al. 2021; Rojas et al. 2022).

From this perspective, research on industrial waste treatment with potential reuse in agricultural systems has been developed (Zhu et al. 2018; Yaashikaa et al. 2022). Specifically, it focused on organic waste pre-treatment to reduce the potential environmental risks and the accumulation of pollutants in the soil system (Wang et al. 2016). In this regard, the novelty is identifying new nutrient sources for generating low-cost organic fertilizers, minimizing the release of organic matter and the amounts of waste to the environment due to its treatment (Freitas et al. 2021).

In this context, the Chilean aquaculture industry is among the world’s largest producers of salmonids, showing an exponential growth projection of 47.6% by 2030 (FAO 2022). This growth implies a constant increase in waste production from the recirculation aquaculture systems (RAS) (Louvado et al. 2021). According to Badiola et al. (2012), RAS is considered one of the most sustainable fish production methods with responsible and adequate management. The process of RAS achieves an effective organic matter elimination and avoids the accumulation of nutrients in the culture ponds (Martins et al. 2010). In turn, recirculation would favor the presence of fluorescent substances as humic and fulvic-like substances associated with fish feeding, metabolic wastes, the constant inflow of water from outside, and the system’s water treatment process (Wang et al. 2021). These systems have an advantage over flow-through systems by reducing the environmental impact due to the removal of solid waste through sedimentation and successive filtering processes (Dauda et al. 2019). These sludges are solid residues from uneaten aquafeeds and waste from fish culture metabolism, such as feces or scales (Fraga-Corral et al. 2022). Therefore, an improper removal process from the culture systems could increase the concentration of nitrogen compounds, generating stress symptoms in the fish and the proliferation of aerobic bacteria that will increase the oxygen demand (Dauda et al. 2019).

In Chile, sludges produced by salmon farming reach 6.6 thousand tons per year (Aranda et al. 2018). Problems derived from their generation are mainly due to the costs associated with final disposal, the release of unpleasant odors caused by microorganisms, and the water content (Mirzoyan et al. 2010). Furthermore, sludges have elevated toxicity factors caused by pathogens, salinity, heavy metals, and antibiotics, mainly oxytetracycline and florfenicol (Mirzoyan et al. 2010; Gorito et al. 2022). However, the benefit of their reuse is to provide nutrients and organic matter to the crops (Maigual-Enriquez et al. 2019).

Therefore, waste management produced in high volume and constant flow generation requires technologies adapted to the needs of salmon farming and focused on improving its handling. For this reason, large-scale drying technologies have been developed to obtain dry sludge of high interest for its potential use in agriculture as an amendment. In this sense, the salmon farming industry utilizes drying technologies to obtain innovative and innocuous amendments and biofertilizers (Yogev et al. 2020; Cristiano et al. 2022), thereby opening up the possibility of conducting studies demonstrating improvements in crop yield and quality compared to traditional fertilizers (Ezziddine et al. 2021), in turn, promoting the adaptation of the salmon farming industry to a circular economy system and the generation of new sources of amendments that can compete with the increasing rise in fertilizer prices (Hebebrand and Laborde 2023).

Thus, this study investigates the thermo-drying process as an alternative to rapid waste conversion in salmon farming and evaluates how the temperature affects sludge nutrient content and its possible toxicity inhibition on seeds of agronomic interest crops. In the same way, this research contributes to the scarce available information related to the revalorization of aquaculture sludges with RAS in Chile.

Materials and methods

Sampling

The sludge samples were collected from the ‘Salmonífera Dalcahue Company’ from Chile (38°38’14.3“S 72°05’46.0“W). The material was obtained from a freshwater recirculation system (RAS) of Salmon salar culture in ponds with a 3.0 parts per thousand (ppt) salinity. The wet sludges (WS) samples are stored in a storage tank before drying, where the first sampling was performed. This WS sample was dried at 40 °C until constant weight (96 h), then stored until analysis. Subsequently, the storage tank transports the material to the drying equipment (Figure 1), which has an operation capacity to dry about 600 kg of wet sludge per cycle, with a yield approximately of 65 kg of dry sludge per cycle, completing a treatment cycle between 4–4.5 h at temperatures ~ 125 °C. Finally, the DS was collected from a bulk bag and stored as dry sludge until analysis.

Figure 1. Thermomechanical drying system for the salmon farming industry.

Physicochemical characterization

The physicochemical characterization of WS and DS were performed using the Test Methods for the Examination of Composting and Compost 2001 (TMECC methodologies) based on the Chilean Standards Compost NCh2880 (Sadzawka et al. 2005). The parameters evaluated were electrical conductivity (EC), pH, humidity, dry matter, organic matter (OM) content, NH₄⁺/NO₃⁻, and C/N ratios. The total nutrient concentrations, such as phosphorus (P₂O₅%), total potassium (K₂O%), iron (Fe), manganese (Mn), and boron (B) were also determined. In addition, the total concentration of different heavy metals such as arsenic (As), cadmium (Cd), copper (Cu), chromium (Cr), mercury (Hg), nickel (Ni), lead (Pb), and zinc (Zn) was determined. Furthermore, salinity was determined based on the chemical characterization of soluble substances such as calcium (Ca²⁺), magnesium (Mg²⁺), potassium (K⁺), sodium (Na⁺), chloride (Cl⁻), sulfate (SO₄^2-), and bicarbonate (HCO₃⁻). The chemical characterization was determined on dry weight.

