Nitrogen Budget and Statistical Entropy Analysis of the Tiber River Catchment, a Highly Anthropized Environment

Round 1

Reviewer 1 Report

Dear authors ,the draft paper you presented addresses a question of interest far beyond the agricultural comunity: the role of agricultural production versus civilian population in the environmental nitrogen pollution. The paper is of great interest to me, as it combines methods to evaluate nitrogen flows in large catchments, determine  N efficiency and detect dispersion of N in environment (budgeting and statistical methods). So I expect to find an added value to the single methods. This added value need to be discussed further in the discussion part. The paper needs to be finalized with a conclusion. 

Below, you find some details

line 31 ff, Introduction

Detailed introduction with many references, although many references before 2010

line 113 ff, Material and methods

line 142: Table 1:  the headings are misleading: Total N Input/UAA (tN/yr) and Total N Output/UAA (tN/yr): should be transferred to Total N Input/UAA*year (Nt/(ha*yr)) and Total N Output/UAA*year (Nt/ha*yr)), Futher, there seems to be a miscalculation in the column Total N Input/UAA (tN/yr)

line 143: Data collection:

statistical data refer mostly to the census of 2010 and animal categories defined according to publications from 2007 and 2016.

However, N losses due to ammonia volatilization refers to literature from 1997 (van der Weerden and Jarvis) and 1998 (Bussink and Oenema). As technical progress may have taken place in the meantime, an update of coefficients should be seriously considered, as ammonia volatization meanwhile may be reduced. Further, 60 % of ammonia deposition is calculated as being re-deposited, according to Asman (1998). Is this share of ammonia considered by calculation of the N in atmospheric deposition?

The rate of denitrification is taken from a reference of 1999 (Smil), here as well, an update of reference seems necessary.

How information on runoff and leaching are derived from Velthoff et al (2009) and transcribed in the calculation is not clearly explained.

line 231: Figure 2: please explain: what is meant with „data normalized to 1 kgN/ha y antropogenic input? What is antropogenic? Dimension should be kg N/(ha*year)?

Definition NUE = nitrogen use efficiency versus „real N efficiency“

Spelling: NO3

No information is given on the water quality (nitrates in groundwater and rivers).

line 265: Results

(line 266 to 268 should be deleted)

line 278: Table 3: it is not clear to which soil cover some of input- and output variable refer to. Does crop uptake refer to agricultural crops or all plants cultivated within the Tiber area? Biological fixation seems to refer to agricultural and natural vegetation. However, it would be helpful to get more information in this respect.

What is meant with preserving capacity? %tage of Nsurplus in relation to Ninput? But there is little leaching and denitrification; but these figure can be questioned, as they are – as ammonia volatilization – calculated from standards.

line 314 ff Figure 5 and 6 (Random Forest Analysis). 

Wording: is overcome used in the sense of exceed?

The figure shows main predictors besides N load due to civil sewage! It is not clear what is summerized under the 5 categories shown in the figure , e.g. what is meant by non biological fixation?

With respect to the liability of the calculation data (as being derived as a standard share from standard data), results of the random forest analysis need to be discussed more thoroughly. E.g., Information on (symbiontic and non-symbiontic) biological N-fixation is not reliably, as fixing capacities depend on many factors.

line 3l4l1 ff Figure 7: Related to this figure in the text is mentioned, that UAA is an important variable associated with total N load. The authors did however not follow this thread investigating further on efficiency of agriculturally used N in the different sub-catchments.  

In fact, for an expert with agricultural background it is not easily comprehensible that the evaluation of N efficiency is not executed with focus on the agricultural part of the land surface (in this case, N loads could be compared with findings from other articles on agricultural performance). 

line 391: Table 3: NUE-definition: which Nin and Nout is referred to; the figures from Table 1?

line 429 ff: Figure 8 and 9: although these figures represent the main topic of the paper, there is not much explanation.

line 459: Discussion: this chapter is doesn’s seem complete.

What is urgentliy missing is a comparison with results of other authors, with reference to the methods applied, the added value of different methods and further reseach needs. 

E.g, what is the role of agriculture in relation to the sewages from the population in the pollution of the Tiber river system and the groundwaters of the catchment?

In addition, the reliability of data (standards) should be discussed.

The results need to be compared to results obtained in other catchments, with same or other methods. If the catchment really shows a low retention capacity for N, with respect to the high efficiencies, does it matter?

e.g., Tanzer and Rechenberger (2020), Sobantka et al. (2013)

The Conclusion could include further research needs, also methodological improvements for future applications.

