Spatiotemporal Land Use and Land Cover Changes and Their Impact on Landscape Patterns in the Colombian Coffee Cultural Landscape (2014–2034)
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors can find my comments in the attached document.
Comments for author File: 
Comments.pdf
Author Response
|
We sincerely thank you for the time and dedication invested in reviewing our manuscript and for the thoughtful and constructive feedback that significantly contributed to improving its scientific rigor and clarity. Your suggestions led us to a comprehensive revision of both the structure and content of the manuscript. In particular, we have incorporated clarifications in key definitions, eliminated redundant or unnecessary sections, corrected inconsistencies in tables and figures, and revised methodological descriptions for conciseness and precision. These and other changes—clearly highlighted in the revised version—reflect our commitment to addressing each of your comments with care and respect. We are grateful for your insightful observations, which have undoubtedly strengthened the contribution of this work.
|
||
|
2. Questions for General Evaluation |
Reviewer’s Evaluation |
Response and Revisions |
|
Does the introduction provide sufficient background and include all relevant references? |
Yes/Can be improved/Must be improved/Not applicable |
[Please give your response if necessary. Or you can also give your corresponding response in the point-by-point response letter. The same as below] |
|
Are all the cited references relevant to the research? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Is the research design appropriate? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the methods adequately described? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the results clearly presented? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
3. Point-by-point response to Comments and Suggestions for Authors
|
||
|
ABSTRACT: Comments 1: The innovation of this study lies in the integration of multi-sensor remote sensing, hybrid predictive models, and landscape metrics within the CCLC, providing a quantitative framework to assess cultural landscape transformations under anthropogenic pressures.
|
||
|
Response 1 : The phrase "The innovation of this study lies in" has been removed, taking into account your suggestions. Although the methods proposed have indeed been extensively documented, these methods have not yet been integrated into a unified methodological framework for heritage agricultural landscapes. Thank you for your contribution; your comment is acknowledged and agreed upon. Accordingly, the text has been modified as follows:
[Updated text in the manuscript: “The integration of multisensor remote sensing, hybrid predictive models, and landscape metrics within the CCLC provides a quantitative methodological framework to evaluate the transformation of cultural landscapes under anthropogenic pressures.” – page number 1, paragraph 1, and line 23-26.]
|
||
|
INTRODUCTION: Comments 2: [The first sentence needs a citation “Cultural landscapes... are experiencing accelerated transformations”]
|
||
|
Response 2: In response to the observations made, the following phrase is cited: "Cultural landscapes... are experiencing accelerated transformations," from the article titled "Progress toward the sustainable development of world cultural heritage sites facing land-cover changes." This article explicitly states the following: "This study confirms that land-cover changes are among serious threats for heritage conservation, with heritage in some countries wherein the need to address this threat is most crucial, and the proposed spatiotemporal monitoring approach is recommended." [Updated text in the manuscript: “Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by UNESCO [1], are experiencing accelerated transformations due to changes in land use and land cover (LULC) [2].” – page number 2 , paragraph 1, and line 39-42.]
|
||
|
Comments 3: [“The sentence “UNESCO has identified 14 threats that caused...” is out of context in relation to what is said before or after. It should be explained in more detail, specifying what these 14 threats are. Alternatively, it could be removed."] |
||
|
Response 3: We agree with this comment. Therefore, we have decided to remove the sentence "UNESCO has identified 14 threats that cause irreversible biophysical and aesthetic damage in conservation areas [5]"because it was out of context, and providing a more detailed explanation would unnecessarily lengthen the manuscript." Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by UNESCO [1], are experiencing accelerated transformations due to changes in land use and land cover (LULC). These changes reflect the direct interaction between human activities and the natural environment, driven by socioeconomic and environmental factors [26]. Over time, human intervention has shaped productive landscapes with high economic, cultural, and ecological value; however, increasing anthropogenic pressure compromises their integrity and sustainability [27]. UNESCO has identified 14 threats that cause irreversible biophysical and aesthetic damage in conservation areas [28]. Furthermore, it is estimated that nearly three quarters of the Earth's surface has been transformed by human activities [29,30], intensifying habitat fragmentation, biodiversity loss, and the degradation of cultural values associated with these landscapes [31]. [Updated text in the manuscript: “Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by UNESCO [1], are experiencing accelerated transformations due to changes in land use and land cover (LULC) [2]. These changes reflect the direct interaction between human activities and the natural environment, driven by socioeconomic and environmental factors [3]. Over time, human intervention has shaped productive landscapes with high economic, cultural, and ecological value; however, increasing anthropogenic pressure compromises their integrity and sustainability [4]. Furthermore, it is estimated that nearly three quarters of the Earth's surface has been transformed by human activities [5,6], intensifying habitat fragmentation, biodiversity loss, and the degradation of cultural values associated with these landscapes [7].” – page number 2 , paragraph 1, and line 40-49.]
|
||
|
Comments 4: [“At the end of this section, it is detailed that spatial analysis and remote sensing techniques are used. This paragraph can be removed as it is not necessary to explain it here."] |
||
|
Response 4: Thank you for noting that the final paragraph of the introduction was unnecessary. In accordance with your recommendation, we have removed that paragraph to maintain clarity and conciseness.
|
||
|
Comments 5: [“A brief introduction to the scientific literature on land use changes in different parts of the world and in other UNESCO sites (if applicable; and if not, it should be mentioned that no studies have been conducted in UNESCO sites) is missing. Additionally, other studies conducted in protected natural areas can be discussed, if there are no studies on UNESCO sites, as well as studies carried out in Colombia."] Response 5: We appreciate the rigor and detail with which you have reviewed our manuscript. In line with your valuable suggestions, we have incorporated additional scientific literature addressing changes in land use in various regions worldwide, as well as in other UNESCO-recognized sites and in Colombia.
[Text to be corrected: “Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by the United Nations Educational, Scientific and Cultural Organization UNESCO [1], are experiencing accelerated transformations due to changes in land use and land cover (LULC) [2]. These changes reflect the direct interaction between human activities and the natural environment, driven by socioeconomic and environmental factors [3]. Over time, human intervention has shaped productive landscapes with high economic, cultural, and ecological value; however, increasing anthropogenic pressure compromises their integrity and sustainability [4]. Furthermore, it is estimated that nearly three quarters of the Earth's surface has been transformed by human activities [5,6], intensifying habitat fragmentation, biodiversity loss, and the degradation of cultural values associated with these landscapes [7]. ” – page number 2 , paragraph 1, and line 40-50.] [Updated text in the manuscript: “Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by the United Nations Educational, Scientific and Cultural Organization UNESCO [1], are experiencing accelerated transformations due to changes in land use and land cover (LULC) [2]. Globally, it is estimated that nearly three quarters of the Earth’s surface have been transformed by human activities [3,4], exacerbating habitat fragmentation, biodiversity loss, and the degradation of cultural values [5]. This phenomenon not only affects natural ecosystems but also areas designated as World Heritage Sites, such as Sagarmatha National Park in Nepal [6], Jeju Island in the Republic of Korea [7] , and the Viñales Valley in Cuba [8], where urbanization, deforestation, and agricultural expansion processes have been documented, compromising their integrity. These changes reflect the direct interaction between human activities and the natural environment, driven by socioeconomic and environmental factors [9]. Over time, human intervention has shaped productive landscapes with high economic, cultural, and ecological value; however, growing anthropogenic pressure jeopardizes their sustainability [10].” – page number 2 , paragraph 1, and line 40-54.]
|
||
|
STUDY AREA Comments 6: [“More information is needed. Six coffee landscapes are mentioned, but it is not clear which ones they are. In this section, it is not specified what land uses exist, what the economic activities of the people living there are, whether there are protected species, or what kind of protection is in place... after all, why is it a UNESCO site?"] |
||
|
Response 6 : Based on the suggestions received, the description of the six zones that comprise the coffee landscape has been clarified. Furthermore, Figure 1 has been updated to clearly identify zones A, B, C, D, E, and F. The main land uses of the CCLC have been specified, highlighting that, in addition to coffee production, commercial and tourism activities are important economic drivers for its inhabitants. In this section, we have decided to clarify why only the core zone is studied, taking into account other suggestions made by the reviewers; for this reason, this information is added. With respect to the question regarding why this site is recognized by UNESCO, the answer is detailed in the Introduction; therefore, it was deemed unnecessary to repeat this information in the study area section.
