The Terraces of the Anti-Atlas: From Abandonment to the Risk of Degradation of a Landscape Heritage

Round 1

Reviewer 1 Report

The manuscript is perfectly done and wrote. this is the first ever in this slandered since I start to review manuscripts about 20 years ago.   well done

Author Response

Comments and Suggestions for Authors

The manuscript is perfectly done and wrote. this is the first ever in this slandered since I start to review manuscripts about 20 years ago.   well done

Thank You for your efforts. Your comments are encouraging.

Submission Date

15 October 2020

Date of this review

16 Oct 2020 13:29:36

Reviewer 2 Report

The terraces of the Anti-Atlas: from abandonment to the risk of degradation of a landscape heritage.

 

The author brings up a very important environmental issue regarding soil degradation as a result of abandonment, list of methods and analysis shows a significant degradation following the first year, and it continued to the fifth year. Differences between 5th and 20 years were less significant. Nonetheless soil filtration remarkably decreed to more than 80% following 5 years of terraces abandonment. However the manuscript is missing ideas and methods that could help in terraces restoration, and to what extend these soils can be restored, some recommendation should be addressed based on literature. 

Abstract

Line 16-17 please rewrite, what you mean by the “to the reduction of its opening and to

its compaction”

 

Line 17 please add standard deviation

 

Line 20-21 The time spans represented in the abstract are varied and hard to compare, “More than 11% of the surfaces are severely degraded after 20 years of abandonment” what was the percentage after 1 and 5 years? 

Introduction

 

Line 94 “galleries”?

 

Line 124 duplicate word “Ravines”

 

Line 134         “The rest being abandoned”, incorporate it with the previous sentences.

 

Line 135 “Just over three percent (3.3%) were discontinued in the year (2017-18)” please rewrite the sentence. What is that was discontinued? 

 

Line 182 duple dots. 

Author Response

Point 1. General comment

The author brings up a very important environmental issue regarding soil degradation as a result of abandonment, list of methods and analysis shows a significant degradation following the first year, and it continued to the fifth year. Differences between 5^th and 20 years were less significant. Nonetheless soil filtration remarkably decreed to more than 80% following 5 years of terraces abandonment. However the manuscript is missing ideas and methods that could help in terraces restoration, and to what extend these soils can be restored, some recommendation should be addressed based on literature. 

See conclusion:

These heritage spaces require special attention to their rehabilitation and conservation. It can be argued that spaces not yet abandoned (24.7%) can be preserved. The strengthening of intergenerational links through training and education (schooling), the promotion of local know-how, the search for crop varieties adapted to climate change (less demanding in terms of water), and the establishment of a solidarity mechanism, such as payment for ecosystem services, between the mountain producing water and the plain using this essential resource for the sustainability of new intensive agricultural systems oriented towards the export of fruits and vegetables to Europe and America. The role of women is very important in the management of these landscapes. It must be strengthened by promoting their emancipation. These landscapes are part of the “Arganeraie” biosphere reserve. They could constitute a research and training laboratory for Ibn Zohr University. Scientific communities and civil society have a lot to do in these mountainous and heritage areas. The establishment and development of museums on cultural heritage in the RBA provided for by National Agency for the Development of Oasis and Argan Zones is an opportunity to be exploited.

Abstract

Point 2.

Line 16-17 please rewrite, what you mean by the “to the reduction of its opening and to its compaction”

The abandonment of the terraces led to the reduction of the plant cover of the soil surface and its opening and to enhance its compaction.

Point 3.

Line 17 please add standard deviation

The risk of runoff is high. A year of abandonment reduced infiltration from 301.8 mm/h (Sd: 105.8 mm/h) to 129.6 mm/h (Sd: 28.9 mm/h). A 5 years of abandonment reduce it to 62.9 mm/h (Sd : 14.9 mm/h).

 

Point 4.

Line 20-21 The time spans represented in the abstract are varied and hard to compare, “More than 11% of the surfaces are severely degraded after 20 years of abandonment” what was the percentage after 1 and 5 years? 

More than 11% of the surfaces are severely degraded after 50 years of abandonment.

This means that currently areas abandoned for 50 years are in a critical state of erosion with a soil area factor of FSS of 69.5. The results show that only this category of terraces that has a critical degree of degradation. The other abandonment ages have medium and low grades.

Introduction

Point 5.

Line 94 “galleries”?

Underground passage, subterranean passage. (https://www.lexico.com/definition/gallery)

Insects living in the soil dig galleries to circulate there.

Point 6.

Line 124 duplicate word “Ravines”

It consists of observing the plot and giving values to the degrees of the factors linked to the erosion process, thus making it possible to classify the plots by degree of erosion: Soil movement, Surface litter, Stones on the soil surface, Witnesses of erosion, Flow model, Ravines, Gullies.

Point 7.

Line 134         “The rest being abandoned”, incorporate it with the previous sentences.

Throughout the territory of Tizerkine - Timzemzit (283.9 ha), only a quarter (24.9%) of the terraces remain maintained by the population and still used for a fairly low and very irregular crop production (grains and straw of cereals, argan fruit) and the rest being abandoned (Figure 2, Table 3).

Point 8.

Line 135 “Just over three percent (3.3%) were discontinued in the year (2017-18)” please rewrite the sentence. What is that was discontinued? 

Just over three percent (3.3%) were abandoned during the year (2017-2018).

Point 9.

Line 182 duple dots. 

and the compact of 20%. After 10 years of abandonment,

 

Submission Date                          15 October 2020

Date of this review                       22 Oct 2020 11:33:42

Reviewer 3 Report

The work appears to be interesting inhabited areas investigation, where the terraces represent a since long time the cultivation system in that area, probably the only one possible to preserve fertility, comfortable (!) territory shaping and erosion control.

No doubt are moved forward on fact that maintenance attention to the of these traditional artifacts can give chance to the territory habitability, the stability, and security of downstream territories.