The elemental determination was performed by the Dumas method based on the combustion of sludge samples (2–3 mg) in an Elemental Analyzer (Thermo, FlashSmart). The samples were burned at 950 °C in a pure oxygen atmosphere. The gases resulting from the combustion are then transported by He carrier gas to a furnace where the chromatographic column is located and operated at 60 °C. The elements were identified by a thermal conductivity detector (TCD). The chromatographic peaks integration gives the percentage proportion of the chemical elements contained in the sample, which are expressed as carbon (C), hydrogen (H), nitrogen (N), and sulfur (S) abundances (w/w%). The oxygen content was estimated by difference from unity after subtracting the ash and elemental C, H, N, and S content.

Thermogravimetric analysis (TGA)

Sludge samples (4–5 mg) were tested for thermal degradation in an inert N₂ atmosphere using a thermogravimetric analyzer (Shimadzu TG60H). The equipment was programmed at a heating rate of 10 °C min⁻¹ from 35 to 800 °C with a carrier at 50 mL min⁻¹ gas flow. During the analysis, the mass of the samples was measured to obtain the thermal degradation profile by thermogravimetry (TG) and derivative thermogravimetry (DrTGA).

Water extractable organic matter (WEOM)

Extracts from dry and wet sludge samples were obtained with water for chromatography (LC-MS grade) at a ratio of 1:10 m/v for 24 h in a horizontal shaker at room temperature. Then, the extracts were centrifuged at 5000 rpm (2200 × g) for 10 min and filtered with qualitative filter paper grade 292. Then, WEOM were stored for physicochemical characterization and spectroscopy analysis.

Spectroscopy analysis

To characterize the functional groups of the sludge samples, an Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) spectroscopy was used. Spectra were performed between 4000 to 600 cm⁻¹ range with a 1 cm⁻¹ resolution (CARY 630 FT-IR Agilent Technologies).

Fluorescence emission spectroscopy (FES) was recorded using Perkin-Elmer LS 45 on WEOM samples (He et al. 2011). Emission spectra were determined from 280 to 480 and 400 to 650 nm using a scanning speed of 300 nm min⁻¹ and an excitation wavelength of 254 and 370 nm, respectively. Also, the fluorescence index (FI) was determined using the emission intensity ratio at 450 nm and 500 nm at 370 nm excitation wavelength. Excitation spectra were determined between 300 to 500 nm at an emission wavelength of 520 nm and a scanning speed of 300 nm min⁻¹. Synchronous-Scan Excitation spectra were determined from 250 to 600 nm with a constant offset (Δλ = 30 nm) and 300 nm min⁻¹ scan speed. Finally, Excitation – Emission Matrix (EEM) spectra were obtained by scanning the emission wavelength from 280 to 550 nm by increasing the excitation wavelength by 5 nm increments from 200 to 500 nm and 1200 nm min⁻¹ scan speed. In this sense, raw data of EEM spectra were processed using the ‘staRdom’ package in RStudio. The results were corrected by blank subtraction, Raman normalization, removal and interpolation of Raman and Rayleigh scattering, and subsequent correction of dilution factor (d10).

Germination assay

The germination test was performed using the aqueous extract of the DS sample at different electrical conductivity values. The assay was according to Luo et al. (2018) and Curaqueo et al. (2020) with some modifications. Before the procedure, the dry weight factor (DWF) was determined according to the following equation considering the moisture content of the samples:

$DWF=\frac{wetmassingrams}{drymassingrams\hspace{0.17em}}$

Considering the DWF, an aqueous extract was prepared in a 1:10 (m/v) ratio with constant agitation for 1 h and then filtered with qualitative filter paper grade 292. The initial electrical conductivity value of the extract (T1 = 5.36 mS cm⁻¹) was determined and then adjusted to half (T2 = 2.60 mS cm⁻¹) and a quarter (T3 = 1.30 mS cm⁻¹) of its initial value with distilled water. The control sample was distilled sterile water (T0). Subsequently, the seeds were sterilized with 75% ethanol for 3 min at constant agitation and washed with sterile distilled water. The assay was performed in triplicate using Petri dishes with ten seeds in each one on filter paper. 4 mL of filtered sample (0.45 µm syringe filters) were added for each plate. Phytotoxicity was evaluated on lettuce (Lactuca sativa L.), radish (Raphanus sativus L.), and tomato (Solanum lycopersicum L.) species. Then, each plate was covered and sealed with parafilm and incubated at 25 ± 1 °C with a 16/8 h light/darkness photoperiod. After 6 d, germinated seeds were recorded, and root length was digitally measured. Finally, the germination percentage (G) and germination index (GI) were determined considering the relative seed germination (RSG) and relative root growth (RRG) using the following equations:

$GI=RSGxRRGx100\mathrm{\%}$

$RSG=\frac{numberofgerminatedseeds\left(sample\right)}{numberofgerminatedseeds\left(control\right)}$

$RRG=\frac{Totalradiclelength\left(sample\right)}{Totalradiclelength\left(control\right)}$

Data analysis

Statistical analysis was performed in RStudio version 2021.09.0 + 351 software. The results were plotted using the package ‘ggplot2’ for the thermogravimetry and spectroscopy analyses, except for EEM spectra, which were developed using the ‘plotly’ package. Student t-test with the Holm-Sidak correction method was used to analyze significant differences (p < 0.05) between WS and DS samples. Finally, the germination test and root length measurements were performed by analyzing images using ImageJ 1.50i software (Tajima and Kato 2013).