[Text to be corrected: “The CCLC is located in the central and western foothills of the Andes, in mid-mountain regions ranging from 1000 to 1900 meters above sea level, with temperatures between 17 and 24°C and precipitation of 1200 to 2300 mm. It extends across the departments of Caldas, Quindío, Risaralda, and Valle del Cauca (Figure 1). This region, composed of six distinctive coffee landscapes, is home to 595.884 inhabitants [32]. The territory is divided into two zones: the main area, covering 141.120 hectares 140.046 rural and 1.074 urban, distributed across 47 municipalities, 411 veredas, and 14 urban centers; and the buffer zone, covering 207.000 hectares, of which 204.542 rural and 2.458 urban, distributed across 51 municipalities, 447 veredas, and 17 urban centers. ” – page number 3 , paragraph 6, and line 116-125.] [Updated text in the manuscript: “ 2.1 Study Area The Colombian Coffee Cultural Landscape (CCCL) is located on the central and western slopes of the Andes mountain range, within mid-elevation mountainous regions ranging from 1,000 to 1,900 meters above sea level. The area is characterized by temperatures between 17 and 24 °C and annual precipitation ranging from 1,200 to 2,300 mm. It extends across the departments of Caldas, Quindío, Risaralda, and Valle del Cauca, with a total population of 595,884 inhabitants. The territory is composed of six distinct landscape zones (Figure 1), which are characterized by the integration of cultural traditions, including expressions of Indigenous and Afro-descendant communities, archaeological heritage, historical town centers, and legacies associated with coffee cultivation. The region also displays a diverse pattern of land use and agricultural production, encompassing extensive coffee plantations, agroforestry systems, grazing areas, patches of native vegetation, and urban centers with commercial and tourism-related activities [32].
The CCCL is divided into two zones. The primary zone corresponds to the core area, which spans 141,120 hectares—of which 140,046 hectares are rural and 1,074 hectares urban—distributed across 47 municipalities, 411 rural settlements (veredas), and 14 urban centers. The buffer zone covers 207,000 hectares, comprising 204,542 hectares of rural land and 2,458 hectares of urban land, distributed across 51 municipalities, 447 veredas, and 17 urban centers.
This study focuses on the core area of the CCCL, where the attributes of Outstanding Universal Value (OUV) that justified its designation as a World Heritage Site are located [33]. According to UNESCO guidelines, the boundary of the inscribed property must ensure that all attributes contributing to its OUV are contained within its limits, while the buffer zone plays a complementary role in protection and management but is not considered part of the designated property [34,35]. The decision to focus on the core area is based primarily on conservation considerations: it is within this zone that land use changes have a direct impact on the authenticity and integrity of the cultural landscape. Secondly, from a methodological perspective, the core area offers a more coherent and clearly defined unit of analysis, in which the application of spatial metrics and predictive modeling can be conducted with greater accuracy [31].” – page number 3-4 , paragraph 6, and line 104-136.]
|
||
|
SATELLITE IMAGE PROCESSING AND APPLICATION OF SPECTRAL INDICES Comments 7: [“Satellite image processing and Application of spectral indices These sections can be reduced. They are explained in too much detail, and it is not necessary. There are articles that already use this methodology, so these can be cited for more information. Table 2 is not necessary and can be placed as supplementary material."] |
||
|
Response 7 : We appreciate your timely suggestion and, accordingly, have reduced the text as follows:
[Text to be corrected: Satellite image processing To generate the composite image for the study period, a median filter was employed to reduce outliers and improve temporal representativeness [41],39]. Additionally, Sentinel-1 SAR images acquired in Interferometric Wide (IW) mode with VV and VH polarizations on a descending orbit were integrated [43,44]. A median filter was applied to reduce speckle noise, and the VV/VH ratio was calculated as an additional band. Finally, all images were aligned to the EPSG:4326 coordinate system and resampled to a spatial resolution of 30 meters, ensuring consistency across sensors.” – page number 9 , paragraph 1 y 2, and line 190-203.] [updated text in the manuscript: “ The processing of Landsat 8 Collection 2 Tier 1 Level 2 imagery included atmospheric correction using the LaSRC algorithm [39]. with cloud and shadow masks generated from the QA_PIXEL band [40], all processed in Google Earth Engine (GEE) [41–43]. For temporal integration, a median filter was applied to produce a composite image [44,45]. Additionally, Sentinel-1 SAR imagery (IW mode, VV/VH polarizations) was processed by applying speckle filtering and computing the VV/VH ratio [46,47]. All layers were standardized to the EPSG:4326 coordinate reference system and resampled to a spatial resolution of 30 meters.” – page number 8 , paragraph 1, and line 180-188.] [Text to be corrected: The spectral characterization of various land cover types in coffee-growing regions requires a methodological approach that enables the reliable discrimination of coffee crops, forests, bamboo, grasslands, water bodies, bareland, and building areas. The present study employed spectral indices with demonstrated capability to differentiate these representative covers of the coffee landscape [22,48,49].
The spectral basis of the analysis was founded on the bands of the Landsat 8 OLI sensor: blue (0.45–0.51 µm), green (0.53–0.59 μm), red (0.64–0.67 μm), and near-infrared (0.85–0.88 μm), which were selected for their ability to capture the distinctive spectral signatures of each cover [50,51]. For the discrimination between coffee crops, forests, and bamboo, the Enhanced Vegetation Index (EVI) was employed, as it demonstrated greater sensitivity in differentiating vegetative structures with varying canopy architectures [52]. The EVI proved particularly effective in distinguishing coffee plantations from other types of vegetation due to its lower saturation under high biomass conditions [53,54].
The differentiation between coffee crops and grasslands was optimized by combining the Normalized Difference Vegetation Index (NDVI) and the Soil-Adjusted Vegetation Index (SAVI). The NDVI enabled the identification of variations in green biomass density, while the SAVI was critical for distinguishing areas of young coffee crops and grasslands, where soil influence is significant [55,56].
The identification of buildings and structures associated with coffee production was carried out using the Normalized Difference Built-up Index (NDBI) and the Urban Index (UI), both of which demonstrated effectiveness in differentiating built-up areas from natural surfaces [57,58]. Additionally, the Bare Soil Index (BSI) [59] facilitated the identification of exposed soil areas, including roads and zones undergoing planting.
For the detection and delineation of rivers and water bodies, the Modified Normalized Difference Water Index (MNDWI) was implemented, which demonstrated superior capability in discriminating water surfaces even in the presence of topographic shadows and riparian vegetation [60,61]. The Normalized Difference Moisture Index (NDMI) complements this analysis by facilitating the identification of areas with varying levels of moisture in the soil and vegetation [62].