Nevertheless, some doubts remain on the methodology adopted, on the presentation of the results and above all on some qualitative rather than quantitative considerations showed in this work; this does not mean that the results are inaccurate but that greater rigor is required in the application of the scientific method. And in particular:

• 60-71 Taking into account article purpose, rainfall characteristics doesn’t seemsvery well described. Is not enough in my opinion the annual amount of rain and the coefficient of variation over a (long) gauging period. Average rainfall =211mm and 54% of coefficient of variation doesn’t describe potential intense rainfall, usually responsible for runoff and erosion. Would be appreciable some mathematical descriptor as IDF curves to be compared with infiltration rate. It also must be pointed out that for dry stone wall collapse, no indication about causes is provided. If abandonment of land cultivation induce wall collapse, due to infiltration rate changing, runoff should destabilize stones from the top of the wall.
• Figure 1: maps aren’t accurate to understand elevation, convergence, slope aspect and so on. Could you provide a morphological or technical map to overlap study area?
• IN Table 1 it is curious to couple conservation state with number of years of abandonment. doesn’t exist other cases with crossed classes? As, for example, Good state of conservation in 5 years of abandonment and vice versa? The classes are probably related but not necessarily cause-and-effect and bijective as affirmed.
• Lines 86-90 as above mentioned, is the 50k topographic background with overlapped GPS surveys available?
• Lines 97-100 Pen Penetrometer provides indication on soil resistance but only on 5-10 mm of depth. When main reason of high infiltration rate and reduced compaction derive from continuous tillage for cultivation, it is risky in my opinion to extend top soil penetration measures and analysis to the entire thickness of the investigate soil.
• 124 “Ravines” it appear 2 times..
• 141–151 I do not think that emigration paragraph should be placed in Results and Discussion session. Maybe better in Introduction to highlight one of the main study reasons.
• Fig 4. Photos show some dry stone wall built, apparently, in a not very stable way. Stones seems to be distributed without a maximum stability criterion (big stones should be placed with larger side in horizontally position, arranged with regular intervals and "saturated" by small stones also placed horizontally as possible with local material). A non correct stone distribution and arrangement create wall areas with reduced resistance, “seeds” for local instability starting.
• 182 two poinst in line
• 181-188 I don’t understand where the showed data are coming from (open surface, plant cover percentage and compaction=. What does it means that the soil is “frozen”?
• Tab 4: Bare Surface, cited into the text, is not presence into the table. Wasn’t an interesting parameter?
• Tab 5: Infiltration rate are very high for cultivated areas. This data aren’t surprising taking by considering infiltration rate of a recently plowed terrace. I’m afraid that such high values are due to surface plowing that is a partial measure of a in-homogeneous soil system and scale effect of a local test is great. It is probably that during test, the water descend vertically till the not worked soil (base below plowing depth) and there goes laterally in base of lateral slope. If this scheme was confirmed, data provided does not refer to average slope hydrologic behavior (two layers: the shallow one very permeable, the lo lower much less permeable) and runoff consideration should be evaluated in this light.
• Moreover standard deviation is quite high (in a normal distribution I expect than 95% of observations in 90,2-513,4 mm/h range) that increase doubts on measuring gauging (not on right field measurement but homogeneous distribution on slopes and an enough samples magnitude). It is maybe possible one or few outlier that casually deviate from normal assumption? In fact standard deviation reduction move from 3.66 and 7.1 times in 1-5 years of non plowing is very high and suggest (in addition to a modification of the shallow hydraulic characteristics) an homogenization of analyzed area.
• Fig 4 I would appreciate boxplot for infiltration rate measurement.
• 231-231 This is realistic but not evidenced in this study. In order to prove that, it should taken into account above mentioned consideration on shallow and deep soil on hydrologic behavior. Effects of high infiltration rate for terraced system on runoff formation, moreover, is relevant in direct relation with relative extension on watershed.
• 241-243 this is a qualitative (maybe right) consideration.
• 243-245 It doesn’t appear as a results.
• Fig 5. It isn’t clear to me where data for histogram are obtained. Original data for SSF calculation are not showed into the article. If SSF came from other publications, it should be indicated.
• 272-274 This affirmation is in general acceptable but not demonstrated in ths work. Moreover I can’t understand with a such high infiltration rate. By observing Fig 4, omitting first 30’ of infiltration rate stabilization for all tests, infiltration rate move from 300 mm/h to 50 mm/h so that abandonment effect on runoff formation can be appreciated only for rainfall intensity over 50 mm/h and duration over 30’; otherwise runoff shouldn’t form. Moreover 1 year of abandonment, only an intensity rainfall over 150 mm/h and 30’ of duration (75mm of cumulative rain), can generate runoff and dry stone wall destabilization. For under 30’ rainfall intensity rainfall should be much higher.
• In my opinion a more complex hydrology governs runoff and water balance into that terraced system and shallow infiltration rate gauging can represent only the “bottleneck” of water infiltration.
• 284-286 Maybe fertility losing influence vegetation non affirmation but there is not an evidence in this work. No fertility measurement have been provided before and after abandonment or, in alternative, in still cultivated or abandoned terraces.

Author Response

Comments and Suggestions for Authors

The work appears to be interesting inhabited areas investigation, where the terraces represent a since long time the cultivation system in that area, probably the only one possible to preserve fertility, comfortable (!) territory shaping and erosion control.

No doubt are moved forward on fact that maintenance attention to the of these traditional artifacts can give chance to the territory habitability, the stability, and security of downstream territories.

Nevertheless, some doubts remain on the methodology adopted, on the presentation of the results and above all on some qualitative rather than quantitative considerations showed in this work; this does not mean that the results are inaccurate but that greater rigor is required in the application of the scientific method. And in particular:

• 60-71 Taking into account article purpose, rainfall characteristics doesn’t seems very well described. Is not enough in my opinion the annual amount of rain and the coefficient of variation over a (long) gauging period. Average rainfall = 211 mm and 54% of coefficient of variation doesn’t describe potential intense rainfall, usually responsible for runoff and erosion. Would be appreciable some mathematical descriptor as IDF curves to be compared with infiltration rate. It also must be pointed out that for dry stone wall collapse, no indication about causes is provided. If abandonment of land cultivation induces wall collapse, due to infiltration rate changing, runoff should destabilize stones from the top of the wall.