Results

Physicochemical characterization

The results obtained from the physicochemical characterization of the sludge samples are summarized in Table 1 and compared with the Chilean Compost Standard NCh2880 (INN 2015). The WS and DS showed differences in some of their parameters. Significant differences were observed for moisture content, dry matter, electrical conductivity, pH, and NH₄⁺/NO₃⁻. The moisture content was 63.0 ± 0.18 and 7.0 ± 0.10% for the samples collected from the storage tank before the drying process and after the thermal treatment, respectively. Additionally, pH evidenced a decrease from 7.4 to 5.3.

Table 1. Physicochemical characterization of wet and dry sludge.

Parameters	Wet sludge	Dry sludge	NCh2880
pH	7.4 ± 0.05	5.3 ± 0.03*	5.5^a −8.5^b
Moisture (%)	63.0 ± 0.18	7.0 ± 0.10*	30^a −45^b
Dry matter (%)	37.0 ± 0.18	93.0 ± 0.10*	70^a −55^b
Organic matter (%)	81.0 ± 0.54	82.0 ± 0.68	>20
Electrical conductivity (dS m^−1)	25.0 ± 0.12	5.4 ± 0.20*	3 <^a; 8 <^b
Organic C (%)	45.0 ± 0.46	45.6 ± 0.61	>11
Total N (%)	7.2 ± 0.10	6.4 ± 0.10*	>0.5
C/N ratio	6.3 ± 0.07	7.1 ± 0.05*	<30
NH4^+ (mg kg^−1)	36223 ± 311	2608 ± 34*	<500
NO3^− (mg kg^−1)	4671 ± 50	292 ± 2.65*	-
NH4^+/NO3^− ratio	7.8 ± 0.07	8.9 ± 0.10*	<3
Total P (g kg^−1)	34.9 ± 1.20	31.9 ± 0.90	-
Total K (g kg^−1)	1.25 ± 0.11	1.5 ± 0.09	-
Total Fe (mg kg^−1)	1150 ± 26	290 ± 5.70*	-
Total Mn (mg kg^−1)	120 ± 4.50	60 ± 1.80*	-
Total B (mg kg^−1)	16 ± 0.90	10 ± 0.60*	-
Total As (mg kg^−1)	<0.01 ± 0.0	0.105 ± 0.01*	15ª − 33^b
Total Cd (mg kg^−1)	<0.01 ± 0.0	<0.01 ± 0.0	0.7^a −9^b
Total Cu (mg kg^−1)	14.0 ± 0.2	24.0 ± 0.3*	70ª − 400^b
Total Cr (mg kg^−1)	28.8 ± 0.5	54.0 ± 0.3*	70ª − 100^b
Total Hg (mg kg^−1)	0.07 ± 0.0	<0.01 ± 0.0*	0.4ª − 2^b
Total Ni (mg kg^−1)	16.9 ± 0.3	27.8 ± 1.1*	25ª − 80^b
Total Pb (mg kg^−1)	0.47 ± 0.01	3.28 ± 0.3*	45ª − 220^b
Total Zn (mg kg^−1)	860 ± 5.2	812 ± 3.7*	200ª − 800^b

Data based on dry weight. Values represent mean ± standard deviation (n = 3). Asterisks indicate statistically significant differences (p < 0.05) between sludges according to the Student t-test. ^a= Class A Compost by Chilean Compost Standard (NCh 2880). ^b=Class B compost by Chilean Compost Standard (NCh 2880).

Electrical conductivity reflected high values in the WS, about 25.0 ± 0.12 dS m⁻¹, whereas, for DS, a significant decrease was observed to 5.4 ± 0.20 dS m⁻¹. Meanwhile, the C/N and NH₄⁺/NO₃⁻ ratios boosted from 6.3 ± 0.07 to 7.1 ± 0.05, and from 7.8 ± 0.07 to 8.9 ± 0.10 in the DS sample. Regarding the organic matter, total K and P concentration, both samples presented a similar behavior, showing low variation in the results. As observed in Table 1, the values of organic matter content were ~ 80%. Likewise, the values registered were 1.25 ± 0.11 (WS) and 1.50 ± 0.09 g kg⁻¹ (DS) for total K concentration, and 34.9 ± 1.20 (WS) and 31.9 ± 0.90 g kg⁻¹ (DS) for P concentration. Regarding the presence of metals, WS evidenced higher contents of total Fe, Mn, Zn, and B, whereas total contents of Cr, Ni, Cu, and Pb were higher values in DS.