These indices were calculated from the selected Landsat 8 bands and were added as additional bands to the annual composite images, providing supplementary information that optimized the classification algorithm's ability to accurately differentiate between LULC classes. The ten calculated indices were incorporated into the original image bands as training samples, thereby enhancing the classification accuracy. Table 2 presents a summary of the spectral indices derived from the annual Landsat 8 composites, designed to optimize the discrimination between the various LULC classes. ” – page number 9 and 10 , paragraph 2 -7, and line 200-240.] [Updated text in the manuscript: “ Spectral characterization of land covers in the CCLC was based on validated indices for discriminating coffee crops, forests, bamboo, grasslands, water bodies, bare soils, and built‐up areas [22,48,49]. Using the bands from the Landsat 8 OLI sensor (blue, green, red, and near‐infrared) [50,51], ten indices were computed: the Enhanced Vegetation Index (EVI) to differentiate canopy structures [52–55]; the Normalized Difference Vegetation Index (NDVI), the Soil-Adjusted Vegetation Index (SAVI), and the Modified Soil-Adjusted Vegetation Index (MSAVI) to distinguish young coffee crops from grasslands [56–61]; the Normalized Difference Built-up Index (NDBI) and the Urban Index (UI) for built‐up areas [62–66]; the Bare Soil Index (BSI) for exposed soils [67,68]; and the Modified Normalized Difference Water Index (MNDWI), the Normalized Difference Moisture Index (NDMI), and the Land Surface Water Index (LSWI) for water bodies and soil moisture [69–74]. These indices were integrated as additional bands into the annual composites, thereby optimizing the discrimination among LULC classes. The full formulas and references are provided in (Table S1). ” – page number 8 , paragraph 2, and line 190-203.]
|
||
|
DEFINITION OF CLASSES Comments 8: [“I believe that the sentence from “this algorithm consolidates prediction through majority... expedite its learning algorithms” can be removed.."] Response 8: We agree with this suggestion, and it has been removed. Supervised classification is a method that employs labeled training samples to guide the categorization of pixels into different land use and land cover (LULC) classes [76]. For this study, a representative set of samples covering the relevant LULC categories in the analysis area was selected. These samples were used to train a classification model based on Random Forest (RF), assigning each pixel to one of the defined classes [77,78]
RF is a multivariate and non-parametric machine learning algorithm that improves classification accuracy by combining multiple decision trees [79]. This algorithm consolidates predictions through majority voting to assign labels to the input data. Its computational efficiency and the absence of tree pruning requirements expedite its implementation, making it faster and more resource-efficient than other machine learning algorithms [80]. Therefore, RF was employed in this study for the multitemporal classification of LULC into seven specific classes.” – page number 9, paragraph 3 and 4, and line 215-228.] [Updated text in the manuscript: “Supervised classification is a method that employs labeled training samples to guide the categorization of pixels into different land use and land cover (LULC) classes [76]. For this study, a representative set of samples covering the relevant LULC categories in the analysis area was selected. These samples were used to train a classification model based on Random Forest (RF), assigning each pixel to one of the defined classes [77,78]. RF is a multivariate and non-parametric machine learning algorithm that improves classification accuracy by combining multiple decision trees [79]. In this study, RF was employed for multitemporal classification of LULC into seven specific classes.” – page number 8, paragraph 3, and line 204-212.]
|
||
|
|
||
|
DEFINITION OF CLASSES Response 9: Following the suggestions, the text has been modified as follows:
[Updated text in the manuscript: “Seven land use and cover (LULC) classes were identified: Buildings, Coffee Crops, Bareland, Forest, Grasslands, Bamboo, and Water Bodies. Supervised classification algorithms employed labeled samples to recognize spectral patterns associated with each class [81]. The trained model was applied to both historical and current images, enabling the detection of changes such as urban expansion or the conversion of forests to agricultural land [82]. Training and validation samples were generated through the visual interpretation of Landsat 8 and Sentinel-1 composite images available on GEE. A detailed description of each class, including its uses and characteristics, is presented in Table 2.” – page number 8, paragraph, and line 213-221.] |
||
|
LULC VALIDATION PROCESS previously can be cited."] Response 10: We agree with this suggestion, and it has been removed.
.” – page number 10, paragraph 1 and 2, and line 234-246.]
[Updated text in the manuscript: “2.3.1.5 LULC Validation Process The validation method employed in this study involved the random partitioning of data into training and validation sets. 80% of the samples were allocated for training the Random Forest classifier, while the remaining 20% was proportionally distributed among categories for validation [80,83]. This data-splitting approach is widely recommended to ensure a robust evaluation of the model [84]. Confusion matrices were generated for each period analyzed, enabling the calculation of statistical metrics, including overall accuracy (OA), user accuracy (UA), producer accuracy (PA), and the Kappa coefficient [85,86]. These metrics are essential for evaluating the performance of LULC classification models, providing a quantitative measure of their accuracy [87]. The described procedure was implemented in a script on Google Earth Engine (GEE), the details of which are presented in Appendix A.” – page number 9, paragraph 1 and 2, and line 223-236.]
|
|
SPATIAL VARIABLES
Response 11: We deeply appreciate the comments and suggestions received. Figure 3 has been expanded. In response to the observation regarding the necessity of including at least one citation when stating that “these variables are commonly employed,” the references [92, 93, and 94] have been incorporated. In addition, three additional variables have been added to the prediction model, as suggested by another reviewer, which has entailed an overall revision of the manuscript to ensure coherence and consistency in the methodological description.
[Updated text in the manuscript: “The spatial variables used for prediction include natural factors: elevation, slope [87], precipitation, temperature, and NDVI vegetation dynamics [88,89]; neighborhood factors: distance to main roads and proximity to bodies of water[90]; and socioeconomic aspects, such as population density [91]. The raster maps of each variable (Figure 3) were standardized to a resolution of 30 m, with NoData values set to 0, and a unified spatial extent and coordinate system (ESRI:31918 – SIRGAS_UTM_Zone_18N) was applied, thereby ensuring consistency across all spatial layers used. These variables are employed in land use/land cover (LULC) change analysis[92–94], as they provide reproducible information regarding the physical and anthropogenic influences on its evolution.” – page number 9, paragraph 3, and line 237-246.]
|
|
PREDICTIONS AND VALIDATION OF MODELS Response 12: We have removed the suggested phrases.
LULC simulation was performed using the Cellular Automata–Artificial Neural Networks (CA-ANN) algorithm, implemented in the MOLUSCE module of QGIS [88]. This approach combines Cellular Automata (CA), which model spatial dynamics, with Artificial Neural Networks (ANN), which capture nonlinear relationships between explanatory variables and land use changes. This combination enables the projection of future scenarios with high accuracy by integrating spatial and temporal patterns.
The process was carried out in two stages. First, the model was trained and validated by predicting the LULC for 2024 based on historical data from 2014 and 2019. For this purpose, explanatory variables and transition matrices were integrated, ensuring consistency in the evolution of spatial patterns. Model validation was performed using the kappa index by comparing simulated images with the observed images from 2024, achieving an overall accuracy of 86.52% and a kappa index of 0.83.
Once the model was validated, it was used to project the LULC for 2034. In this case, the input data consisted of the LULC from 2019 and 2024, along with the same explanatory variables and transition matrices used in the validation stage. This procedure ensures continuity in the simulation of spatial changes and enables the evaluation of land cover evolution based on trends observed in previous periods. The CA-ANN model was configured with the following parameters: 350 iterations, a neighborhood size of 1 × 1 pixels, a learning rate of 0.005, five hidden layers, and a momentum of 0.010. This methodological approach guarantees the validity and reliability of the LULC projections, allowing for the analysis of possible change scenarios in the CCLC.” – page number 12, paragraph 2-6, and line 255-284.]
[Updated text in the manuscript: “The MOLUSCE plugin is widely used in modeling land use changes [88]. It provides robust methods for evaluating the correlation between LULC data and spatial variables. Natural, socioeconomic, and neighborhood factors have been identified as key drivers of land use change [89]. LULC simulation was performed using the Cellular Automata-Artificial Neural Network (CA-ANN) algorithm, which is implemented in the MOLUSCE module of QGIS [88]. This approach combines Cellular Automata (CA), which model the spatial dynamics, with Artificial Neural Networks (ANN) that capture the non-linear relationships between explanatory variables and land use change.