 

The maximum daily rains at Ait Baha station with return period 2 years, 10 years and 100 years are respectively 42.6 mm, 64.0 mm and 90.6 mm [10] (Daide, 2016).

 

Daide F. 2016. Modélisation Hydraulique des crues du bassin versant de l’oued Tamraght, Souss Massa. Master Sciences et Techniques : Géoressources et Environnement. Faculté des Sciences et Techniques – Fès. Université Sidi Mohammed Ben Abdellah. Maroc. Pp: 63.

 

See 280-286

The supporting walls of the terraces deteriorate more and more with the age of abandonment because of the runoff that creates the paths with the slope. The stones fall and water circulation holes are formed. With the lack of maintenance, the walls of the terraces fulfill their role of conserving soil (fertility) and water less and less.

 

 

• Figure 1: maps aren’t accurate to understand elevation, convergence, slope aspect and so on. Could you provide a morphological or technical map to overlap study area?

 

Thank You for this observation: The figure is inside the paper!

 

 

• IN Table 1 it is curious to couple conservation state with number of years of abandonment. doesn’t exist other cases with crossed classes? As, for example, Good state of conservation in 5 years of abandonment and vice versa? The classes are probably related but not necessarily cause-and-effect and bijective as affirmed.

 

I agree. The effects of land abandonment on the evolution of terraces are multiple and complex. Moreover, their impacts are not linear according to age with land degradation. In some places, the overpressure of goat grazing and tillage allowed the re-establishment of natural vegetation and the reduction of erosion processes. In other places (marly substrates), the stopping of tillage by the arar caused the closure of the surface (compaction and crust of the surface) and the accentuation of runoff and therefore erosion.

 

See methodology 2.2.1: Line 77-84

Prospecting the land with the peasants showed us that the land was abandoned in waves, during the same period and on the same slopes. The abandonment age was roughly estimated and was used as an indication for the cessation of agricultural activities on the terraced plots. Different, but approximately similar places can be found in the same age group.

At the scale of the plot cultivated in terraces, determining the age of abandonment is fairly easy for recent ages. But for a space or a slope, it is difficult. We made an approximation.

 

 

• Lines 86-90 as above mentioned, is the 50k topographic background with overlapped GPS surveys available?

The locations surveyed with the farmers were referenced by GPS and transferred to a 1/50000 topographic map. For some places, a sketch was made in the field on topographic map.

 

• Lines 97-100 Pen Penetrometer provides indication on soil resistance but only on 5-10 mm of depth. When main reason of high infiltration rate and reduced compaction derive from continuous tillage for cultivation, it is risky in my opinion to extend top soil penetration measures and analysis to the entire thickness of the investigate soil.

 

In these arid areas and on soils poor in organic matter (<1%) and with a silty texture, it has been shown that the compaction of the first few centimeters of the surface (<2 cm) governs the infiltration of water into the soil (Sabir, 1994; Sabir et al., 2004; Roose et al., 2010). The measurement of PEN by a pocket penetrometer is a good indicator which can be easily measured.

 

• 124 “Ravines” it appear 2 times..

 

It consists of observing the plot and giving values to the degrees of the factors linked to the erosion process, thus making it possible to classify the plots by degree of erosion: Soil movement, Surface litter, Stones on the soil surface, Witnesses of erosion, Flow model, Ravines, Gullies.

 

• 141–151 I do not think that emigration paragraph should be placed in Results and Discussion session. Maybe better in Introduction to highlight one of the main study reasons.

 

You are right. However, emigration is both a cause and a consequence of the degradation of the terraces. Several authors, and recently in an interesting paper, Boseli et al. (2020) have shown that human factors (emigration, lack of intergenerational communication) are the most relevant in the dynamics of terracing systems and that natural factors (water availability) are less involved.

 

• Fig 4. Photos show some dry stone wall built, apparently, in a not very stable way. Stones seems to be distributed without a maximum stability criterion (big stones should be placed with larger side in horizontally position, arranged with regular intervals and "saturated" by small stones also placed horizontally as possible with local material). A non correct stone distribution and arrangement create wall areas with reduced resistance, “seeds” for local instability starting.

 

You are right. The walls of the terraces do not really follow the rules of construction with dry stones. Several authors have mentioned it in the High and Anti-Atlas (Ziyadi M. 2001; Roose et al., 2010). Good wall construction to support terraces requires a lot of expertise and especially hard work: collecting and calibrating stones, arranging in the wall for good stability, etc. (Sabir, M., El-Khoury D. L., Salman, M. 2020. Field guide for hill land reclamation and water management. Beirut, FAO.  http://www.fao.org/3/ca8381en/CA8381EN.pdf).

 

• 182 two poinst in line

 

and the compact of 20%. After 10 years of abandonment,

 

• 181-188 I don’t understand where the showed data are coming from (open surface, plant cover percentage and compaction=. What does it means that the soil is “frozen”?

 

See the methodology: 2.3. Characterization of soil surface conditions.

Open surface means that it is made up of loose aggregates, that it has not covered with a crust. The cover represents the part of the surface covered with vegetation and litter. The compaction of the surface is measured by the PEN and the bulk density of the soil.

A “frozen” soil is a vernacular qualifier given by farmers to soils that are compacted and covered with crust.

 

 

• Tab 4: Bare Surface, cited into the text, is not presence into the table. Wasn’t an interesting parameter?

 

The proportion of the bare surface and that of the covered surface is 100% (Bare surface + Covered surface = 100%). A surface, it is either covered or bare. And so, we preferred to use only the% of covered area. See the methodology.

