Effect of Animal ByProducts Fertilization on Durum Wheat in Mediterranean Conditions: Preliminary Results

This study aims to evaluate the effects of new-BioFertilizing Amendments (BFAs) deriving from fast organic matter decomposition of Animal ByProducts (ABPs) in comparison with ordinary soil organic amendments (compost), mineral N-fertilizers and no fertilization, on durum wheat development and production in a field trial under Mediterranean conditions. Results showed taller plants with heavier spikes and greater vigor in plots fertilized with BFAs when compared to no fertilization and N-fertilization, respectively. Likewise, BFAs fertilization resulted in higher protein content, gluten content, protein yields and higher values of yellow index with respect to no fertilization and N-fertilization. In contrast, lower values for test weight in correspondence of BFAs fertilization as well as no statistically significant differences on grain yield and gluten index were found. These preliminary results suggest that replacing N-fertilization with BFAs can be effective to ensure crop quality and yield stability in Mediterranean conditions.


Introduction
In order to feed the increasing world population, there is a need to produce about 50% more food by 2050. Since the 1960s, the availability of food per capita has increased more than 30% supported by a greater use of nitrogen (N) fertilizers (increase of about 800%) and irrigation water (increase of more than 100%) [1]. At the current levels of land exploitation this would require an increased cropland area between 6% and 21% more than the cropland area of 2010 (i.e., 90-325 million hectares vs 1567 million hectares) [2]. Nevertheless, intensive soil management practices (e.g., moldboard plough) as well as expanding world cropland area are bound to engender significant increases in greenhouse gas (GHG) emissions on one hand, and increasing land degradation and/or desertification particularly in the arid, semi-arid and dry sub-humid areas [3,4] on the other hand. As an outcome, increasing land overexploitation will exacerbate global warming and affect soil degradation, desertification and crop yields [1]. Concerning wheat, yields have declined at a global scale by 1.8% between 1981 and 2010 due to climate change [5]. In Europe, wheat-yield declines of 2.5% since 1989 have been reported, amplified by a drying trend leading to further yield declines of 5% or more in Italy [6]. Likewise, projected rising temperatures are expected to reduce global wheat yields by about 4%-6% for every degree of temperature rise [7,8]. Focusing on cereal-forage systems in the Mediterranean growing conditions of Sardinia, predicted durum wheat yield reductions between 16.2% and 19.0% for simulated increasing air temperature (from +1 • C to +6 • C) and decreasing rainfall (from −5% to −30%) scenarios have been reported [9]. Moreover, climate change is considered to exacerbate land degradation and potentially degradation of various byproducts matrices, both of animal and/or vegetal origin, resulting in a new bio-fertilizer that can be used in agriculture (Italian Ministry of Agricultural, Food And Forestry Policies -MIPAAF, Registration No. 0027872 /19) and included in the class of soil amendments listed in the Italian Legislation (Annex 2 of the National Legislative Decree 75/2010).
This current study is a part of a project with the aim of evaluating the effects of these new amendments, in comparison with traditional compost, ordinary N-fertilizers and no fertilization on yield and quality of cereal-forage cropping systems of Sardinia, as well as their impact on the main physical, chemical and microbiological characteristics of soils, both on a plot scale and in open field conditions, over a 3-year time span.
In this project the new biofertilizing amendments (BFAs) are evaluated in order to assess whether they can supplement or replace chemical fertilizers in the ordinary agronomic management of the cereal-forage cropping systems of the Mediterranean basin, such that they can lead to a new model of economically competitive agriculture, more efficient in the use of resources. An approach that is more responsible towards the environment and society and, hence, fully oriented towards sustainable production.
While several studies on the influence of biowaste compost [27], composted sewage sludge [28] and biochar [29,30] on durum wheat have been carried out, to our knowledge no similar studies have been published about the fertilization effect of BFAs on durum wheat.
Here we report some preliminary results obtained during a field trial conducted in 2019. Table 1 lists the BFAs used in this study (CS/N, CS 2, CS 3, CS 4, CN), as well as their composition and application on the trial plots. The effects of BFAs on durum wheat growth and production have been compared with a binary N-fertilizer (CMIN, diammonium phosphate: N 18% and P 2 O 5 46%) and the control (0-treatment, U) ( Table 1). The distribution of some BFAs has determined effects on Soil Organic Matter (SOM) and wheat growth when compared to the control (0-treatment, U). Table 2 shows the effects of the different BFAs, N-fertilizer and control on the percentage content of SOM (soil sampling 0-5-cm depth), on some plant traits (i.e., plant height, spike weight and density) and on the normalized difference vegetation index (NDVI) [31], which is an index of vigor of plants, either at booting and flowering. Some BFAs treatments tend to increase the SOM content in comparison with the control. Indeed, higher values of SOM have been observed in plots treated with CS/N and CS 4 (4.31% and 3.29%, respectively) and are significantly different from 0-treatment (U), from the other BFAs CS 2 and CS 3 and from CN and CMIN. Means followed by the same letter (in bold) do not differ significantly at p ≤ 0.05 by LSD.