The process was carried out in two stages. First, the model was trained and validated to predict the LULC for 2024 using historical data from 2014 and 2019, by integrating explanatory variables and transition matrices. Validation was performed using the kappa index, achieving an overall accuracy of 86.52% and a kappa index of 0.83. Subsequently, the validated model was used to project the LULC for 2034, incorporating the LULC data of 2019 and 2024, along with the same explanatory variables and transition matrices employed in the validation stage. The CA-ANN model was configured with 350 iterations, a neighborhood size of 1 × 1 pixels, a learning rate of 0.005, five hidden layers, and a momentum of 0.010.” – page number 11, paragraph 2 and 3, and line 250-270.]
|
|||
|
TRANSITION MATRICES AND CHORD DIAGRAMS 'to visualize these transitions, chord diagrams are an effective...', is not needed.”] Response 13: In accordance with your recommendations, both the transition matrices and chord diagrams have been eliminated. Additionally, the final paragraph of this section has been removed.
M??= (?,?=1,2,3…?)
where: ● Mᵢⱼ represents the land area transformed from type i to type j. ● i and j denote the land use types before and after the transition, respectively. ● When, i = j, it represents the unchanged land area. ● n is the total number of land use types (n = 7 in this study).
To visualize these transitions, chord diagrams are an effective representation tool [92]. These diagrams depict flows between categories using chords, where the thickness of each chord reflects the magnitude of the transformations. This visual representation facilitates the identification of dominant patterns in territorial dynamics and allows for an intuitive interpretation of the main trends in land use change.” – page number 12, paragraph 5 and line 292-296.]
[Updated text in the manuscript: “To analyze LULC changes, transition matrices were employed as a fundamental tool [89]. In the matrix, rows and columns correspond to the land use types in two distinct periods (T₁ and T₂). The values within the matrix (Mij, where i, j =1,2,…,n) represent the areas that have experienced changes among the different land use types during both periods. Based on previous studies [90,91]. ” – page number 11, paragraph4, and line 277-282.]
|
|
MOVING WINDOW METHOD The last paragraph “this process generated grid maps...” can be removed.”] Response 14: In line with the suggestions provided, the paragraph beginning with “The selected indices included fragmentation metrics…” has been removed, as has the final paragraph “This process generated grid maps…”.
The window size was determined after evaluating multiple scales, with the objective of minimizing distortion in the results and maximizing the representativeness of spatial patterns. For each window position, ten landscape indices were calculated using FRAGSTATS 4.2 software [102], considering data from four temporal periods (2014, 2019, 2024, and 2034).
The selected indices included fragmentation metrics (patch density [PD], edge density [ED], largest patch index [LPI], and perimeter-area fractal dimension [PAFRAC]), connectivity metrics (cohesion and division), and measures of spatial heterogeneity using the contagion index (CONTAG), aggregation index (AI), Shannon diversity index (SHDI), and Shannon evenness index (SHEI).
This process generated grid maps representing the spatial distribution of the landscape indices, thereby enabling an analysis of the temporal evolution of landscape patterns throughout the study period.” – page number 12 and 13, paragraph 5,6 and 7 and line 280-300.]
[Updated text in the manuscript: “The moving window method was implemented to generate a grid map of landscape indices, allowing for the evaluation of the spatial variability of landscape patterns at different scales [100,101]. A 100 m × 100 m window was used, systematically shifted from the upper-left corner of the study area. At each position, landscape indices were calculated and assigned to the central cell.
The window size was determined after evaluating multiple scales, with the objective of minimizing distortion in the results and maximizing the representativeness of spatial patterns. For each window position, landscape indices were calculated using FRAGSTATS 4.2 software [102], considering data from four temporal periods (2014, 2019, 2024, and 2034).” – page number11, paragraph 5 and 6, and line 283-294.]
|
|
|
|
MOVING WINDOW METHOD Response 15: In accordance with the suggestions, the section has been streamlined, the final paragraph removed. Table 5 presents the grouped classification index along with the intensity values assigned to each category. The LUDI ranges from 100 to 400. A value of ?? =100 represents areas dominated by low-intervention uses (e.g., coffee crops, ??=1), whereas ?? = 400 indicates a complete transformation of the land (e.g., urban areas, ??=4. Intermediate values reflect proportional combinations of land uses, thereby quantifying gradients of anthropogenic impact on the landscape.
(5)
where: Ld represents the comprehensive land use intensity index for the study area. Bi corresponds to the intensity index assigned to the category. Ci indicates the percentage of the surface occupied by the category. i denotes the land use category number. n is the total number of evaluated categories. Table 5 presents the grouping of the land use degree classification index, along with the intensity values assigned to each category. Table 5. Assignment of the Land Use Degree (LUDI) Classification Index
The index ranges between 100 and 400, where low values (Ld = 100) indicate low anthropogenic intervention, whereas high values (Ld = 400) reflect intense land transformation, as seen in urban areas.
This method allows for a quantitative and comparative assessment of land use changes, facilitating the identification of spatial and temporal patterns in landscape transformation. Furthermore, the LUDI is a key tool for territorial planning and environmental assessment, as it integrates information on the magnitude of human activity and its impact on natural ecosystems.” – page number 12, paragraph 6, page 13, paragraph 1 and 2 and line 352-380.]
[Updated text in the manuscript: “The Land Use Degree Index (LUDI) is a quantitative indicator that measures the magnitude and depth of transformations in land use/land cover (LULC). This index facilitates the analysis of the interaction between natural systems and anthropogenic activities in landscape configuration, providing a comprehensive perspective on land occupation [103]. The degree of land use is defined at multiple levels based on the equilibrium state of natural ecosystems under socioeconomic influences, thereby establishing a classification index (Table 3). For its calculation, an intensity value is assigned to each land use category to reflect the degree of anthropogenic alteration. The LUDI is computed as the weighted sum of the percentage of area occupied by each category and its respective intensity index, as expressed in the following equation:
where: ● Ld represents the comprehensive land use intensity index for the study area. ● Bi corresponds to the intensity index assigned to the category. ● Ci indicates the percentage of the surface occupied by the category. ● i denotes the land use category number. ● n is the total number of evaluated categories. Table 3. Assignment of the Land Use Degree (LUDI) Classification Index
.” – page number12, paragraph 5 and 6, and line 280-290.] |
||||||||||||||||||||||
|
|
|
SPATIAL AUTOCORRELATION Response 16: In response to the reviewer’s suggestions, the first paragraph has been removed. Additionally, the content originally in the third paragraph has been repositioned to serve as the opening of this section, and a brief explanation of Local Indicators of Spatial Association (LISA) has been incorporated, along with appropriate citations [104,105]. [Text to be corrected: “The global spatial autocorrelation model is useful for identifying the overall pattern of spatial distribution of a variable in a given region. However, this model does not quantify the spatial association between a specific element and its adjacent regions [104]. For this reason, it is essential to analyze the degree of correlation between a local geographic feature and similar features in neighboring areas.
The Moran's I index quantifies the similarity of attribute values among spatially adjacent units. Its value ranges from -1 to 1: a value greater than 0 indicates positive spatial autocorrelation; a value equal to 0 suggests that there is no spatial correlation; and a value less than 0 indicates negative spatial autocorrelation, implying a dispersed pattern [105]. To examine the spatial autocorrelation characteristics in the degree of land use, this study implemented two complementary techniques: the global Moran's I index and Local Indicators of Spatial Association (LISA) [106]. The combined application of these methods enables the identification and characterization of spatial clustering patterns, facilitating an accurate evaluation of the variations in land use intensity throughout the study area.” – page number 14, paragraph 4,5 and line 282-396.]
[Updated text in the manuscript: “To analyze the characteristics of spatial autocorrelation in land use intensity, this study implemented two complementary techniques: the global Moran’s I index and the Local Indicators of Spatial Association (LISA) [104,105]. The LISA facilitate the identification of significant local clusters and spatial outliers, thereby complementing the global analysis provided by the Moran’s I index. The combined application of these methods characterizes spatial clustering patterns and enables an accurate assessment of variations. The Moran’s I index quantifies the similarity of attribute values among spatially adjacent units. Its value ranges from –1 to 1, where a value greater than 0 indicates positive spatial autocorrelation; a value equal to 0 suggests no spatial correlation; and a value less than 0 indicates negative spatial autocorrelation, implying a dispersed pattern [106].” – page number12, paragraph 2 and 3, and line 317-328.]
|
|
SELECTION AND CALCULATION OF LANDSCAPES PATTERNS A whole series of details are given about the indices. Is there any variation made by the authors compared to the traditional indices? If no variation has been made, Table 6 can be included as an appendix.”] Response 17: We have followed your suggestions and removed the second paragraph of this section. No modifications have been made to the traditional indices; therefore, Table 6 is included in the Appendix.