 

• Tab 5: Infiltration rate are very high for cultivated areas. This data aren’t surprising taking by considering infiltration rate of a recently plowed terrace. I’m afraid that such high values are due to surface plowing that is a partial measure of a in-homogeneous soil system and scale effect of a local test is great. It is probably that during test, the water descend vertically till the not worked soil (base below plowing depth) and there goes laterally in base of lateral slope. If this scheme was confirmed, data provided does not refer to average slope hydrologic behavior (two layers: the shallow one very permeable, the lo lower much less permeable) and runoff consideration should be evaluated in this light.

 

What you are describing here is quite true for the crop profiles of a sloping land with two distinct layers: the surface (15-20 cm) plowed is porous and the lower compact and less porous. The water flow is lateral (hypodermic). It is also observed in superficially plowed agricultural land (with 20 cm plow) which gives a plow base.

In this case at the Anti-Atlas, the terraces are almost horizontal. And so the transfer of water (flux) from surface to depth is essentially vertical. The outer double ring plays a protective role to prevent side flow. Runoff is generated mainly by the conditions of the soil surface: If the surface is open water seeps easily and if the surface is closed and compacted, even if the soil is still dry, runoff occurs. Runoff appears more quickly during thunderstorms and heavy rain events.

 

• Moreover standard deviation is quite high (in a normal distribution I expect than 95% of observations in 90,2-513,4 mm/h range) that increase doubts on measuring gauging (not on right field measurement but homogeneous distribution on slopes and an enough samples magnitude). It is maybe possible one or few outlier that casually deviate from normal assumption? In fact standard deviation reduction move from 3.66 and 7.1 times in 1-5 years of non plowing is very high and suggest (in addition to a modification of the shallow hydraulic characteristics) an homogenization of analyzed area.

 

On very heterogeneous soils on the surface (coarse elements, fine soils) the infiltration capacity is quite variable. The semivariograms show that the lag is less than 2 m. The standard deviations are high. When these surfaces undergo transformations making them more homogeneous from a surface condition point of view (soil compaction, crust), the surface layer governs the infiltration capacity. Therefore, the variability of the infiltration decreases. We checked the results, there are no outliers.

 

• Fig 4 I would appreciate boxplot for infiltration rate measurement.

 

Table 5 gives the values of the standard deviations and coefficient of variation for the final infiltration rates. These are the averages of the last two values obtained at 40 and 45 minutes. It would be difficult to make boxplot for all points of the 5 curves, mainly 5, 10 and 20 years of retirement.

 

• 231-231 This is realistic but not evidenced in this study. In order to prove that, it should taken into account above mentioned consideration on shallow and deep soil on hydrologic behavior. Effects of high infiltration rate for terraced system on runoff formation, moreover, is relevant in direct relation with relative extension on watershed.

 

It is absolutely true. The study did not measure runoff. We were tempted to work by rainfall simulation. But due to the difficulties in the field and the lack of resources, we opted for the double rings method, just measuring the infiltration. The runoff deduction is made because leaving terraces reduces infiltration.

We propose to write this:

With the reduction of infiltration capacity, the abandoned terraces over large areas are believed to be responsible for the increasingly flashing, intense and frequent floods in the Anti-Atlas and the surrounding plains.

 

• 241-243 this is a qualitative (maybe right) consideration.

This is true.

 

• 243-245 It doesn’t appear as a results.

We remove "result" and we propose to write:

Taking this into account, the Moroccan Government, with the support of the UNDP, has developed a payment mechanism for ecosystem services (PES) based on the maintenance and planting (almond trees) of the terraces in the Western Anti-Atlas [25].

 

• Fig 5. It isn’t clear to me where data for histogram are obtained. Original data for SSF calculation are not showed into the article. If SSF came from other publications, it should be indicated.

 

The qualification of the SSF factor is done through sheets in the form of a matrix of several sub-indicators (6) ranging from the description of the state of the surface (cover) to the density of erosion forms, etc. in each site (See methodology). It is very long (several pages) and laborious. The SSF factor is a synthesis of these sub-indicators.

 

• 272-274 This affirmation is in general acceptable but not demonstrated in ths work. Moreover I can’t understand with a such high infiltration rate. By observing Fig 4, omitting first 30’ of infiltration rate stabilization for all tests, infiltration rate move from 300 mm/h to 50 mm/h so that abandonment effect on runoff formation can be appreciated only for rainfall intensity over 50 mm/h and duration over 30’; otherwise runoff shouldn’t form. Moreover 1 year of abandonment, only an intensity rainfall over 150 mm/h and 30’ of duration (75mm of cumulative rain), can generate runoff and dry stone wall destabilization. For under 30’ rainfall intensity rainfall should be much higher.

 

This could be true if we had measured the infiltration capacity under rainfall simulation. But with the double-ring method, which forces infiltration under a water head of 100 mm, the hydraulic conductivity at unsaturation is measured (Darcy's law).

The stormy rains that generated a flood that caused a lot of damage in the area in 2015 upstream of the Oued Souss watershed had between 122 and 150 mm in 48 hours. It turns out that it is the surface condition that determines the generation of runoff.

 

• In my opinion a more complex hydrology governs runoff and water balance into that terraced system and shallow infiltration rate gauging can represent only the “bottleneck” of water infiltration.

 

It is totally true. As said above, the rain simulation could make more sense to the dynamics of water infiltration into the soil on these abandoned terraces. May be in the future with young scientists.

 

• 284-286 Maybe fertility losing influence vegetation non affirmation but there is not an evidence in this work. No fertility measurement have been provided before and after abandonment or, in alternative, in still cultivated or abandoned terraces.

 

We don't have any measurements of soil fertility. We have referred to the work of [19] Aziki (1983).

Aziki S., 1983. Irrigated agriculture and the development of a rapidly changing southern Moroccan rural space: the case of Souss-Aval. 3rd Cycle Geography thesis, University of Aix Marseille 2, 387 p.