Results
Concerning bio-morphologic traits, the higher values of plant height and spike weight registered with ABP treatments confirm the positive effect of these amendments. In particular, all the plots treated with BFAs show a higher average plant height as well as a higher average spike weight in comparison with no fertilization and N-fertilization, respectively. The best results for plant height were registered for CS/N (73.4 cm), CN (71.3 cm), CS 2 and CMIN (70.9 cm and 70.3 cm, respectively). In contrast, untreated plots show the lowest average plant height (66.1 cm).
As for spike weight, BFAs exhibit a general positive effect with highest values for CS 2, CS 4 and CS 3 (3.69 g, 3.61 g and 3.59 g, respectively) while CMIN and CS/N (3.03 g and 2.87 g, respectively) are not statistically different from the untreated plots (2.61 g). No effects were observed on the spike density ( Table 2).
The effects of treatments with BFAs can also be appreciated visually in Figure 1 both from the aerial photo (a) and from the calculated NDVI map (b), where plots treated with BFAs show a darker green color corresponding to higher NDVI values. Indeed, the quantitative evaluation of the positive effect of BFAs was estimated by the normalized difference vegetation index (NDVI) either at booting or flowering.  Table 3 shows the mean comparisons of the treatments. Concerning grain yield, no statistically significant differences among treatments were found. Regarding grain quality, measured on the basis of test weight, 0-treatment (U) exhibits the best results (84.2 kg hL −1 ), followed by CMIN and CS/N (83.6 kg hL −1 and 83.4 kg hL −1 , respectively). On the other hand, the lowest values were registered for CS 2 and CS 4 (82.4 kg hL −1 and 82.2 kg hL −1 , respectively). On the contrary, no difference were observed for 1000 kernel weight (Table 3).  The NDVI exhibits overall highest values in correspondence of CS 2 and CS 4 (0.412 and 0.278 for CS 2 and 0.399 and 0.277 for CS 4 at booting and flowering, respectively). These positive results are confirmed to a lesser extent for CN and CS 3 (0.384 and 0.273 for CN and 0.331 and 0.238 for CS 3 at booting and flowering, respectively). Oppositely, CMIN and U, with 0.324 and 0.217 and 0.238 and 0.121 at booting and flowering, respectively, denote the lowest NDVI values conveying less vigor compared to the other treatments. Table 3 shows the mean comparisons of the treatments. Concerning grain yield, no statistically significant differences among treatments were found. Regarding grain quality, measured on the basis of test weight, 0-treatment (U) exhibits the best results (84.2 kg hL −1 ), followed by CMIN and CS/N (83.6 kg hL −1 and 83.4 kg hL −1 , respectively). On the other hand, the lowest values were registered for CS 2 and CS 4 (82.4 kg hL −1 and 82.2 kg hL −1 , respectively). On the contrary, no difference were observed for 1000 kernel weight (Table 3). Technological quality, measured on the basis of protein content, protein yield, yellow index, gluten content and gluten index, denotes statistically significant differences among the treatments, with the important exception of gluten index. In particular, protein content and gluten content exhibit the same trend with CS 2, CS 4 and CN having the highest values (13.1% and 10.5%, 12.9% and 10.2%, 12.6% and 10.1%, respectively) while untreated plots register the lowest values (11.1% and 8.6%, respectively). Similarly, protein yield (protein content multiplied by the grain yield) shows more or less the same trend with higher values for CN (0.406 Mg ha −1 ), CS/N (0.393 Mg ha −1 ), CS 2 (0.389 Mg ha −1 ) and CS 3 (0.371 Mg ha −1 ) and the lowest one for U (0.289 Mg ha −1 ).
A positive trend among treatments for yellow index can be found in favor of BFAs: CS 3 (15.07 b*), CN (15.05 b*) and CS 4 (15.03 b*) show the highest values and CMIN and U the lowest ones (14.76 b* and 14.77 b*, respectively). In contrast, no significant trend for gluten index was found, CS/N and U showing the highest observed values (96 and 95, respectively).