According to hierarchical systems theory, landscape indices are classified into three scales: patch metrics, class metrics, and landscape metrics [101]. For this analysis, class and landscape metrics were prioritized given their relevance in characterizing the fragmentation, diversity, and spatial configuration of the study area.
For this study, six class metrics were selected: Patch Density (PD), Edge Density (ED), Largest Patch Index (LPI), Fractal Dimension of the Perimeter Area (FRAC), Cohesion, and Division; and four landscape metrics: Contagion Index (CONTAG), Aggregation Index (AI), Shannon Diversity Index (SHDI), and Shannon Evenness Index (SHEI). All landscape pattern indices were calculated using Fragstats 4.2. Table 6 describes the formula and definition of the landscape pattern indices.” – page number 14 and 15, paragraph 6,8 and 7 and line 298-413.]
[Updated text in the manuscript: “Landscape pattern indices are quantitative tools that condense information on the structure and spatial configuration of the landscape, reflecting its composition and temporal dynamics. These indices are widely used to assess the impact of LULC change on landscape configuration, as this phenomenon is one of the primary determinants of its spatial pattern [107,108].
For this study, six class metrics were selected: Patch Density (PD), Edge Density (ED), Largest Patch Index (LPI), Fractal Dimension of the Perimeter Area (FRAC), Cohesion, and Division; and four landscape metrics: Contagion Index (CONTAG), Aggregation Index (AI), Shannon Diversity Index (SHDI), and Shannon Evenness Index (SHEI). All landscape pattern indices were calculated using Fragstats 4.2. For a detailed description of the formulas and definitions, please refer to the Appendix (Table A1).” – page number12, paragraph 5 and 6, and line 329-341.]
|
|
LULC CLASSIFICATION AND VALIDATION In Figure 4, it is unclear what each of the 11 maps represents; this should be indicated in the figure description. Additionally, the land uses are not clearly visible, so it might be helpful to make them larger. This section could be combined with the next one, as it continues discussing the same topic, rewritten the title, for example, Analysis of LULC changes.”] Response 18: We appreciate your suggestions. The following modifications have been made to the manuscript: Revision of the First Sentence, the wording of the opening sentence has been modified to correct its structure and improve clarity.
Clarification of Figure 4: The content of each of the 11 maps has been specified, including the conventions corresponding to each year in the figure description. In addition, the images have been enlarged to enhance their visualization, and a complete version in excellent resolution has been attached in the figures folder.
Unification of Sections: The content has been consolidated into a single section entitled 3.1 Spatiotemporal Analysis of LULC Changes in the CCLC, which is now subdivided into:
The LULC classification for the period 2014–2024 identified seven main categories: Buildings, Coffee crops, Bareland, Forest, Grasslands, Bamboo, and Water. The spatial and temporal distribution of these categories is presented in Figure 4, which shows the predominance of coffee crops, a gradual expansion of building areas, and a significant reduction in grassland cover.” – page number 12 and 15, paragraph 1 and line 417-422.]
Figure 4. Maps of the spatial distribution of CCLC LULC (2014–2024).
[Updated text in the manuscript: “3.1 Spatiotemporal Analysis of LULC Changes in the CCLC 3.1.1 LULC Classification and Validation for the period 2014–2024, seven LULC classes were defined: Buildings, Coffee Crops, Bareland, Forest, Grasslands, Bamboo, and Water. Figure 4 presents the spatiotemporal distribution through 11 annual maps, clearly highlighting the predominance of coffee crops, the expansion of built-up areas, and the reduction of grasslands.” – page number13, paragraph 5 and 6, and line 343-348.]
Figure 4. Annual spatial distribution of land use and land cover in the Colombian Coffee Cultural Landscape (CCLC): (a) 2014; (b) 2015; (c) 2016; (d) 2017; (e) 2018; (f) 2019; (g) 2020; (h) 2021; (i) 2022; (j) 2023; (k) 2024. – page number15, line 349-351.]
|
|
CHARACTERISTICS ANALYSIS OF LULC STRUCTURE sentences that are unnecessary and quite standard, such as "the analysis of the total area and its percentage... a comprehensive understanding of LULC change dynamics." In the result section just the results should be detailed, not what they allow us to do, which, in any case, should be mentioned in the methodology. In forest areas the years of the % and ha data are missing. The paragraph “these patterns of changes suggest... by progressive urban-rural development” can be removed. Table 8 could be removed, as Figure 5 already provides a summary of the evolution, and it is not necessary to present it in such detail, especially when Table 9 also offers a summary of this table. If the authors consider it important, they could include it as an annex.”] Response 19: In response to the reviewers' suggestions, the following modifications have been made to the manuscript: Redundant expressions such as “the analysis of the total area and its percentage... an integral understanding of the dynamics of LULC change” have been removed. The results are now presented directly, including the specific data by year and the forest areas expressed in hectares. The paragraph concluding with “these change patterns suggest... a progressive urban–rural development” has been eliminated. Finally, Table 8 has been incorporated as supplementary material and is now designated as “Table S2.”
The structure and composition of LULC during the 2014–2024 period reveal significant dynamics in territorial transformation, characterized by heterogeneity in the distribution and evolution of land cover. The quantitative analysis of spatiotemporal patterns, expressed in hectares (ha) and percentages (%), as shown in Figure 5 and Table 8, reveals substantial changes in the seven identified cover categories.
The analysis of the total area and its percentage distribution over different periods allowed for a comprehensive understanding of LULC change dynamics. Coffee crops represents the dominant land cover, exhibiting a notable increase from 69.38% (96855.25 ha) in 2014 to 77.91% (108690.67 ha) in 2019, followed by a gradual decrease to 72.89% (101601.42 ha) in 2024, indicating an intensification of coffee cultivation activity in the first half of the period. Grasslands experienced the most drastic transformation, with a reduction from 15.71% (21934.85 ha) to 5.02% (6998.83 ha) during the study period.
Bare land increased from 1.83% (2558.99 ha) in 2014 to 7.44% (10315.12 ha) in 2018, reaching its maximum of 8.09% (11277.60 ha) in 2024. Forest areas maintained a variable yet relatively stable presence, ranging between 5.01% (6999.22 ha) and 7.87% (10984.44 ha).
Bamboo cover exhibited moderate fluctuations between 4.57% (6348.62 ha) and 6.69% (9322.93 ha). Building areas showed steady but moderate growth, increasing from 1.46% (2040.41 ha) to 2.35% (3273.92 ha), with the most significant expansion occurring between 2014 and 2016 (an increase of 1.05%), reflecting progressive urban–rural development. Water surfaces consistently maintained the smallest proportion of the territory, fluctuating between 0.30% (421.09 ha) and 1.57% (2165.11 ha).
These patterns of change suggest a significant transformation of the landscape, characterized primarily by the consolidation of coffee cultivation, the reduction of grasslands, and an increase in bare land, accompanied by progressive urban–rural development.” – page number 17, paragraph 2,6 and 7 and line 437-468.]