Elemental analysis

Table 2 describes the elemental characterization of the wet and dry sludge samples. The quantification reported a higher concentration in the dry sludge sample, except for the oxygen content. The N amount raised from 2.9 ± 0.1 to 6.7 ± 0.4% in the DS and WS, respectively. Similarly, the C concentration reported a variation from 35.1 ± 1.5 to 39.9 ± 0.1% concerning the WS and DS samples, respectively. Furthermore, H content showed a similar trend as N and C, increasing from 5.2 ± 0.2 to 5.8 ± 0.0%. On the contrary, the O content value decreased from 56.8 ± 1.6 to 47.2 ± 0.2 for the wet and dry sludge samples, respectively.

Table 2. Elemental composition and indicators of wet and dry sludge.

Sludge	Elemental percentage (wt %)					C/N	H/C	O/C
	N	C	H	S	O^a
Wet	2.9 ± 0.1	35.1 ± 1.5	5.2 ± 0.2	0.0 ± 0.0	56.8 ± 1.6	12.0 ± 0.8	0.2 ± 0.0	1.6 ± 0.1
Dry	6.7 ± 0.4*	39.9 ± 0.1*	5.8 ± 0.0*	0.3 ± 0.0*	47.2 ± 0.2*	6.0 ± 0.3*	0.2 ± 0.0	1.2 ± 0.0*

^aThe oxygen content was calculated by difference. Values represent mean ± standard deviation (n = 3). Asterisks indicate statistically significant differences (p < 0.05) between sludges according to Student t-test.

Thermogravimetric analysis

The thermal profile of sludge samples was plotted as TGA in Figure 2a, and a detailed analysis was summarized in Table 3. The results obtained for wet sludge analysis can be described in four stages and related to the peaks obtained on the DrTGA (Figure 2b). The first stage (33–200 °C) concerns the dehydration process and volatilization of low molecular weight volatile compounds, representing a decrease equivalent to 16.2% of the initial sample. Then, a mass loss of 33.8% was recorded between 200 and 375 °C, possibly due to a devolatilization process caused by organic compounds’ decomposition of low thermal stability. In the third stage (375–600 °C), a mass decrease of 24.1% was associated with highly stabilized and condensed complex aromatic compounds. A final stage was observed between 600–800 °C, which maintained a relatively constant mass that remained a residual mass of 25.4%.

Figure 2. Thermogravimetry analysis curve (TGA) (a) and derivative thermogravimetry (DrTGA) analysis for wet and dry sludge.

Table 3. Thermogravimetric analysis stages of wet and dry sludge.

Sludge	Mass loss (%) 1^st stage	Mass loss (%) 2^nd stage	Mass loss (%) 3^rd stage	Mass loss (%) 4^th stage	Residue (%)
Wet	16.2	33.8	24.1	0.8	25.1
Dry	6.8	43.6	29.8	-	19.8

Thermal degradation of dry sludge was characterized by three stages and associated with the results obtained from the DrTGA plot (Figure 2b). In the first stage (33–150 °C), the initial weight loss was 6.8%, corresponding to thermal devolatilization of light volatile matter content and loss of moisture. Then, the second stage (150–400°C) reported a mass reduction of 43.6%. The third stage (400–800 °C) recorded a constant degradation process and 29.8% mass loss until the end of the analysis.

FT-IR spectroscopy

The FT-IR spectra of wet and dry sludge samples are presented in Figure 3. An intense broad band around 3270 cm⁻¹ was associated with O-H stretching vibrations due to hydroxyl groups from alcohol or phenolic substances. This peak was followed by a doublet around 2918 and 2849 cm⁻¹ related to the C-H stretching vibration of aliphatic structures. The sharp band at 1623 cm⁻¹ was associated with the carboxylic and ketonic group due to C=O stretching vibration and possibly to the C=C stretching of aromatic structures. Close to this peak, around 1541 cm⁻¹ was and detected a band concerning C=N stretching and N-H deformation vibration of amides. The peak at 1403 cm⁻¹ was characterized by bending and C-O stretching vibration in the phenolic O-H group. A low-intensity peak at 1230 cm⁻¹ was described possibly as NH⁺ deforming vibration. Finally, a large peak at 1021 cm⁻¹ was observed, corresponding to the C-O stretching.

Figure 3. FT-IR spectra of wet and dry sludge.

Fluorescence spectroscopy

The sludge’s WEOM characterization using fluorescence emission spectroscopy (FES) is shown in Figure 4a. For the emission spectra at 254 nm, both samples showed a broad band at 430 nm. At an excitation wavelength of 254 nm, two characteristic bands at ~ 340 or ~430 nm for both spectra were detected. Furthermore, the emission spectra detected at an excitation wavelength of 370 nm (Figure 4b) showed a broad band centered around 447 nm, corresponding to the highest results in the dry sludge sample.

Figure 4. Spectra of wet and dry sludge of emission at 254 nm (a) and 370 nm (b) excitation wavelength, excitation at 520 nm emission wavelength (c) and Synchronous-scan spectra (d).