Discussion
Considering the detrimental effects of soil degradation and declining yields, improved cropland management should strive to: (i) increase soil fertility by enhancing SOM; (ii) replace chemical fertilizers with green manures to restore soil fertility and preserve yield stability [6,32]. The preliminary results obtained in this study seem to confirm the effectiveness of this approach.
Concerning grain yield, data show no significant differences among treatments (Table 3). Therefore, BFAs can be a remarkable alternative to ordinary N-fertilizations. Comparing the effects of mineral fertilization with increasing doses of compost in Southern Italy, Pasqualone et al. [27] found no significant differences on durum wheat yields in correspondence of high levels of compost (12 Mg ha −1 ). Likewise, Fecondo et al. [27] found that the use of 40 Mg ha −1 of compost significantly increased durum wheat yields in comparison with mineral fertilization in rotation with tomato. Other studies confirm the positive effect of compost on bread wheat yields in Australia [29] and in Italy by comparing the effects of organic commercial fertilizers and compost on durum wheat yields in an organic agriculture Plants 2020, 9, 1094 6 of 13 context [30,33]. In the light of existing literature, the preliminary results presented in this study seem to open the way to the replacement of ordinary N-fertilizers with no detrimental effects on durum wheat yield, at least in the Mediterranean rainfed cereal-forage cropping systems.
Concerning the relationships between grain yield, plant height and spike weight, plots fertilized with BFAs are generally taller and exhibit heavier spikes with respect to the control and N-fertilization (Table 2). Interestingly, plants treated with BFAs show greater vigor than plants with ordinary N-fertilization as observed by NDVI measured at booting and flowering ( Table 2). The positive effect on plant vigor resulting in taller plants has been described on durum wheat at increasing levels of compost application [28], although not significantly different from mineral fertilization and combined compost + mineral fertilization. Our preliminary observations highlight a general positive effect of BFAs on plant growth and development resulting in higher grain quality and protein content (Table 3). Nonetheless, this result needs being confirmed in the next cropping seasons and by further investigations due to lacking studies about this specific topic.
As for grain quality, the negative correlation between yields and test weight is well known by plant physiologists (Table 3) and is associated with the relationship between test weight and spike weight ( Table 2). This result is due to the reduced competition among caryopses for the translocation of photosynthates from the stem and flag-leaf during grain-filling: fewer caryopses per spike in small spikes uptake more nutrients resulting in a larger grain size. Even though a somewhat negative effect on test weight was registered when compared with the control and N-fertilization, all plots treated with BFAs show high values of test weight, thus confirming the good to excellent grain quality (Table 3). However, a different, positive effect of compost in comparison with mineral fertilization on test weight had already been found by Fecondo et al. [27] and, to a lesser extent, by Pasqualone et al. [28].
Regarding the technological quality, protein and gluten content show significant differences among the treatments in favor of BFAs with respect to ordinary N-fertilization and no fertilization (U). This result is in agreement with Fecondo et al. [27], who found that the application of 40 Mg ha −1 significantly increased protein content by 26% and 14% in comparison with no fertilization and mineral fertilization, respectively. Conversely, Vaccari et al. [34] found no significant differences between the control (0-fertilization) and two biochar treatments (30 Mg ha −1 and 60 Mg ha −1 , respectively) on grain N content. Noteworthy, protein yield shows more or less the same trend as protein and gluten content. Therefore, since the grain yields do not differ significantly between all treatments, the increase in protein yields is due solely to the positive effects of BFAs on the grain protein content (Table 3).
In contrast, no statistical differences among treatments for gluten index were found (Table 3), thereby confirming that technological quality does not seem to be negatively affected by replacing N-fertilization with BFAs. Interestingly, Leogrande et al. [33] found a significant difference between a commercial organic fertilizer and a compost for gluten index. However, since their study was carried out under organic conditions, it was not possible to compare the effects of biofertilizers (i.e., organic fertilizer and compost) with conventional mineral fertilization. As for yellow index of milled whole meal, a statistically significant difference among treatments was found. This result is in agreement with Pasqualone et al. [28] who found a browning effect on whole meal at increasing levels of compost application. However, the specific effect of BFAs on this quality trait must be thoroughly evaluated in the oncoming cropping seasons.
The positive correlation found both for biologic and productive data suggests that the contribution given by BFAs to the plant development can likely result in a better grain quality. Conversely, only test weight seems to be positively correlated both to N-fertilization and no fertilization with respect to BFAs. Nevertheless, given the general high values of test weight, replacing ordinary fertilization with BFAs does not seem to prevent from the large-scale use of these amendments.