[Updated text in the manuscript: “3.1.2 Characteristic Analysis of LULC Structure (2014–2024) The quantitative analysis of spatiotemporal patterns, expressed in hectares (ha) and percentages (%) are presented in Figure 5 and Table S2, reveals the changes across the seven identified land cover categories during the 2014–2024 period. Coffee, the dominant cover, increased from 69.38% (96,855.25 ha) in 2014 to 77.91% (108,690.67 ha) in 2019, followed by a gradual decrease to 72.89% (101,601.42 ha) in 2024. Grasslands experienced a reduction from 15.71% (21,934.85 ha) in 2014 to 5.02% (6,998.83 ha) in 2024. Bareland increased from 1.83% (2,558.99 ha) in 2014 to a range of 7.44–8.09% (reaching 11,277.60 ha) in 2024. Forest areas decreased from 5.01% (6,999.22 ha) in 2014 to 4.70% (6,537.58 ha) in 2019, then increased to 5.45% (7,598.64 ha) in 2024. Bamboo cover fluctuated, rising from 6.30% (8,787.77 ha) in 2014 to 6.69% (9,322.93 ha) in 2020 and then declining to 5.35% (7,452.90 ha) in 2024. Built-up areas grew steadily from 1.48% (2,066.30 ha) in 2014 to 2.51% (3,508.12 ha) in 2024, with the most notable expansion occurring between 2014 and 2016 (an increase of 1.05%). Lastly, the water surface, remaining the smallest cover, increased from 0.30% (421.09 ha) in 2014 to 0.85% (1,185.96 ha) in 2024.” – page number16, paragraph 5 and 6, and line 369-383.]
|
|
LULC PREDICTION represented larger to better observe the evolution, or the authors could consider showing a map that highlights the land use changes, rather than the land uses themselves. It is noteworthy that some values in Table 9 do not match those in Table 8. What is the reason for this discrepancy? There is no mention of the location of the land uses or the changes themselves.”] Response 20: We appreciate your comments, which have been addressed as follows:
Restructuring of the Prediction Section: Section 3.2.2, LULC Prediction Analysis for 2034, was reorganized to enhance coherence and eliminate redundancy. It is now presented as Section 3.2, 2034 LULC Prediction, and subdivided into Section 3.2.1, Prediction-based Changes.
Cartographic Update: Previous maps were replaced with larger versions to improve visual clarity. In addition, comparative maps highlighting spatial changes between periods were incorporated (Figure 7).
Correction of Table Inconsistencies: Discrepancies were identified between the values reported in Table 9 and those in Table 8. These inconsistencies were corrected, ensuring that both tables are derived from the same data source and spatial extent. Table 8 has been moved to the supplementary materials and is now referred to as Table S2.
[Text to be corrected: “3.2.2 LULC Prediction Analysis for 2034 The analysis of LULC changes during the periods 2014–2019, 2019–2024, and 2024–2034 reveals significant spatial variations. According to the results presented in Figure 6 and Table 9, substantial transformations were observed in the analyzed land cover types. The urbanized area exhibited a non-linear expansion pattern, increasing from 2066.30 ha, 1.47% in 2014 to 2567.72 ha, 1.83% in 2019, representing an increase of 24.27%. This growth accelerated between 2019 and 2024, reaching 3508.12 ha, 2.51%, corresponding to an additional increase of 36.62%. The projection for 2034 indicates an expansion up to 23726.66 ha, 16.95%.
Coffee crops, which constitute the dominant land cover, experienced notable fluctuations. Initially, they expanded from 96855.25 ha, 69.17% in 2014 to 108298.52 ha 77.36% in 2019, showing an increase of 11.81%. However, this trend reversed in subsequent periods, decreasing to 101601.42 ha , 72.60% in 2024, with a further reduction projected to 86679.92 ha, 61.89% for 2034, representing a cumulative loss of 14.69% relative to 2024. Natural areas exhibited significant transformations. Forest cover displayed a non-linear decreasing trend, declining from 6999.22 ha in 2014 to 6537.58 ha in 2019 a decrease of 6.60%, then experiencing a temporary increase to 7598.64 ha in 2024 an increase of 16.23%; however, the projection for 2034 indicates a drastic reduction to 1489.63 ha, representing a loss of 80.40% relative to 2024. Grasslands experienced the most dynamic transformation among natural covers. Starting from 21934.85 ha in 2014, they decreased to 5791.18 ha in 2019, a decline of 73.60%, followed by a moderate increase to 6998.83 ha in 2024, an increase of 20.85%. The model projects a further increase for 2034, reaching 8852.86 ha, which represents an increase of 26.49% relative to 2024. Secondary covers also exhibited significant changes. Bamboo showed an expansion trend, increasing from 8787.77 ha in 2014 to 14454.67 ha in 2034, an increase of 64.48%. Bare land displayed high temporal variability, increasing from 2558.99 ha in 2014 to 11277.60 ha in 2024, then decreasing to 3554.86 ha in 2034, a decline of 68.48%. Water bodies experienced an initial growth from 421.09 ha in 2014 to 1,185.96 ha in 2024, with a projected reduction to 864.89 ha in 2034, a decline of 27.07%.” – page number 12 and 13, paragraph 5,6 and 7 and line 280-300.]
[Updated text in the manuscript: “3.2 LULC Changes 2014–2034 The analysis of LULC dynamics between 2014 and 2034 reveals a continuous process of territorial transformation, characterized by urban expansion, the loss of natural areas, and the conversion of agricultural land cover. Figure 7 displays the LULC change maps corresponding to the periods 2014–2019, 2019–2024, and 2024–2034, respectively, while Table S3 details the transitions (in hectares) between categories for each interval. During the period 2014–2019, the transition from Bareland to Building was particularly notable, with 10,396 hectares converted; this represented the most significant transformation, followed by the change from Coffee to Building with 1,894 hectares. Additionally, 1,252 hectares of Forest were transformed into Bareland, while 1,515 hectares of Grassland were lost in favor of urban and agricultural uses. Urban expansion accumulated an increase of 14,000 hectares, concentrating the majority of the conversions from vegetative or agricultural covers. Between 2019 and 2024, the conversion from natural covers to urban uses persisted. The most pronounced transitions included Bareland to Building with 11,738 hectares, followed by Grassland to Building with 3,563 hectares. During this period, conversions from Coffee and Forest to Grasslands also occurred, accounting for 3,514 hectares and 2,174 hectares, respectively. The spatial pattern illustrated in the map indicates an intensification of urban dispersion and a progressive fragmentation of the forest. Finally, during the period 2024–2034, projected based on modeling of previous transitions and spatial variables, an acceleration in Building expansion is anticipated, with an estimated increase of 17,374 hectares originating from Bareland (10,113 hectares), Grassland (3,901 hectares), and Coffee (1,512 hectares). Furthermore, a continuous loss of forest cover is foreseen, with 2,508 hectares converted primarily into Bareland and Grasslands. The figures indicate that the transformation will be concentrated in already active urban expansion corridors, thereby compromising areas of ecological connectivity.” – page number 17 , paragraph 1 , page number 18 paragraph 1,2,3 and line 391-416.] Figura 7. Change maps 2014–2034 in CCLC
|
|
LANDSCAPE TRANSFORMATION ANALYSIS Response 21: Considering the suggestions, we have integrated section 3.2.3 “Landscape Transformations: Analysis of LULC Gains and Losses” into section 3.2 “LULC Changes 2014–2034”. The text has been reformulated as follows:
The analysis of LULC gains and losses over the period 2014–2034 (Figure 7) reveals significant transformation patterns based on the temporal intervals used for LULC prediction. During the first period 2014–2019, a considerable expansion of coffee crop areas was observed (+11443.27 ha), in contrast to a pronounced reduction in grasslands (-16143.67 ha). Building areas registered moderate growth (+501.42 ha), whereas natural land covers, such as forests, experienced a slight decrease (–461.64 ha).
The second period (2019–2024) exhibited different dynamics, characterized by a significant expansion of bare land (+5722.08 ha) and a substantial decrease in coffee crop areas (- 6697.10 ha). Building areas maintained their growth trend (+940.40 ha), whereas forests showed a moderate recovery (+1061.06 ha). Bamboo experienced a notable reduction (–1801.93 ha), while grasslands exhibited a slight recovery (+1207.65 ha).