The excitation spectra of the wet and dry sludge extracts are presented in Figure 4c. A peak at 392 nm was recorded for both samples, while a slight shift was observed in the peak from 342 to 352 nm in the wet and dry sludge samples, respectively.

Synchronous-scan spectra (Figure 4d) for wet and dry sludge extracts were determined. These results reported the presence of a peak centered at ~280 nm with a higher intensity in the wet sludge sample, possibly associated with the thermal degradation of the protein substances in the dry sludge sample. A second peak was observed centered at ~350 nm in the wet sludge sample, and a lower resolution shoulder was reported in the dry sludge sample. The last band was determined centered around 390 nm in both samples but with a higher intensity for the dry sludge sample.

According to Excitation-Emission Matrix (EEM) contours (Figure 5), the region I did not present quantifiable peaks in both sludge samples evaluated. Region II reported low levels associated with the presence of aromatic protein substances, which completely disappeared after the drying process. Region III showed a peak centered around 245/430 nm (Ex/Em) in the wet sample, which decreased significantly in the dry sample. In turn, region IV, associated with soluble microbial by-product-like substances, reported a peak centered around 280/360 nm (Ex/Em), which decreased significantly after thermic treatment. Region V showed a different trend by increasing its intensity from the peak around 345/435 (Ex/Em), associated with the presence of humic acid-like substances.

Figure 5. Excitation – emission matrix (EEM) spectra of wet sludge (a) and dry sludge (b).

Germination assay

The results obtained to evaluate the phytotoxic effect of the DS aqueous extract are shown in Figure 6 and Table 4. The germination percentage (G) showed favorable results in control and T3 treatments, even above the manufacturers’ germination values and low germination rate for T2. In this sense, L. sativa and S. lycopersicum reached germination values of 96.7% and 90%, respectively, when T3 treatment was applied, while for R. sativus, the germination percentage was 70%. For T2 treatment, R. sativus reported germination values of 50%, while L. sativa y R. sativum recorded germination values of 86.7 and 70%, respectively.

Figure 6. Effect of dried-sludge aqueous extract at different EC values on germination of seeds of (a) L. sativa, (b) R. sativus, and (c) S. lycopersicum (n = 3).

Table 4. Effect of dry sludge aqueous extract at different EC values on the germination rate on L. sativa, R. sativus, and S. lycopersicum seeds.

Species	T1		T2		T3		Control		Minimum germination (%)^(c)
	G(^a)	GI(^b)	G	GI	G	GI	G	GI
L. sativa	0	0	86.7	11.8	96.7	26.2	96.7	-	75.0
R. sativus	0	0	50.0	33.0	70.0	39.2	93.3	-	80.0
S. lycopersicum	0	0	70.0	6.3	90.0	19.2	93.3	-	75.0

Germination percentage (%)^(a). Germination index^(b). Described by the manufacturer^(c). T1: 5.36 mS cm⁻¹. T2: 2.60 mS cm⁻¹. T3: 1.30 mS cm⁻¹. Control: Distilled sterile water. Values represent mean (n = 30).

Regarding the germination index (GI), T2 and T3 showed better values for all evaluated species, while T1 presented zero percent values for all seeds. In this sense, T2 registered GI values of 11.8%, 33.0%, and 6.3% for L. sativa, R. sativus, and S. lycopersicum, respectively; meanwhile, the results of GI for the same crops were 26.2%, 39.2% and 19.2% under T3 showed a visible salt dilution effect.

Discussion

The current study found that WS’s parameters were similar to those reported in other studies on the physicochemical characterization of aquaculture wastes, including fish excretion (Kurniawan et al. 2021) and brackish RAS (Zhang et al. 2016). However, moisture, pH, and EC values are strongly affected after the thermal treatment. In this sense, according to Chilean Standards Compost NCh2880 (INN 2015), the dry sludge complies with several parameters, including organic matter content, an interesting level of nitrogen, C/N ratio, NH₄⁺/NO₃⁻ ratio, and lower levels of heavy metals (As, Cd, Cu, Cr, Hg, Ni, Pb) to be classified as class A compost. In turn, parameters such as electrical conductivity and total Zn are within the upper limits of the standard and can be classified as class B compost according to the NCh2880 standard.

On the other hand, the decrease of pH coupled with the increase of C/N and NH₄⁺/NO₃⁻ ratios in DS could be due to the volatilization of nitrogen compounds in the form of ammonia (NH₃) and nitrous oxide (N₂O) as a result of the high temperatures (~125°C) of the drying equipment (Cáceres et al. 2018). According to Fan et al. (2011), temperature is one of the main factors that affect ammonia volatilization and N transformation. In this case, it caused a decrease in substances such as ammonium (NH₄⁺) of almost 93% and nitrate (NO₃) of 94% (Zeng et al. 2013; Cáceres et al. 2018; Wang et al. 2020). Furthermore, high temperatures could promote cell wall degradation and denaturation of proteins and genetic material (Russell 2003). Consequently, this could favor microorganisms’ death, making sludge innocuous. In addition, this process, together with the degradation of complex compounds of high molecular weight, contributes to CO₂ generation (Ge et al. 2015) and decreases pH by releasing H⁺ into the medium. Therefore, the increase in NH₄⁺/NO₃⁻ ratio could affect the development of germination assays because a high value of NH₄⁺/NO₃⁻ ratio could be considered phytotoxic, and an excessive NH₄⁺ amount can cause plant growth inhibition due to rhizosphere acidification (Zhu et al. 2021), affecting the seed testa rupture and the emergence of the radicle (Weitbrecht et al. 2011).