 

 

Submission Date

15 October 2020

Date of this review

06 Nov 2020 00:03:49

Round 2

Reviewer 3 Report

I appreciated the author's effort to adapt the article to the requests for improvement.
However, I must point out that doubts already expressed and not resolved persist, mainly relating
to considerations on surface, medium and deep water circulation (which affect the surface runoff),
which do not seem supported by only the infiltration measures carried out. The study of water
circulation should be deepened (Tab 5) and compared at least with an adequate rainfall descriptor
(IDF curves are suggested) especially for return times compatible with the years of abandonment
of the terraces.
Erosion measures (and not estimates) would also be appropriate to make certain statements less
qualitative and more quantitative, reducing their uncertainty.

Author Response

Responses to reader 3:

All my responses are in this file. No information is added to the paper.

 

Comments and Suggestions for Authors

The work appears to be interesting inhabited areas investigation, where the terraces represent a since long time the cultivation system in that area, probably the only one possible to preserve fertility, comfortable (!) territory shaping and erosion control.

No doubt are moved forward on fact that maintenance attention to the of these traditional artifacts can give chance to the territory habitability, the stability, and security of downstream territories.

Nevertheless, some doubts remain on the methodology adopted, on the presentation of the results and above all on some qualitative rather than quantitative considerations showed in this work; this does not mean that the results are inaccurate but that greater rigor is required in the application of the scientific method. And in particular (line order follows not revised version):

60-71 Taking into account article purpose, rainfall characteristics doesn’t seems very well described. Is not enough in my opinion the annual amount of rain and the coefficient of variation over a (long) gauging period. Average rainfall =211mm and 54% of coefficient of variation doesn’t describe potential intense rainfall, usually responsible for runoff and erosion. Would be appreciable some mathematical descriptor as IDF curves to be compared with infiltration rate.

To carry out a frequency analysis of precipitation intensities, we need measurements at short time steps of the intensity (mn, hour). However, these data do not exist. This is why we used the annual mean value and the coefficient of variation to characterize the precipitation.

It also must be pointed out that for dry stone wall collapse, no indication about causes is provided. If abandonment of land cultivation induces wall collapse, due to infiltration rate changing, runoff should destabilize stones from the top of the wall.

The damage to the walls is due to surface runoff, but also by grazing animals in the area and especially also by the lack of maintenance. The stones of the body of the wall are destabilized over time and can slip and weaken the wall. It's not just surface runoff. As can be seen from the photos (fig 3).

NOT RESOLVED

 

Figure 1: maps aren’t accurate to understand elevation, convergence, slope aspect and so on. Could you provide a morphological or technical map to overlap study area?

RESOLVED

 

 

IN Table 1 it is curious to couple conservation state with number of years of abandonment. doesn’t exist other cases with crossed classes? As, for example, Good state of conservation in 5 years of abandonment and vice versa? The classes are probably related but not necessarily cause-and-effect and bijective as affirmed.

Prospecting the land with the peasants showed us that the land was abandoned in waves, during the same period and globally on the same slopes. The abandonment age was roughly estimated and was used as an indication for the cessation of agricultural activities on the terraced plots. Different, but approximately similar places can be found in the same age group.

NOT RESOLVED

 

Lines 86-90 as above mentioned, is the 50k topographic background with overlapped GPS surveys available?

RESOLVED

 

Lines 97-100 Pen Penetrometer provides indication on soil resistance but only on 5-10 mm of depth. When main reason of high infiltration rate and reduced compaction derive from continuous tillage for cultivation, it is risky in my opinion to extend top soil penetration measures and analysis to the entire thickness of the investigate soil.

These measurements (PEN, CV, bulk density) are to characterize the soil surface. They are not generalized over the entire depth of the soil. Several studies have shown that the compaction of the soil surface is generally done on the first cm (2, 5 cm, 7 cm). It is this part of the soil that determines its infiltration capacity.

NONE INFORMATION ADDED

 

124 “Ravines” it appear 2 times..

RESOLVED

 

141–151 I do not think that emigration paragraph should be placed in Results and Discussion session. Maybe better in Introduction to highlight one of the main study reasons.

RESOLVED

 

Fig 4. Photos show some dry stone wall built, apparently, in a not very stable way. Stones seems to be distributed without a maximum stability criterion (big stones should be placed with larger side in horizontally position, arranged with regular intervals and "saturated" by small stones also placed horizontally as possible with local material). A non correct stone distribution and arrangement create wall areas with reduced resistance, “seeds” for local instability starting.

side comment that could have been ignored

182 two poinst in line

RESOLVED

 

181-188 I don’t understand where the showed data are coming from (open surface, plant cover percentage and compaction). What does it means that the soil is “frozen”?

RESOLVED

 

Tab 4: Bare Surface, cited into the text, has not presence into the table. Wasn’t an interesting parameter?

As indicated before, bare ground + covered ground = 100%. Whether you use one or the other, it's the same thing!

NOT RESOLVED

 

TAB 5: Infiltration rate are very high for cultivated areas. This data aren’t surprising taking by considering infiltration rate of a recently plowed terrace. I’m afraid that such high values are due to surface plowing that is a partial measure of a in-homogeneous soil system and scale effect of a local test is great. It is probably that during test, the water descend vertically till the not worked soil (base below plowing depth) and there goes laterally in base of lateral slope. If this scheme was confirmed, data provided does not refer to average slope hydrologic behavior (two layers: the shallow one very permeable, the lo lower much less permeable) and runoff consideration should be evaluated in this light.

Moreover standard deviation is quite high (in a normal distribution I expect than 95% of observations in 90,2-513,4 mm/h range) that increase doubts on measuring gauging (not on right field measurement but homogeneous distribution on slopes and an enough samples magnitude). It is maybe possible one or few outlier that casually deviate from normal assumption? In fact standard deviation reduction move from 3.66 and 7.1 times in 1-5 years of non plowing is very high and suggest (in addition to a modification of the shallow hydraulic characteristics) an homogenization of analyzed area.

