Regional Drivers behind the Burning of Remanent Forests in Michoacán Avocado Belt, Central Mexico
Round 1
Reviewer 1 Report
Please check the attached document.
Comments for author File: 
Comments.pdf
Author Response
Reply: All the comments of the reviewer were addressed, and changes in the manuscript text were performed. The results and discussion sections now include and emphasize the atmospheric and meteorological variables that drive forest fire occurrence within the studied region. We also rewrote specific sections of the manuscript in order to show the relationship with global atmospheric processes in the study region, such as El Niño and La Niña phenomena.
Please see the attached file
Author Response File: 
Author Response.docx
Reviewer 2 Report
In “Regional drivers behind the burning of remnant forests in Avocadoland, Central Mexico,” the authors compare the incidence of wildland fire (intentionally set or otherwise) over an 18-year period across central Michoacán state (Mexico) with the location of avocado orchards. They also estimate fire return interval during this period, as well as proximate drivers of the patterns observed.
Major comments
While the data presented are both interesting and worthy of publication in the right context, I do not find that the manuscript as currently written provides the structure and connection between data and objectives in a compelling way. Based on the title, I expected to see visually (i.e., in a spatially-explicit way) and quantitatively (i) how fires set in the avocado-growing region were reducing or eliminating natural forest remnants, (ii) a breakdown of the drivers of this forest conversion, and (iii) how fire size/number together with conversion were changing over time. Figure 4, for example, shows the distribution of fires and avocado orchards and agricultural lands in the study area, which was presumably once all forest. I was hoping, however, to see, on this or a similar map, the distribution of current forest remnants (ideally by forest type), as well as the distribution of remnants that were burned by fires during the time period evaluated – all of which I think readers would find very interesting.
Given the fire overlap and spatial analyses conducted, I also expected to see how the current fire return interval (FRI) varied across the study region and how that compares to estimates of historical or natural FRI. Presumably, fires are set to clear forests and understory vegetation in advance of establishing avocado orchards (and perhaps to maintain orchards, although this isn’t stated). As such, I would assume that FRIs now are much shorter than under natural conditions, but the only reference to natural or historical FRI (L340) is overly general (“between a couple of years to decades”) and not useful for comparison. Instead, the results presented included number of fires each year by municipality, a coarse-scale aggregate estimate for percent forest fragment type burned (L256), total area burned annually, the relationship between average number of fires and area burned per month and climate variables, intent, and distance-related proximate drivers. I am not implying that these results would not be of interest to readers, but a stronger and more structured argument for why these and not other results are of interest to the reader needs to be established in the introduction and buttressed in the discussion.
Lastly, other than to dispel the perception stated in lines 392-93 that ENSO events have a positive effect on fire number and size, I do not understand the emphasis on atmospheric variables, including the timing of el Niño/la Niña periods. If the authors are interested in showing how intentionally set fires are reduced or augmented under certain climatic conditions, that should be stated as an objective. But in general, it seems like if one is talking about the two-thirds of fires that are intentionally set, people inclined to set fires, especially for economic benefit, are going to set fires. Otherwise, I would think it more relevant to present data on and assess the degree to which fires are limited by atmospheric factors, but also ignition type and fuels.
Minor Comments
While the authors clearly have a strong command of English, I recommend a careful editing by a native English speaker; I made some changes but there were more than I had the time to address.
I am not insisting, but would suggest the authors consider replacing “Avocadoland” with “Michoacan Avocado Belt” in the title.
L126 Replace “of” with “in the”
Table 1, footnote 2: is this the percentage canopy cover? If so, say that, otherwise it’s not clear to me what % surface covered means.
Section 2.2 – a lot of unnecessary information here; this section could be made more concise.
Fig. 2 – This diagram was a confusing and might benefit from more explanation in the caption. Also, i* is not defined, nor is it clear how the bracket connects the right and left sides.
Figs. 1, 3 & 4 – A distance scale (e.g., 50 km) on the local map in lieu of or in addition to the Lat/Lon hashmarks would be helpful.
It’s not clear how an inactive weather station would be useful.
Fig. 4 – Would like to see forest remnants on this map.