Regarding EC values, this parameter is associated with salinity levels in the sample, which can cause adverse effects on soil and plant growth (Jiang et al. 2015) and also a possible disruption of soil physical properties, ionic toxicity, and osmotic stress (Ravindran et al. 2017). Based on the high values of EC in WS, this sample was not considered in the germination assay. However, it is crucial to consider that salinity can cause osmotic and oxidative stress that could imply a delay in seedling growth (Longo et al. 2021). Therefore, it is recommended the dilution of DS extracts during germination assay.

Phosphorus concentration and organic matter are in the range reported by other studies, such as showed minimal variation in the results, indicating that high temperatures would not affect this parameter of interest. However, organic matter presented higher values than that described for other agricultural residues, such as cattle and poultry manure, with 73.9% and 65.2%, respectively (Bustamante et al. 2008). Regarding P, some research reports similar values with results ranging from 35.9 to 25.5 g kg⁻¹ in tilapia RAS (Monsees et al. 2017). Idrovo-Novillo et al. (2018) also reported a lower value in the P evaluation for poultry manure samples. Among these residues, broiler chicken, hen, and quail manures were described, with values of 9.23, 14.56, and 14.91 g kg⁻¹, respectively. In turn, studies of livestock and poultry manures report values of 8.92 and 9.60 g kg⁻¹ (Torres-Climent et al. 2015).

The thermal treatment decreased Fe, Mn, Zn, and B contents in DS, whereas K, Cu, Cr, Ni, and Pb increased. In the case of K, the concentration was lower than those reported in studies of livestock and poultry manures, between 20.2–38.4 g kg⁻¹ (Torres-Climent et al. 2015; Idrovo-Novillo et al. 2018). Additionally, according to studies conducted by Hepp (2012), in which several samples of fish farm sludge were evaluated, the concentration of these nutrients varied between 536–3,122 mg kg⁻¹ for available N, 8.5–75.5 g kg⁻¹ for total P, 71–173 mg kg⁻¹ for available K and 50% for organic matter content. Regarding Fe, Mn, and B, Belmeskine et al. (2023) studies report a lower trend in their results with values of 33.1, 0.45, and 11.1 g kg⁻¹ in fish farm sludge.

The results of the heavy metal composition showed an increase in concentration in the DS sample, and, except for Zn and Ni, the analysis of the WS reported low values. In this sense, the solubility of heavy metals can be affected by the medium’s pH, temperature, and salinity (Márquez et al. 2008). These conditions could allow the transport and accumulation of heavy metals in the sludge. Therefore, the drying process at high temperatures could increase the concentration of these phytotoxic elements.

Furthermore, the concentration of soluble ions related to the salinity of the sludge was analyzed (Table 5). After the thermal treatment, the results showed a significant decrease in K⁺, Na⁺, Cl⁻, SO₄^2-, and HCO₃. In contrast, Ca²⁺ and Mg²⁺ concentrations increased for DS. For this reason, the electrical conductivity reflected a significant decrease in its values, a parameter directly related to salinity. This decrease could be attributed to processes such as the precipitation of mineral salts and the transformation of small molecules due to high temperatures (Wang et al. 2016).

Table 5. Soluble ions characterization of wet and dry sludge.

Soluble ions (mg L^−1)	Wet sludge	Dry sludge
Ca^2+	62 ± 0.4	324 ± 5.3*
Mg^2+	17 ± 0.3	73 ± 1.1*
K^+	133 ± 0.9	31 ± 2.3*
Na^+	1290 ± 10.5	591 ± 7.6*
Cl^−	2216 ± 12.2	886 ± 4.9*
SO4^2-	955 ± 9.3	19 ± 0.6*
HCO3^−	18422 ± 103.7	2007 ± 14.6*

Values represent mean ± standard deviation (n = 3). Asterisks indicate statistically significant differences (p < 0.05) between sludges according to Student t-test.

Meanwhile, compounds such as HCO₃⁻ are susceptible to reacting with ionic substances to form insoluble carbonates, causing their precipitation with Ca²⁺ or Mg⁺² (Wang et al. 2020). After a thermal degradation process, this would cause their dissociation in the CO₂ form and solubilization of these elements (Barceló et al. 2002). Finally, ammonia volatilization may be a factor in salinity variation because combining this compound with Ca²⁺ and Mg²⁺ can lead to crystal formation, affecting the germination assay (Jiang et al. 2016).