English: some expressions seem misleading, such as N species, migration of nitrogen, to overcome, to input

Author Response

The authors thank the comments and observations of the reviewer 1 which certainly made an added value to the study.
The authors' responses to the observations made by reviewer 1 are in red and represented in the text in bold type.

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript describes a study, based on secondary data, to determine a nitrogen budget for the highly anthropized Tiber River catchment in Italy. The study also undertook a Statistical Entropy Analysis to evaluate the nitrogen efficiency of the river and its sub catchments to consider the distribution of nitrogen into the environment.

This is an interesting article that fits within the scope of the journal. Use of the English language is adequate but there are a few issues that need to be addressed. Paragraphing is a serious issue, with many paragraphs being very long and could easily be broken up to improve readability. There are also a few typos, for example in the abstract at line 14 there is a stray ‘s’ and there is superfluous text between lines 266 and 268 which appear to come from the journal guidance. There are a few other examples in the text. Please check carefully.

The introduction is interesting and provides some good background with a strong rationale for the work but it is far too long, too detailed and contains a lot of unnecessary information. I suggest that this section is slimmed down considerably and properly paragraphed.

In contrast the Materials and Methods section does not contain enough detail. Sufficient information should be included to allow the study to be duplicated. The following issues need to be addressed:

- Line 121. Please explain what the mentioned database is and what it is used for.

- Line 129-141 is all of this text strictly relevant?

- Table 1 what is the source of the data presented in this Table?

- Line 145 the Soil System Budget needs to be briefly explained.

- Line 149 what is meant by ‘released in the frame…’?

- Line 150 is 2010 data the most recent available? It is almost 12 years old. How does the age of this data affect the study findings?

- Line 152 what exactly is meant in this context by ‘nitrogen unit’. Please be precise.

- Line 153 what exactly are the site-specific coefficients?

- Line 155 what is the source of the ‘accessory  information’?

- Line what are the animal categories, what typology was used, where did the data come from?

- Line 161-164 where did all of this data come from?

- Paragraph between Lines 170 and 180 needs greater explanation. Give a summary of the GAINS model,

- Line 170 the hyperlink should be converted to a reference.

Please ensure non-main stream terms are fully explained e.g. Line 197 ‘out-of-bag' data.

Figure 2 is not of sufficient quality for publication. It is rather blurred and some of the text is hard to read.

The Results section is interesting but is again far too long and contains some unnecessary text. Some of the text in this section might be better moved to the Discussion section. Please stick to the highlights. It is unclear why reference 76 at line 170 is underlined.

Author Response

The authors thank the comments and observations of the reviewer 2, which certainly made an added value to the study.
The authors' responses to the observations made by reviewer 2 are in red and represented in the text in bold type.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Authors,

Your contribution is interesting, but needs for corrections, explanations and broader interpretations.

Some details are herewith below.

Line 14

...matter, s by leaching...

Please correct.

Line 59

This risk is augmented by the large use of water for irrigation...

This is an important statement, but the authors did not discuss it in the text.

Lines 182-184

Runoff and leaching were calculated as the 4% of manure application and synthetic fertilization and the 18% of Surplus corrected for runoff (mean value for Italy derived from Velthof and co-workers), respectively.

You mean the runoff and leaching of what? The explanation is too fragmentary.

Line 223

...dioxide and nitrogen oxides (?₂?, ??_?).

?₂? is named Nitrous oxide. Nitrogen dioxide formula is ??₂.

Line 266-268

This section may be divided into subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experi268 mental conclusions that can be drawn.

The sentence seems to be a remark of the reviewer?

Line 279
... it was evident a low preserving capacity (15.83%)

Please make the evidence to become clear. The same to other figures in this and the next paragraphs as well.

Line 330

...the significant and negative correlation between N balance and non-symbiotic N fixation (r = -0.379...

What is a criterion to assess this level of correlation as significant? Please use the sign "minus".

Lines 330-333

...N outputs exceed N inputs when non-symbiotic N fixation was reduced by high availability of N at the soil (that is total N losses were low). This high availability of N in the soil should be due to the agricultural practices that enriched the soil with nitrogen through its fertilization.

This is quite understandable qualitative statement. What about the influence of the variables on the result of quantitative calculations? More details are needed.