As indicated in the first correction, this system described by the reader exists in agricultural land cultivated on the surface by a disc plow. However, this is not the case here. The tillage is done by the hoe and the "soil" is a deposit of fine particles and large stones. No formation of an impermeable layer is observed at depth. There is no plow base.

NOT RESOLVED

 

Fig 4 I would appreciate boxplot for infiltration rate measurement

NOT IMPORTANT BUT NOT RESOLVED

 

231-231 This is realistic but not evidenced in this study. In order to prove that, it should taken into account above mentioned consideration on shallow and deep soil on hydrologic behavior. Effects of high infiltration rate for terraced system on runoff formation, moreover, is relevant in direct relation with relative extension on watershed.

With the reduction of infiltration capacity, the abandoned terraces over large areas are believed to be responsible for the increasingly flashing, intense and frequent floods in the Anti-Atlas and the surrounding plains.

This is an observation to be verified by future mapping and spatialization research of infiltration processes at the watershed scale.

NOT RESOLVED

 

241-243 this is a qualitative (maybe right) consideration.

NO INFORMATION ADDED

 

243-245 It doesn’t appear as a results.

NO INFORMATION ADDED

 

Fig 5. It isn’t clear to me where data for histogram are obtained. Original data for SSF calculation are not showed into the article. If SSF came from other publications, it should be indicated.

NO INFORMATION ADDED

 

272-274 This affirmation is in general acceptable but not demonstrated in ths work. Moreover I can’t understand with a such high infiltration rate. By observing Fig 4, omitting first 30’ of infiltration rate stabilization for all tests, infiltration rate move from 300 mm/h to 50 mm/h so that abandonment effect on runoff formation can be appreciated only for rainfall intensity over 50 mm/h and duration over 30’; otherwise runoff shouldn’t form.

Moreover 1 year of abandonment, only an intensity rainfall over 150 mm/h and 30’ of duration (75mm of cumulative rain), can generate runoff and dry stone wall destabilization. For under 30’ rainfall intensity rainfall should be much higher.

In my opinion a more complex hydrology governs runoff and water balance into that terraced system and shallow infiltration rate gauging can represent only the “bottleneck” of water infiltration.

I think that we cannot make this kind of comparison between the appearance of runoff under natural rains and the infiltration capacity measured by the double rings to draw conclusions about the values of the rainfall intensities that could produce the runoff.

The double rings measure the infiltration capacity under a hydraulic load of 10 cm (ponding) while the runoff is produced under natural rain under the effect of modification of the soil surface (crusting and compacting of the soil surface).

NO INFORMATION ADDED AND NOT RESOLVED

 

 

284-286 Maybe fertility losing influence vegetation non affirmation but there is not an evidence in this work. No fertility measurement have been provided before and after abandonment or, in alternative, in still cultivated or abandoned terraces.

Indeed, we did not measure soil fertility in the strict sense (N, P, K, etc.). But the fact that the soil is eroded implies a loss of fertility. We don't need evidence for that. Besides, the writing is in parenthesis.

NO INFORMATION ADDED

 

Submission Date

15 October 2020

Date of this review

04 Dec 2020 10:56:48

 

Round 3

Reviewer 3 Report

The work appears to be interesting inhabited areas investigation, where the terraces represent a since long time the cultivation system in that area, probably the only one possible to preserve fertility, comfortable (!) territory shaping and erosion control.

No doubt are moved forward on fact that maintenance attention to the of these traditional artifacts can give chance to the territory habitability, the stability, and security of downstream territories.

Nevertheless, some doubts remain on the methodology adopted, on the presentation of the results and above all on some qualitative rather than quantitative considerations showed in this work; this does not mean that the results are inaccurate but that greater rigor is required in the application of the scientific method. And in particular (line order follows not revised versione):

60-71 Taking into account article purpose, rainfall characteristics doesn’t seems very well described. Is not enough in my opinion the annual amount of rain and the coefficient of variation over a (long) gauging period. Average rainfall =211mm and 54% of coefficient of variation doesn’t describe potential intense rainfall, usually responsible for runoff and erosion. Would be appreciable some mathematical descriptor as IDF curves to be compared with infiltration rate.

It also must be pointed out that for dry stone wall collapse, no indication about causes is provided. If abandonment of land cultivation induce wall collapse, due to infiltration rate changing, runoff should destabilize stones from the top of the wall.

 

NOT RESOLVED

 

Figure 1: maps aren’t accurate to understand elevation, convergence, slope aspect and so on. Could you provide a morphological or technical map to overlap study area?

RESOLVED

 

 

IN Table 1 it is curious to couple conservation state with number of years of abandonment. doesn’t exist other cases with crossed classes? As, for example, Good state of conservation in 5 years of abandonment and vice versa? The classes are probably related but not necessarily cause-and-effect and bijective as affirmed.

NOT RESOLVED

 

Lines 86-90 as above mentioned, is the 50k topographic background with overlapped GPS surveys available?

RESOLVED

 

Lines 97-100 Pen Penetrometer provides indication on soil resistance but only on 5-10 mm of depth. When main reason of high infiltration rate and reduced compaction derive from continuous tillage for cultivation, it is risky in my opinion to extend top soil penetration measures and analysis to the entire thickness of the investigate soil.

NONE INFORMATION ADDED

 

124 “Ravines” it appear 2 times..

RESOLVED

 

141–151 I do not think that emigration paragraph should be placed in Results and Discussion session. Maybe better in Introduction to highlight one of the main study reasons.

RESOLVED

 

Fig 4. Photos show some dry stone wall built, apparently, in a not very stable way. Stones seems to be distributed without a maximum stability criterion (big stones should be placed with larger side in horizontally position, arranged with regular intervals and "saturated" by small stones also placed horizontally as possible with local material). A non correct stone distribution and arrangement create wall areas with reduced resistance, “seeds” for local instability starting.

side comment that could have been ignored

182 two poinst in line

RESOLVED

 

181-188 I don’t understand where the showed data are coming from (open surface, plant cover percentage and compaction). What does it means that the soil is “frozen”?