The elemental analysis results highlighted a reduction of C/N (50%) for the DS due to increased elemental nitrogen percentage. Studies by Monsees et al. (2017) report a C/N ratio in freeze-dried sludge from tilapia RAS of about 9.2% with C and N values of 35.6 and 3.9%, respectively. Parameters such as the H/C ratio reported no variation in their values. Conversely, the O/C ratio showed a slight decrease affected by high temperatures. These ratios indicate the aromaticity degree of the organic matter and the presence of O-containing polar groups such as phenols, alcohols, or carboxylic groups (Aranganathan et al. 2019).

Based on the thermogravimetric results, the differences between WS and DS curves are related to the thermal treatment, affecting the degradation profiles. In this context, dry sludge evidenced a lower content of moisture and the disappearance of the fourth stage, which is also reported by Maigual-Enriquez et al. (2019) for RAS sludge dried at 105 °C. Therefore, the first stage, in which occur dehydration and volatilization of low molecular weight volatile compounds (Zhang et al. 2022), showed a lower mass reduction for DS. Both samples evidenced a second stage, possibly associated with the devolatilization of aliphatic hydrocarbons, peptides, and phenolic, carboxyl, and carbonyl structures (Zhou et al. 2022). The mass reduction in the second stage for DS was higher than for WS. This behavior could be attributed to a higher decomposition of carboxyl groups, lipids, protein structures, and biodegradable organic matter than the wet sludge sample. The third stage was also observed for WS and DS. According to Languer et al. (2020), this stage is associated with highly stabilized and condensed complex aromatic compounds. Some authors propose that the accumulation of these compounds may be related to processes in aquatic environments. These include the leaching of plant organic matter directly into the water or through the soil profile, the leaching of soil humic and fulvic acid into water, and/or polymerization reactions among phenolic, aldehyde, and amine functional groups from biological products in natural waters (Sharma and Anthal 2016). In the final stage, the remaining mass is related to the content of inorganic material in the sludge sample and carbonate decomposition (Naqvi et al. 2021).

The FT-IR spectra of wet and dry sludge showed a similar trend, but the dry sludge spectrum showed a higher intensity, possibly associated with a higher concentration of functional groups due to the effect of the drying process. Regarding the large peak at 1021 cm⁻¹, several authors reported that for aquaculture sludges, the C-O stretching is associated with substances like polysaccharides, alcohols, and aliphatic ethers (Kowalski et al. 2018; Cristina et al. 2020; Haouas et al. 2021; Xiaoyan et al. 2021; Xiao et al. 2022). Studies by Pérez-San Martín et al. (2023) compile information about the spectroscopy characterization of organic amendments, highlighting its advantages and disadvantages and the infrared analysis function.

The fluorescence emission spectroscopy (FES) technique is also used to analyze the sludge samples’ stability degree of organic matter. Some research has focused on discussing the presence of humic-like substances in recirculation aquaculture systems. This suggests that the successful and efficient management of RAS is essential in controlling the accumulation and quality of dissolved organic matter in the system (Hambly et al. 2015). The presence of humic- and fulvic-like compounds is a clue of partially-stabilized organic matter. These compounds can emit with some intensity by stability condensed aromatic rings and unsaturated aliphatic carbon chains (He et al. 2011). According to Wang and Zeng (2018), p. 340 and 430 nm bands are related to components with simple structures and small degrees of conjugated chromophores o to high molecular weight compounds with a great degree of conjugation, respectively.

The DS showed a broad band centered around 447 nm at an excitation wavelength of 370 nm, corresponding to their highest value ((Figure 4b). This result is comparable to those obtained in compost characterizations from other organic wastes and is related to the presence of fulvic-like substances (Wei et al. 2014). From this plot, it is feasible to determine the fluorescence index (FI) described to evaluate the degree of aromaticity of the organic matter. Hence, a high FI value (~1.7 - ~2.0) reflects an organic matter of microbial origin, whereas a low FI value (~1.3 - ~1.4) is related to organic matter of an allochthonous origin (terrestrial) (Yu et al. 2011). In this regard, the calculated ratios were 1.67 and 1.74 for the wet and dry sludge samples, respectively, suggesting that the organic matters were microbially derived. Studies related to this index indicate values above 2.0 for various mature composts, i.e. highly stabilized organic material (Wei et al. 2014).

The excitation spectra of the wet and dry sludge extracts showed two characteristic peaks with a higher intensity in the dry sludge sample (Figure 4c). Some studies have reported that the peak obtained at around ~390 nm possibly relates to fluorophores originating from the polycondensation of carbonyl groups and phenolic structures (He et al. 2011). Other authors have described that this peak is associated with humic-like substances during the composting of agricultural residues (Huang et al. 2009). In turn, the shoulder detected at 342/352 nm is not characteristic of all types of stabilized waste (Wei et al. 2014) and could be associated with non-condensed organic matter.