Materials and Methods
This study was carried out in the AGRIS experimental farm of San Michele, southern Sardinia (Ussana, Italy) (Lat. 39 • 24 N; Long. 09 • 05 E; 114 m ASL) in 2019. The farm is located in the Campidano plain, a major rainfed durum wheat growing area of Sardinia.
The climate of the study area is typically Mediterranean, with warm and dry summers and mild winters. Mean temperature values range from 4.8 • C in January to 33.0 • C in August. Total mean annual rainfall of the area is 450 mm, mostly concentrated between autumn and early spring. During the experimental period (October 2018-June 2019) we observed on average a similar trend for maximum temperatures and lower values for minimum temperatures with significant reduction of precipitation on February and March followed by heavy rain events on May when growth and development of wheat were already completed ( Figure 2). The positive correlation found both for biologic and productive data suggests that the contribution given by BFAs to the plant development can likely result in a better grain quality. Conversely, only test weight seems to be positively correlated both to N-fertilization and no fertilization with respect to BFAs. Nevertheless, given the general high values of test weight, replacing ordinary fertilization with BFAs does not seem to prevent from the large-scale use of these amendments.

Materials and Methods
This study was carried out in the AGRIS experimental farm of San Michele, southern Sardinia (Ussana, Italy) (Lat. 39°24′ N; Long. 09°05′ E; 114 m ASL) in 2019. The farm is located in the Campidano plain, a major rainfed durum wheat growing area of Sardinia.
The climate of the study area is typically Mediterranean, with warm and dry summers and mild winters. Mean temperature values range from 4.8 °C in January to 33.0 °C in August. Total mean annual rainfall of the area is 450 mm, mostly concentrated between autumn and early spring. During the experimental period (October 2018-June 2019) we observed on average a similar trend for maximum temperatures and lower values for minimum temperatures with significant reduction of precipitation on February and March followed by heavy rain events on May when growth and development of wheat were already completed ( Figure 2). The experimental field is located in a hilly environment, typical of the low-fertility, rainfed durum wheat growing areas of Sardinia and many parts of the Mediterranean basin. The only cultivar used for this experiment is Karalis, a leading durum wheat cultivar in the last 10-15 years in Sardinia. The experimental field is located in a hilly environment, typical of the low-fertility, rainfed durum wheat growing areas of Sardinia and many parts of the Mediterranean basin. The only cultivar used for this experiment is Karalis, a leading durum wheat cultivar in the last 10-15 years in Sardinia.
A pedological characterization of the site was carried out; the profile ( Figure 3) was described (Table 4), samples collected and characterized [35] ( Table 5).
The new BFAs tested in this study originate from a physical method based on the use of radio frequencies (microwaves). This non-biological process can be applied both on animal and vegetal by-products, does not provide a fermentation phase and is capable of rapidly changing the highly degradable organic biological material into a stabilized and sanitized product without risk of contamination. Further, by applying such a high MW source, the electromagnetic reaction is so strong to break up in small fragments the bigger macromolecules of humic and fulvic acid chains; this modification event determines a "short chained" molecular structure phenomenon, a special structural feature that renders these molecules easier to be uptaken by the plant. Treatments (see descriptions on Table 1 and chemical data on Table 6) were applied following a randomized complete block design with 3 replications. In the field trial each plot was 5.9-m-long and 1.5-m-wide with an approximate surface of 10 m 2 and rows spaced 0.18 apart (Table 7 and Figure 4). Plant density at sowing was about 300 viable seeds m −2 .
The transformation process of organic matter occurs within special machines called "fast organic matter degradation reactors" and is completed in about 1 hour, unlike the 60 − 90 days required by a natural fermentation process based on biological degradation (composting), greatly accelerating the BFA recovery and fertilizer production process. The microwave (MW) treatment acts in a dose-dependent manner and is destructive both for cells and bacterial populations, resulting in organic matter degradation (DNA denaturation) to reach a complete fast-humification. As a result, this rapid cell degradation leads to fungal reduction, bacterial, parasitological and virological inactivation (according to Annex IV, Chapter III "Standard transformation methods", letter G, Processing method 7 of EU Regulation No. 142/2011) up to the biostabilization of organic materials.