Lines 480-484

 It should be a best practice the reduction of the N-based chemical fertilizers at aiming to increase the percentage of non-symbiotic and biological N fixation and to reduce N loads through a reduction of the livestock number, to practice crop rotation, and to maximize the use of crops species able to increase the efficiency of nitrogen uptake.

This statement is categorical, but is doubtable to the some extent. Reduction of the N-based chemical fertilizers – yes, this is a measure, indeed. If you reduce the livestock number, then how should you feed the people? If you wrote "to practice crop rotation", then there is no crop rotation in the Tiber catchment? This reduces the value of your research. It is figuring out that a task of a Tiber catchment N balance optimization is simple – steady use the old agricultural recommendations. 

"...maximize the use of crops species able to increase the efficiency of nitrogen uptake" - it is a declaration. Please indicate which species did you mean, and how and whom will apply these species. This is important.

Lines 486-490

The surplus calculation may be considered a good indicator of the environmental impacts due to agricultural practices, but information about the type of N that is released into the environmental matrices and on nitrogen compounds and load variations allows for a more detailed analysis of N input and output fluxes, enabling more effective management actions.

What is a semantic load of this sentence? In its initial part, you put forward considerations about nitrogen, then use the word "but" to contrast this part with the second part of the sentence, but in reality the opposition did not work out.

Lines 490-492

In conclusion, RSE analysis was a very useful method to evaluate the system’s efficiency and dispersion of various nitrogen compounds into the systemic environmental matrices.

In conclusion, it should be useful to give a word about Nitrogen budget, Random Forest Analysis, and Soil type distribution at the catchment and sub-catchment level. Or these sections of your paper you assess as not important? A word about the future land use change needed would be important.

Author Response

The authors thank the comments and observations of the reviewer 3, which certainly made an added value to the study.
The authors' responses to the observations made by reviewer 3 are in red and represented in the text in bold type.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors, 
the comprehensibility of the text has improved, but still there remain some points unclear. Further, it is definitely necessary to compare results in the discussion with those from other scientists, even though the combination of NUE at catchment level and SEA is new.  A conclusion would be a good opportunity to point out further research needs, also improved data requirements.
I will point out some major remaining obstacles:
line 146:  statistical data for 2010 reveal for the agriculturally used areas a very high efficiency, in all sub-catchments; it would be helpful to know what inputs are included and if the inputs are gross or net data, with or without reference to gaseous N emission. The link of this soil surface budget to the soil system budget is not clear, are there other main factors or is it just the larger area, comprising of UAA and non-UAA? Comparing the figures of Table 1 and 2 I have the impression that the data basis is the same, but the reference area is the catchment. 
line 290: In relation to soil system budget, you refer to the population in the different catchments, but it is not clear how this effects the budgets or other results. Sewage (sludge) or waste water is not considered in the budget. There is no information if/how wastewater is treated and whether/how it is included in the budget (you refer to it later, Figure 3, line 311). In case wastewater is collected and treated more or less centrally, outlet of the waste water treatment plant would affect water quality of the rivers. Otherwise, direct discharge of wastewater in rivers would pollute the rivers.
As I understand your approach, the budget covers the whole catchment area. The share of biological fixation seems to me quite high with respect to fertilized crop. Are these figures from the census?
The percentages of leaching and runoff were taken from Velthof et al. [66] as national average figures. Small variation from this average will affect N retention. Retention to my understanding means that N remains in the soil. 
line 333: it is unclear what is meant by N losses from non-nitrogen-fixing plants: in my understanding, these plants rather consume N for their growth (N consumption?)
line 339: You refer to the influence of the number of residents and the amount of sewage sludges, which have no significant effect on the N retention. I have not yet understood how this in reality should happen (as retention refers to soil and not to the rivers of the catchments), and how you include this in your analysis. 
line 398: influence of soil type distribution is not clear to me as runoff and leaching are determined as standards. At which point soil type enters the analysis? 
line 462: wording: RSE_product; what is meant, the crop uptake, or are there other products to be considered?
line 470: as leaching and runoff and gaseous N-emissions are defined as depending variables (e.g., Velthof et al,) it is not so surprising that a change of the input-variables “mineral or organic fertilizers” would have an effect. 
line 494: you deduce, that a reduction of 50% in N loads would allow for a better system performance. But “real” fertilization levels in this example are not so high, whereas “derived, virtual” N-fixation rates seem not realistic. To the findings of your analysis, it is therefore necessary to add results from corresponding research and compare the results.

Author Response

Please, see the attachment.

Author Response File: Author Response.pdf