The final projection period (2024–2034) shows the most significant transformations. Building areas exhibit a substantial increase (+20218.54 ha), accompanied by a considerable reduction in coffee crop areas (–14921.50 ha) and bare land (–7722.74 ha). Natural ecosystems display mixed trends: forests experience severe losses (–6109.01 ha), while grasslands (+1854.03 ha) and bamboo (+7001.77 ha) show significant gains. Water bodies maintain minor fluctuations throughout all periods, with a final loss of 321.07 ha in the last period.
These change patterns suggest an intensification of urbanization at the expense of traditional productive areas and natural ecosystems, particularly during the 2024–2034 period. The observed transformation indicates a substantial reconfiguration of the landscape, with natural and agricultural land covers being replaced by urbanized areas, which has significant implications for territorial sustainability. ” – page number 20 and 21, paragraph 1 -4, and line 511-557.]
[Updated text in the manuscript: “Analysis of the Gains and Losses in LULC Figure 8 and Table S4. During the first period (2014–2019), an expansion in coffee cultivation was observed (+30,255.81 ha), in contrast with a reduction in grasslands (–19,634.14 ha). Built-up areas exhibited moderate growth (+1,511.91 ha), whereas natural covers such as forests showed a relative loss (–5,481.41 ha). Additionally, bare soil areas increased (+5,265.54 ha).
The second period (2019–2024) was characterized by a dynamic expansion of bare soil (+10,199.34 ha) and a reduction in coffee cultivation (–26,937.27 ha). Built-up areas maintained their growth trend (+2,116.06 ha), while forests experienced a moderate recovery (+6,363.20 ha). Likewise, grasslands registered an increase (+6,107.44 ha), in contrast with the loss of bamboo observed during the previous period. The final period (2024–2034) exhibited drastic transformations. Built-up areas recorded the highest increase (+20,839.17 ha), at the expense of a decrease in coffee cultivation (–11,519.86 ha) and bare soil (–8,230.56 ha). Natural covers exhibited mixed behaviors: forests suffered considerable losses (–6,058.02 ha), whereas grasslands (+5,554.49 ha) and bamboo (+7,001.77 ha) expanded.” – page number 18 and 19, paragraph 1,2,3, and line 425-440.]
Note: the data have changed because climate variables were added to the prediction, following the reviewers' recommendation; therefore, you will find differences in the results when comparing the texts.
|
|
Comments 22: [“I wonder why the land use projection in Figure 7 is so different compared to the previous years, especially if, as stated, the authors have created a "business as usual" scenario. Is that correct? The type of scenario used should be clearly explained in the methodology. Response 22: We thank the reviewer for raising this important question regarding the visual difference between the projected 2034 LULC map and the maps from previous years, especially in the context of the Business as Usual (BAU) scenario employed.
(2) Future LULC Prediction, and (3) Analysis of the Spatiotemporal Characteristics of LULC Changes and Landscape Patterns, as presented in Figure 2. 20,”– page number 6, paragraph 1 and line 160-162.]
[Updated text in the manuscript: ““This study is structured into three main components: (1) LULC classification, (2) projection of future LULC under a Business as Usual (BAU) scenario, and (3) spatiotemporal analysis of LULC changes and landscape patterns”– page number 6, paragraph 1, and line 160-162.]
[Updated text in the manuscript: “The modeling was conducted under a Business as Usual (BAU) scenario. This scenario assumes that the land use transition probabilities and spatial driving forces observed during the calibration period (2019-2024) will persist into the projection period (2024-2034). Consequently, the model extrapolates existing trends, including accelerated transformations if such dynamics were detected in the recent past.”– page number 12, paragraph 1, and line 262-266. ]
|
|
LAND USE TRANSFER some land uses. What is the reason for this?”] Response 23: We appreciate your recommendation. In response to your suggestions, the following revisions have been made:
The transfer matrix allowed the quantification of conversions among different LULC categories, as presented in Table 10. For its representation, chord diagrams were employed (Figure 8), with the thickness of the links indicating the magnitude of the transfers. The analysis was conducted for the periods 2014–2019, 2019–2024, and 2024–2034. During 2014–2019, coffee crops exhibited high stability, maintaining 76.4%, 78042.71 ha of their original area, although transfers occurred toward grasslands 16.1%, 16456.88 ha and bamboo 6.3%, 6423.15 ha. Grasslands retained 21.7%, 2212.15 ha of their initial cover, transferring 3.3%, 337.66 ha to building areas and 14.0%, 1433.36 ha to bare land. Forest areas maintained 14.6%, 1488.49 ha, transferring 47.3%, 4834.76 ha to coffee crops and 1.9%, 192.44 ha to bare land. The period 2019–2024 demonstrated a consolidation in coffee crops with a retention of 79.7%, 81361.24 ha, although notable transfers occurred toward bare land 8.0%, 8200.29 ha and bamboo 5.6%, 5712.38 ha. Building areas increased their retention to 53.5%, 1373.13 ha, incorporating 1.6%, 1612.91 ha from coffee crops. Grasslands reduced their retention to 8.5%, 868.23 ha, transferring 36.1%, 3683.05 ha to coffee crops. During 2024–2034, a more pronounced transformation was observed. Coffee crops maintained 81.8%, 83526.43 ha of their area, but transferred 12.6%, 12826.72 ha to bamboo and 5.1%, 5241.61 ha to grasslands. Bare land exhibited a retention of 27.4% 2802.47 ha, receiving a transfer from building areas 73.4%, 7495.13 ha. Grasslands retained 26.9%, 2745.65 ha, while forest areas preserved 14.6%, 1487.69 ha. The analysis of the complete period (2014–2034) revealed significant cumulative patterns. Coffee crops maintained a permanence of 76.4%, 74029.52 ha, with transfers to building areas of 22.5%, 21750.55 ha and minimal transfers to grasslands 0.1%, 108.99 ha. Grasslands retained 27.7%, 6082.01 ha, transferring 22.6%, 4952.91 ha to coffee crops. Forest areas preserved 18.8%, 1319.06 ha, transferring 55.3%, 3870.29 ha to coffee crops. These dynamics suggest three dominant territorial processes: Intensification of coffee cultivation, evidenced by its high retention (>75%) and significant absorption of other covers, particularly forests and grasslands; Progressive urbanization, reflected in the steady increase of building areas, shifting from a retention of 53.5% to an expansion that incorporated more than 22% of coffee areas; Transformation of natural covers, manifested in the reduction of forests—retaining only 18.8%—and the fluctuation of grasslands with a final retention of 27.7%. ” – page number 22 and 13, paragraph 1-6 and line 539-579.]
[Updated text in the manuscript: “ 3.2.4 Land Use Transfer Analysis The transition matrix enabled the quantification of conversions among Land Use and Land Cover (LULC) categories for the periods 2014–2019, 2019–2024, 2024–2034, and the cumulative period 2014–2034 (Table S3). Chord diagrams were employed for visualization (Figure 8), where the thickness of the links represents the magnitude of transitions. Between 2014 and 2019, coffee crops retained 80.9% (78,042.71 ha) of their original area, while 2.9% (2,782.34 ha) transitioned to grasslands and 7.4% (7,105.42 ha) to bamboo. Grasslands preserved only 10.1% (2,212.15 ha) of their initial cover, with losses to built-up areas (1.5%, 337.66 ha) and bare land (6.6%, 1,433.36 ha). Forest areas maintained 21.4% (1,488.49 ha), with significant transfers to coffee (69.4%, 4,834.76 ha) and bare land (2.8%, 192.44 ha). In 2019–2024, coffee retained 79.5% (81,361.24 ha), but transferred 8.0% (8,200.29 ha) to bare land and 5.6% (5,712.38 ha) to bamboo. Built-up areas preserved 53.5% (1,373.13 ha) and incorporated 1.6% (1,612.91 ha) from coffee zones. Grasslands showed a reduced persistence of 10.1% (868.23 ha), with 17.2% (3,683.05 ha) converted to coffee. During 2024–2034, coffee retained 81.7% (90,081.57 ha), with transfers of 8.4% (919.98 ha) to forest and 7.3% (6,578.64 ha) to bamboo. Bare land maintained 27.2% (3,047.04 ha), receiving 17.3% (7,267.85 ha) from built-up areas. Grasslands retained 1.0% (2,682.10 ha), while forest areas preserved 14.2% (1,540.63 ha).” – page number19, paragraph 2,3, and line 443-461]
|
|
ANALYSIS OF THE COMPREHENSIVE LAND-USE DEGREE It would be interesting to comment on the location of what is observed in the maps.”] Response 24: We appreciate the reviewer’s thoughtful observation regarding the need for clearer justification and spatial specification in our interpretation of land occupation patterns. In response, we revised the paragraph to explicitly link the LUDI-based interpretation to spatial analysis results. Two key spatial trends were incorporated.