Distance to social factors – this needs more development in intro
L228 “varied year-to-year” rather than “presented…”
Table 2 – I think it would be easier for the reader to understand if the data were presented as a series of colored lines in a line graph. The actual table could go into an appendix.
L252 use “area” instead of “extension” where the Spanish “extensión” would fit (e.g., also L272). The use of “extent” later in the sentence is fine as is. The entire paragraph could benefit from careful editing of the English.
L255-256 Do these numbers come from this analysis? If so, this seems like essential information to present spatially and with finer resolution.
L297 The construction of this sentence is confusing. Do the authors mean that fires >500 ha had an average RH of 52.8% compared to fires of <50 ha that had an average RH of 64.9%? If so, it should be reported this way, and not as written.
Section 3.3 and discussion - I question the use of “arson.” I suppose any unsanctioned intentional fire can be called arson, but if these fires are set by landowners growing avocados, then it seems more appropriate to call them something like illegal agricultural burning.
L328 Authors may want to reconsider using the term mega-fire for fires >200 ha, as this term is already common in the fire literature, where in the U.S., the Interagency Fire Center defines a megafire as >40,500 ha (100,000) acres.
L333 “model” is redundant, as the M of GLM stands for model.
L337 (244) “at or below 2000 m” rather than “inside.” It is unclear why elevation/altitude is relevant. The authors should establish this in the introduction or methods. If the orchards occur from 1100-2900 masl, what is the significance of 1945 m?
L339 Hopefully there are more precise estimates for the natural fire return intervals (FRI) of the forests in the study area than “2 years to decades.” I believe the utility of the contemporary FRI analysis depends on being able to compare it to the historical/natural FRI.
Section 4.1 seems like it would be better characterized as economic or socioeconomic drivers, as the implication is that burning for agricultural benefit underlies the trend. Yet the treatment of this topic is unsatisfying because it deals largely in generalizations. L352-353 talks about a well-established cultural management of fire in Neotropical forests, but the authors do not make any explicit connection between the inhabitants of this land and the fires – citation 38 deals with FRI in Mexico State and citation 40 refers to the PNW of the US and Canada. Nor is it clear that the cultural aspects referred to are Indigenous in origin. I find it unlikely that Indigenous groups would be burning the forest every couple of years, unless it was for management of specific grasslands or milpas. Do we suspect that these forests were experiencing such frequent fires before the avocado orchards established?
L352 – Are these forests actually neotropical? The authors mention pine and oak species earlier and I suspect most of these forests, especially those at 2000 masl or higher are pine-oak forests that behave more like temperate forests (Nearctic in origin). More needs to be said about the natural role of fire in these forests to understand how much current conditions represent a deviation from this. If some of these are broadleaf forests, I would think that fire would play very little natural role in their ecology.
L393 – Post hoc should be the start of a new sentence.
Section 4.2 Other than being conducive to lighting fires for land clearing or not, is there any indication of whether climatic conditions during certain times of the year are causing fires to burn out of control or beyond what the people setting the fires intend to burn?
L407 “high vulnerability in the appearance…” I don’t understand this.
L431 There are only 58 references.
Author Response
The comments of the reviewer were sound and valuable. Considering comments and suggestions, the manuscript included a new figure (figure 5). It depicts fires occurring in forests near the agricultural frontier, roads, and existing avocado plantations.
The introduction was rewritten considering expert knowledge regarding the relationship between induced fires and avocado plantation establishment. “One of the tools used for land-use change is the spread of fires in forested areas, [22-25]. Fires weaken the forest trees, favoring the presence of pests such as bark beetles (Dendroctonus spp.). According to Mexican legislation, outbreaks of these pests must be treated by felling and fumigating the infested trees, timber can later be sold later. This situation offers the possibility of having land clear of vegetation and enough monetary resources to start an avocado plantation and provide inputs for it during the first years until fruit production begins, normally between the 4th and 5th year [24]. Even though national and local legislation strongly limit land-use changes from native forest to other uses, like the Sustainable Forest Development Law [26], the lack of effective law enforcement mechanisms has allowed the occurrence of massive illegal land-use changes following fires mainly to agricultural and urban uses [8, 13, 22]. Nevertheless, the role of fire and local changes to fire regimes associated with avocado orchards has not been evaluated.”.