RESOLVED

 

Tab 4: Bare Surface, cited into the text, has not presence into the table. Wasn’t an interesting parameter?

NOT RESOLVED

 

TAB 5: Infiltration rate are very high for cultivated areas. This data aren’t surprising taking by considering infiltration rate of a recently plowed terrace. I’m afraid that such high values are due to surface plowing that is a partial measure of a in-homogeneous soil system and scale effect of a local test is great. It is probably that during test, the water descend vertically till the not worked soil (base below plowing depth) and there goes laterally in base of lateral slope. If this scheme was confirmed, data provided does not refer to average slope hydrologic behavior (two layers: the shallow one very permeable, the lo lower much less permeable) and runoff consideration should be evaluated in this light.

Moreover standard deviation is quite high (in a normal distribution I expect than 95% of observations in 90,2-513,4 mm/h range) that increase doubts on measuring gauging (not on right field measurement but homogeneous distribution on slopes and an enough samples magnitude). It is maybe possible one or few outlier that casually deviate from normal assumption? In fact standard deviation reduction move from 3.66 and 7.1 times in 1-5 years of non plowing is very high and suggest (in addition to a modification of the shallow hydraulic characteristics) an homogenization of analyzed area.

NOT RESOLVED

 

Fig 4 I would appreciate boxplot for infiltration rate measurement

NOT IMPORTANT BUT NOT RESOLVED

 

231-231 This is realistic but not evidenced in this study. In order to prove that, it should taken into account above mentioned consideration on shallow and deep soil on hydrologic behavior. Effects of high infiltration rate for terraced system on runoff formation, moreover, is relevant in direct relation with relative extension on watershed.

NOT RESOLVED

 

241-243 this is a qualitative (maybe right) consideration.

NO INFORMATION ADDED

 

243-245 It doesn’t appear as a results.

NO INFORMATION ADDED

 

Fig 5. It isn’t clear to me where data for histogram are obtained. Original data for SSF calculation are not showed into the article. If SSF came from other publications, it should be indicated.

NO INFORMATION ADDED

 

272-274 This affirmation is in general acceptable but not demonstrated in ths work. Moreover I can’t understand with a such high infiltration rate. By observing Fig 4, omitting first 30’ of infiltration rate stabilization for all tests, infiltration rate move from 300 mm/h to 50 mm/h so that abandonment effect on runoff formation can be appreciated only for rainfall intensity over 50 mm/h and duration over 30’; otherwise runoff shouldn’t form.

Moreover 1 year of abandonment, only an intensity rainfall over 150 mm/h and 30’ of duration (75mm of cumulative rain), can generate runoff and dry stone wall destabilization. For under 30’ rainfall intensity rainfall should be much higher.

In my opinion a more complex hydrology governs runoff and water balance into that terraced system and shallow infiltration rate gauging can represent only the “bottleneck” of water infiltration.

NO INFORMATION ADDED AND NOT RESOLVED

 

 

284-286 Maybe fertility losing influence vegetation non affirmation but there is not an evidence in this work. No fertility measurement have been provided before and after abandonment or, in alternative, in still cultivated or abandoned terraces.

NO INFORMATION ADDED

Author Response

Responses to reader 3:

All my responses are in this file. No information is added to the paper.

 

Comments and Suggestions for Authors

The work appears to be interesting inhabited areas investigation, where the terraces represent a since long time the cultivation system in that area, probably the only one possible to preserve fertility, comfortable (!) territory shaping and erosion control.

No doubt are moved forward on fact that maintenance attention to the of these traditional artifacts can give chance to the territory habitability, the stability, and security of downstream territories.

Nevertheless, some doubts remain on the methodology adopted, on the presentation of the results and above all on some qualitative rather than quantitative considerations showed in this work; this does not mean that the results are inaccurate but that greater rigor is required in the application of the scientific method. And in particular (line order follows not revised version):

60-71 Taking into account article purpose, rainfall characteristics doesn’t seems very well described. Is not enough in my opinion the annual amount of rain and the coefficient of variation over a (long) gauging period. Average rainfall =211mm and 54% of coefficient of variation doesn’t describe potential intense rainfall, usually responsible for runoff and erosion. Would be appreciable some mathematical descriptor as IDF curves to be compared with infiltration rate.

To carry out a frequency analysis of precipitation intensities, we need measurements at short time steps of the intensity (mn, hour). However, these data do not exist. This is why we used the annual mean value and the coefficient of variation to characterize the precipitation.

It also must be pointed out that for dry stone wall collapse, no indication about causes is provided. If abandonment of land cultivation induces wall collapse, due to infiltration rate changing, runoff should destabilize stones from the top of the wall.

The damage to the walls is due to surface runoff, but also by grazing animals in the area and especially also by the lack of maintenance. The stones of the body of the wall are destabilized over time and can slip and weaken the wall. It's not just surface runoff. As can be seen from the photos (fig 3).

NOT RESOLVED

 

Figure 1: maps aren’t accurate to understand elevation, convergence, slope aspect and so on. Could you provide a morphological or technical map to overlap study area?

RESOLVED

 

 

IN Table 1 it is curious to couple conservation state with number of years of abandonment. doesn’t exist other cases with crossed classes? As, for example, Good state of conservation in 5 years of abandonment and vice versa? The classes are probably related but not necessarily cause-and-effect and bijective as affirmed.

Prospecting the land with the peasants showed us that the land was abandoned in waves, during the same period and globally on the same slopes. The abandonment age was roughly estimated and was used as an indication for the cessation of agricultural activities on the terraced plots. Different, but approximately similar places can be found in the same age group.

NOT RESOLVED

 

Lines 86-90 as above mentioned, is the 50k topographic background with overlapped GPS surveys available?