Further, by applying such a high MW source, the electromagnetic reaction is so strong to break up in small fragments the bigger macromolecules of humic and fulvic acid chains; this modification event determines a "short chained" molecular structure phenomenon, a special structural feature that renders these molecules easier to be uptaken by the plant. Treatments (see descriptions on Table 1 and chemical data on Table 6) were applied following a randomized complete block design with 3 replications. In the field trial each plot was 5.9-m-long and 1.5-m-wide with an approximate surface of 10 m 2 and rows spaced 0.18 apart (Table 7 and Figure 4). Plant density at sowing was about 300 viable seeds m −2 .  Before the harvest, the following biologic traits were measured to assess the effects of treatments on plant growth and development: heading date (number of days from planting), plant height (cm) and average spike weight at maturity (g). Between booting and flowering, high resolution aerial imagery were collected using an Unmanned Aerial Vehicle (UAV esacopter, model HYPERION by Dronelab, Tortolì, Italy) implemented with a multispectral camera, a Parrot SEQUOIA, that can measure four bands separately: green (550 nm), red (660 nm), red edge (735 nm) and near infrared (NIR−790 nm). Aerial imagery was processed and the normalized difference vegetation index (NDVI) [31] was obtained from red and NIR bands (Figure 1) based on the following equation: Subsequently, the NDVI map was georeferenced using an open source geographic information system [37] and sample areas, consisting of approximately 4000 pixels each, were created for each plot to evaluate the levels of wheat vigor.
Grain yield was estimated at harvest by a plot combine and, after the harvest, grain samples from each plot were analyzed to evaluate the grain quality. Protein content (%), gluten content (%), test weight (kg hL −1 ) were measured using a Foss Infratec TM 1241 grain analyzer (Foss, Hilleroed, Denmark). The whole meal samples were analyzed using a color reader CR-10 Konica Minolta for color measures. Gluten index was determined following the ICC standard method No. 158 by using the Glutomatic 2200 system (Perten Instruments AB, Huddinge, Sweden). Data were statistically analyzed by ANOVA using the GenStat 19 th edition [38]. Multiple comparison means were made using the Fisher's LSD method (p ≤ 0.05). Before the harvest, the following biologic traits were measured to assess the effects of treatments on plant growth and development: heading date (number of days from planting), plant height (cm) and average spike weight at maturity (g).
Between booting and flowering, high resolution aerial imagery were collected using an Unmanned Aerial Vehicle (UAV esacopter, model HYPERION by Dronelab, Tortolì, Italy) implemented with a multispectral camera, a Parrot SEQUOIA, that can measure four bands separately: green (550 nm), red (660 nm), red edge (735 nm) and near infrared (NIR−790 nm). Aerial imagery was processed and the normalized difference vegetation index (NDVI) [31] was obtained from red and NIR bands (Figure 1) based on the following equation: Subsequently, the NDVI map was georeferenced using an open source geographic information system [37] and sample areas, consisting of approximately 4000 pixels each, were created for each plot to evaluate the levels of wheat vigor.
Grain yield was estimated at harvest by a plot combine and, after the harvest, grain samples from each plot were analyzed to evaluate the grain quality. Protein content (%), gluten content (%), test weight (kg hL −1 ) were measured using a Foss Infratec TM 1241 grain analyzer (Foss, Hilleroed, Denmark). The whole meal samples were analyzed using a color reader CR-10 Konica Minolta for color measures. Gluten index was determined following the ICC standard method No. 158 by using the Glutomatic 2200 system (Perten Instruments AB, Huddinge, Sweden). Data were statistically analyzed by ANOVA using the GenStat 19 th edition [38]. Multiple comparison means were made using the Fisher's LSD method (p ≤ 0.05).

Conclusions
Even though based on a 1-year trial only, BFAs fertilization resulted in taller plants with greater vigor and heavier spikes as well as significant differences concerning grain quality (i.e., protein content, gluten content, protein yields and yellow index) when compared to the control (no fertilization) and N-fertilization. Furthermore, despite lower values of test weight in correspondence of BFAs fertilization were registered, all plots treated with BFAs showed high values of this trait, thereby confirming the positive effect of these amendments on grain quality. Interestingly, no statistically significant differences on grain yield were found.
These preliminary results suggest that replacing ordinary fertilization with BFAs, with or without applying conservation tillage practices, can be an effective RMP for improving soil fertility, as well as crop quality and yield stability in Mediterranean conditions. Although these potentially positive effects, many aspects need further studies: (1) the medium and long term evolution of physical-chemical and microbiological soil fertility; (2) the medium and long term effects on the evolution of the SOM; (3) the role of different cropping systems in CO 2 sequestration; (4) the medium and long terms effects of BFAs on durum wheat production (yield and quality) with a specific attention on technology (i.e., yellow index and protein content).
In order to give an answer to some of these open questions, BFAs will not be applied on soil in the next two cropping seasons to assess their residual effect on soil fertility as well as on durum wheat yield and quality.