Please see attached file.
Author Response File: Author Response.docx
Reviewer 3 Report
Olivares-Marinez et al. presents a compelling, well-designed, well-written, and highly relevant article on the regional drivers behind fire occurrence/reoccurrence in remnant forests in Avocadoland, Central MX. I found it very well written and important ecologically and from a management standpoint. I have only minor suggestions for improvements as in my comments on the pdf, including: (1) if you have room, include more of the key findings in the abstract especially in relation to how the situation could be improved; (2) discuss further the data limitations around using fire records to plot the fire centroids and what that could have done to bias distance measures (I comment on that in a number of places); (3) the discussion of "fire regime" needs a bit more work - did you actually describe the regime or something else (for instance, there's no discussion of fire intensity, severity, patch sizes etc); (4) you could expand on the conclusions and conservation findings much more - including - would an early detection system help identify the conditions of fire starts, would seasonal road closure limit starts, would forest protection of the fragments limit fires spreading into orchards? There's more that you could pull from your findings to reduce forest losses and fire reoccurrence. I do have problems visualizing how a dry forest system - at least during the periods that are dry - has not had much natural fire ignitions - is lightning not a factor? Are the forests therefore all not fire dependent? If so, might there be type conversions happening whereby forests intolerant of fires are replaced by tolerant ones and species intolerant of frequent fires may begin to go locally extinct? Conservation implications? What about climate change and the incidence of fires - might there be even more atmospheric conditions for fire starts to result in even further forest losses? More can be pulled from your findings in relation to conservation - please do that.
Comments for author File: Comments.pdf
Author Response
The abstract was rewritten in order to include the key findings suggested by the reviewer. Data limitations and fire location and centroid description including the implication in the overall analysis were rewritten. The fire regime section was extended and completed in order to encompass a broader perspective.
Please see the attachment
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
First, I'll note that the version of the manuscript I received is a PDF of the version the authors edited in track changes, which is difficult to interpret, so apologies if I ask for edits that have been made.
I appreciate that the authors made many changes to the manuscript based upon my comments. Aside from several smaller changes throughout, the revision of Figure 2 was extremely helpful - the new Figure 2 really helped me understand how the analyses handled fire occurrence data. Nice work.
I appreciate that the primary result is that most fires are anthropogenic and the atmospheric influence only mediates the severity, etc. of fires. That is a valuable insight, worth publishing. Distance to roads and orchards is inversely correlated with fire incidence. I still don't see how those two are separate factors - maybe I don't understand the landscape in question, but they seem like they should be positively correlated (orchards near roads). I think the authors still need to emphasize either a) that orchards are not near roads, or b) dispel concerns about whether they are correlated. Otherwise, it is not clear to me that both are actually important factors in determining fire incidence.
Author Response
Please see the attached file.
Author Response File: Author Response.docx
Reviewer 2 Report
The manuscript has definitely improved. I like the figure at the top of page 10, although the caption is not in the right place in my pdf. A color for forest that stands out better would improve the figure. The manuscript would still benefit from another revision of the English, but it is acceptable. Here are a few editing suggestions:
Pg 10 bottom paragraph: the English in this paragraph needs editing for improved clarity.
“According to the curated database, most of the area burned occurs over three main forest types: pine (25%); pine-oak…”
I’m not sure what the next sentence is saying. The average fraction of trees burned in each fire was 6.4%?
Fig. 7. Average number of fires per month in the MAB from 2000 to 2017…
L 1604 remove “it”: “humidity was inversely proportional.”
L 1886 Typo: burnings
L 1901 add “that”: “fire that occurred”
L 1994 suggest: “advisable to use this method with polygons…”
L 2025 suggest: In the MAB, distance to roads is (or ‘has been’)
L 2208 sentence is too dramatic. Delete: Irremediably, soaring
L 2236 “homologation” is an extremely uncommon word in English. Suggest using simpler words to convey the concept.
Author Response
Please see the attached file.
Author Response File: Author Response.docx