RESOLVED

 

Lines 97-100 Pen Penetrometer provides indication on soil resistance but only on 5-10 mm of depth. When main reason of high infiltration rate and reduced compaction derive from continuous tillage for cultivation, it is risky in my opinion to extend top soil penetration measures and analysis to the entire thickness of the investigate soil.

These measurements (PEN, CV, bulk density) are to characterize the soil surface. They are not generalized over the entire depth of the soil. Several studies have shown that the compaction of the soil surface is generally done on the first cm (2, 5 cm, 7 cm). It is this part of the soil that determines its infiltration capacity.

NONE INFORMATION ADDED

 

124 “Ravines” it appear 2 times..

RESOLVED

 

141–151 I do not think that emigration paragraph should be placed in Results and Discussion session. Maybe better in Introduction to highlight one of the main study reasons.

RESOLVED

 

Fig 4. Photos show some dry stone wall built, apparently, in a not very stable way. Stones seems to be distributed without a maximum stability criterion (big stones should be placed with larger side in horizontally position, arranged with regular intervals and "saturated" by small stones also placed horizontally as possible with local material). A non correct stone distribution and arrangement create wall areas with reduced resistance, “seeds” for local instability starting.

side comment that could have been ignored

182 two poinst in line

RESOLVED

 

181-188 I don’t understand where the showed data are coming from (open surface, plant cover percentage and compaction). What does it means that the soil is “frozen”?

RESOLVED

 

Tab 4: Bare Surface, cited into the text, has not presence into the table. Wasn’t an interesting parameter?

As indicated before, bare ground + covered ground = 100%. Whether you use one or the other, it's the same thing!

NOT RESOLVED

 

TAB 5: Infiltration rate are very high for cultivated areas. This data aren’t surprising taking by considering infiltration rate of a recently plowed terrace. I’m afraid that such high values are due to surface plowing that is a partial measure of a in-homogeneous soil system and scale effect of a local test is great. It is probably that during test, the water descend vertically till the not worked soil (base below plowing depth) and there goes laterally in base of lateral slope. If this scheme was confirmed, data provided does not refer to average slope hydrologic behavior (two layers: the shallow one very permeable, the lo lower much less permeable) and runoff consideration should be evaluated in this light.

Moreover standard deviation is quite high (in a normal distribution I expect than 95% of observations in 90,2-513,4 mm/h range) that increase doubts on measuring gauging (not on right field measurement but homogeneous distribution on slopes and an enough samples magnitude). It is maybe possible one or few outlier that casually deviate from normal assumption? In fact standard deviation reduction move from 3.66 and 7.1 times in 1-5 years of non plowing is very high and suggest (in addition to a modification of the shallow hydraulic characteristics) an homogenization of analyzed area.

As indicated in the first correction, this system described by the reader exists in agricultural land cultivated on the surface by a disc plow. However, this is not the case here. The tillage is done by the hoe and the "soil" is a deposit of fine particles and large stones. No formation of an impermeable layer is observed at depth. There is no plow base.

NOT RESOLVED

 

Fig 4 I would appreciate boxplot for infiltration rate measurement

NOT IMPORTANT BUT NOT RESOLVED

 

231-231 This is realistic but not evidenced in this study. In order to prove that, it should taken into account above mentioned consideration on shallow and deep soil on hydrologic behavior. Effects of high infiltration rate for terraced system on runoff formation, moreover, is relevant in direct relation with relative extension on watershed.

With the reduction of infiltration capacity, the abandoned terraces over large areas are believed to be responsible for the increasingly flashing, intense and frequent floods in the Anti-Atlas and the surrounding plains.

This is an observation to be verified by future mapping and spatialization research of infiltration processes at the watershed scale.

NOT RESOLVED

 

241-243 this is a qualitative (maybe right) consideration.

NO INFORMATION ADDED

 

243-245 It doesn’t appear as a results.

NO INFORMATION ADDED

 

Fig 5. It isn’t clear to me where data for histogram are obtained. Original data for SSF calculation are not showed into the article. If SSF came from other publications, it should be indicated.

NO INFORMATION ADDED

 

272-274 This affirmation is in general acceptable but not demonstrated in ths work. Moreover I can’t understand with a such high infiltration rate. By observing Fig 4, omitting first 30’ of infiltration rate stabilization for all tests, infiltration rate move from 300 mm/h to 50 mm/h so that abandonment effect on runoff formation can be appreciated only for rainfall intensity over 50 mm/h and duration over 30’; otherwise runoff shouldn’t form.

Moreover 1 year of abandonment, only an intensity rainfall over 150 mm/h and 30’ of duration (75mm of cumulative rain), can generate runoff and dry stone wall destabilization. For under 30’ rainfall intensity rainfall should be much higher.

In my opinion a more complex hydrology governs runoff and water balance into that terraced system and shallow infiltration rate gauging can represent only the “bottleneck” of water infiltration.

I think that we cannot make this kind of comparison between the appearance of runoff under natural rains and the infiltration capacity measured by the double rings to draw conclusions about the values of the rainfall intensities that could produce the runoff.

The double rings measure the infiltration capacity under a hydraulic load of 10 cm (ponding) while the runoff is produced under natural rain under the effect of modification of the soil surface (crusting and compacting of the soil surface).

NO INFORMATION ADDED AND NOT RESOLVED

 

 

284-286 Maybe fertility losing influence vegetation non affirmation but there is not an evidence in this work. No fertility measurement have been provided before and after abandonment or, in alternative, in still cultivated or abandoned terraces.

Indeed, we did not measure soil fertility in the strict sense (N, P, K, etc.). But the fact that the soil is eroded implies a loss of fertility. We don't need evidence for that. Besides, the writing is in parenthesis.

NO INFORMATION ADDED

 

Submission Date

15 October 2020

Date of this review

04 Dec 2020 10:56:48

 

Author Response File: Author Response.docx