Synchronous scan (Figure 4d) represents the summation of the spectra of different fluorophores present in dissolved organic matter and is generated from simultaneous scanning of the excitation and emission spectra with a constant wavelength (He et al. 2011; Guo et al. 2012). In this sense, the spectra obtained can be divided into Region A (250–308 nm) related to the presence of protein-like substances and mono-aromatic compounds. Region B (308–365 nm) is associated with the presence of the fulvic-like substance, and Region C (363–550 nm) for the identification of humic-like substances (Song et al. 2015). The peaks at 350 and 390 nm are a critical factor that could represent the presence of stabilized compounds such as fulvic acid-like and humic acid-like substances, respectively. In this sense, several studies reported that the accumulation of humic substances represents about 95% of dissolved organic matter in aquatic ecosystems with concentrations from 0.5 mg L⁻¹ to 50 mg L⁻¹ (Lieke et al. 2021), and the accumulation in RAS can reach 18 mg L⁻¹ (Meinelt et al. 2010).

EEM fluorescence spectroscopy was used to determine the content of fluorescent substances associated with organic matter in both samples. This characterization can be described by five distinct regions represented by specific dissolved organic matter. Regions I and II are related to amino acids derived from simple aromatic proteins such as tyrosine-like and tryptophan-like with peaks between 250–330 nm Emission wavelength (Em)/200–250 nm Excitation wavelength (Ex)/, and 330–380 nm Em/200–250 nm Ex, respectively; Region III is reported with peaks at 380–550 nm Em/200–250 nm Ex and is related to fulvic acid-like substances; Region IV is associated with microbial by-product-like material as a result of microbial protein degradation, reporting peaks at 250–380 nm Em/250–350 nm Ex; Region V includes wavelengths at 380–550 nm Em/250–500 nm Ex and is directly associated with humic acid-like substances (Chen et al. 2003; Duan et al. 2021; Jiang et al. 2022). In this sense, it is essential to state that the presence of peaks in the region V is considered an indicator of sludge stabilization. According to Coban et al. (2020), humic acids are characterized by their solubility at high pH levels, which favors their dissolution and transport in water flows, especially if they come from acidic pH environments with lower solubility. In the case of fulvic acids, they describe a wide pH range in which they can dissolve and accumulate.

The germination of R. sativus, L. sativa, and S. lycopersicum seeds was an assay to evaluate the phytotoxic effect of the DS aqueous extract and its dilutions. Both species are essential because they are widely used in family farming, consumed in the country, and considered model plants. The inhibitory properties of DS aqueous extract were associated with the high EC values reported for the samples ranging from 0.1 mS cm⁻¹ (control) to 5.36 mS cm⁻¹ (T3). The assay reported a significant decrease in the phytotoxic, related to low levels of EC in control, T2, and T3 treatments compared to T1, where the saline solution was more concentrated (5.36 mS cm⁻¹). The treatment T1 inhibited the seed germination for all species assayed, confirming that the high values of EC, NH₄⁺/NO₃⁻ ratio, and the presence of soluble ions and trace metals completely interrupt the seeds growth.

The germination percentage (G) showed favorable results in control and T3 treatments, even above the manufacturers’ germination values and low germination rate for T2. Regarding the germination index (GI), all extracts except the control were lower than 50%. A GI > 80% indicates that there are no phytotoxic substances or these are in a low concentration; contrarily, IG values between 50 and 80% are related to a moderate presence of phytotoxic substances, and IG < 50 are associated with a high amount of phytotoxic substances (Zucconi et al. 1981; Asquer et al. 2017). In this sense, Weitbrecht et al. (2011) described that the germination process can be morphologically divided into three phases. Phase I is the process of imbibition or absorption of water by the seed and its size increase. Phase II is when the seed testa ruptures, and the radicle emerges. Phase III is the process of endosperm rupture to give way to root elongation, completion of germination, and subsequent seedling growth. Therefore, it is highly probable that Phase I may be the most affected in the seed inhibition process caused by the high salinity of the DS sample, preventing the correct absorption of water from the medium. In turn, phase III could be affected by high NH₄⁺ concentrations decrease radicle elongation (Luo et al. 2018). This DS salinity is critical when using salmon farming sludge without a correct pre-treatment or stabilization process.

Conclusions

The characterization of the dry sludge proves that the drying process is advantage in reducing water content and volume, allowing a better management for soil application. In addition, the thermal process did not affect the content of organic matter and nutrients such as phosphorus and potassium and toxic compounds such as ammonium, soluble ions associated with salinity, and possible pathogens showed a decrease in the dry sludge compared to wet sludge.

This research also reports the presence of proteins, humic, and fulvic substances in the sludge samples, considering that the high temperatures during the drying process do not negatively affect the presence of humic substances.

The results establish the importance of drying sludge from salmon industry recirculation systems as a potential material for use in agriculture and improving degraded soils, however, we recommend considering that applying a higher dose of dry salmon sludge to the soil could affect some soil properties due to the heavy metals or high salinity that could affect seed germination.

Acknowledgments

This work was supported by National Agency for Research and Development (ANID)/Scholarship Program/Doctorado Nacional/2020 – 21201805 (A. P-SM) and VIPUCT 2020REGGC07 Project (G.C.). We also thank the companies Salmonífera Dalcachue and Circular Solutions for facilitating the sludge samples.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the National Agency for Research and Development (ANID), Scholarship Program/Doctorado Nacional/2020 [21201805]; VIPUCT [2020REGGC07 Project].