” – page number 12 and 13, paragraph 5,6 and 7 and line 280-300.]
[Updated text in the manuscript: “The spatial analysis of the LUDI index between 2014 and 2024 reveals two distinct territorial dynamics in the CCLC. First, a persistent concentration of high-intensity land use is evident in the central and southwestern zones, particularly around the urban peripheries of Armenia, Calarcá, and La Tebaida, where High-High LISA clusters emerge with statistical significance. These areas reflect sustained urban growth and spatial consolidation of land transformation. Second, a gradual decline in Moran’s I values from 2014 to 2024 suggests an early phase of spatial dispersion of moderately transformed areas (LUDI 200-300), expanding into eastern and northern rural municipalities such as Pijao and Filandia. Together, these patterns illustrate a dual trend of urban densification and rural fragmentation, with implications for landscape heterogeneity and future governance challenges.” – page number21, paragraph 2, and line 483-493.]
|
|
SCALE ANALYSIS OF LANDSCAPE-LEVEL CHANGE Perhaps Figure 14 could also be included as an annex.”] Response 25: We appreciate your suggestion to integrate the scale analysis of landscape-level changes into the previous section. In response, we have merged the content of Sections 3.4.1 and 3.4.2 into a unified subsection under the title 3.4 Spatiotemporal Characteristics of Landscape Patterns.” – page number 12 and 13, paragraph 5,6 and 7 and line 280-300.]
[Updated text in the manuscript: “The period 2014-2034 reveals a growing trend toward fragmentation in the CCLC, with differentiated spatial behaviors among land cover types. At the class level, the patch density index (PD) exhibits a significant increase in building areas, rising from 4.7 in 2014 to 37.1 in 2034, indicating an accelerated urbanization process. In contrast, coffee crops show a slight decrease (from 4.3 to 3.7), and natural covers such as forests and grasslands demonstrate a marked reduction (from 10.9 to 4.9 and 22.2 to 11.6, respectively), suggesting consolidation. Bareland increases substantially (from 5.8 to 19.0), while bamboo and water bodies reflect divergent trends, with bamboo decreasing from 23.1 to 5.7 and water increasing from 1.3 to 16.1, suggesting greater fragmentation or proliferation of aquatic features.
The edge density index (ED) also presents notable changes. Coffee crops increase in internal fragmentation (from 159.48 to 186.8), while forest cover shows simplification (from 36.36 to 10.8). Building areas increase sharply in edge complexity (from 10.25 to 121.7), and bareland and water bodies also reflect this trend. In contrast, grasslands and bamboo decrease notably in edge complexity. The trends in PD and ED are illustrated in Figure 13.
The Largest Patch Index (LPI) shows that coffee crops maintain stable dominance with a minor decline (from 21.84 to 21.79), while forest, grassland, and bamboo decrease markedly, indicating loss of dominant continuous patches. Bareland and water maintain low values, and building areas remain constant at 0.11, highlighting dispersed urban growth.
The Perimeter-Area Fractal Dimension (PAFRAC) reflects increasing complexity in urban patches (from 1.44 to 1.5834) and moderate complexity in coffee crops and bareland. Conversely, natural covers (forest, grassland, bamboo) show slight simplification of patch shapes, while water bodies increase slightly in complexity.
The COHESION index illustrates contrasting connectivity patterns. Building areas remain relatively stable (from 78.21 to 77.64), while coffee crops maintain high cohesion (~99.83). Forest and grassland undergo fragmentation (falling from 79.16 to 58.74 and 86.04 to 51.19, respectively), and bamboo also declines (from 63.60 to 50.20). Interestingly, water bodies increase in cohesion (from 64.10 to 70.17), possibly due to expansion or interconnection of aquatic patches.
The DIVISION index shows a slight increase in fragmentation for coffee crops (from 0.90 to 0.908), while all other covers—including building, bareland, forest, grassland, bamboo, and water—maintain a value of 1.00, indicating a consistently high level of spatial separation and low internal cohesion.
At the landscape level, broader patterns are evident. The Contagion Index (CONTAG) increases in 2019 and 2024 compared to 2014, suggesting more aggregated land cover configurations. However, in 2034, the decrease in CONTAG values reflects a fragmented landscape structure, primarily driven by urban expansion. The spatial maps show greater clustering of natural covers in 2019-2024, and more dispersion by 2034, especially in central areas of the territory.
The Aggregation Index (AI) further supports this, with minimum values declining from 10.34 in 2014 to 7.24 in 2034, particularly in central zones affected by urban development. In contrast, maximum AI values remain stable (between 95.40 and 95.59), especially in the northern and southern rural areas, dominated by forest and water.
The Shannon Diversity Index (SHDI) and Shannon Evenness Index (SHEI) show that zones with high diversity and balanced distribution are consistently located in the central and southern regions, while the north remains more homogeneous. Over time, the increase in SHDI and SHEI values reflects greater heterogeneity in land composition and spatial distribution. These trends are illustrated in Figure 14, which shows the spatial configuration of CONTAG, AI, SHDI, and SHEI in 2034. This pattern, shaped by mountainous topography and urban development, emphasizes the progressive fragmentation and decline of ecological connectivity in the CCLC. ”– page number 26- 27, line 538-600.]
|
|
LIMITATIONS cartography process or how the persistence of land uses can sometimes obscure the results (see Pontius' articles).”] Response 26: In line with the suggestions provided, the paragraph beginning with “The selected indices included fragmentation metrics…” has been removed, as has the final paragraph “This process generated grid maps…”.
Among the study's limitations are: (1) Although the simulations were validated with a high Kappa coefficient, the climatic variables—precipitation and temperature—that influence agricultural productivity and the natural dynamics of the landscape were omitted [110]. (2) The absence of an analysis of the CCLC buffer zone, which would limit the understanding of the extent and displacement of LULC changes and their impacts on landscape patterns. Future studies should incorporate these variables and assess the influence of buffer zones on ecological connectivity and territorial sustainability.” – page number 35, paragraph 6 and line 216-823.]
[Updated text in the manuscript: “Despite achieving an overall accuracy of 88%, the remote sensing analysis carries uncertainties, particularly due to confusion among spectrally similar land cover types. As noted by Pontius et al.[124], high persistence of dominant classes may obscure significant changes in minor but ecologically important covers. In this regard, the future use of intensity analysis and error-adjusted metrics would allow for more precise differentiation between actual changes, random variations, and apparent persistence..” – page number30, paragraph 7, and line 699-704.
|
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThis manuscript mainly focuses on the change in land use in the Colombian coffee cultural landscape, and the study area is innovative and engaging. Meanwhile, this study used solid data, advanced AI models, and statistical and spatial methods to complete the data analysis. The results of this study are very reliable and explain the objectives and answer the major questions in this study. However, some minor problems need to be addressed and modified before it can be published. My recommendation for this manuscript is acceptance with minor revisions.
Comments for author File: Comments.pdf
Author Response
|
Thank you very much for taking the time to thoroughly review our manuscript and for providing valuable feedback that has significantly improved its clarity. We have carefully addressed each of your comments and implemented the corresponding revisions, clearly highlighted in the updated manuscript. These include clarifications of terminology, simplification and specification of the manuscript title, corrections to equations and tables, improved descriptions of coordinate systems, and the addition of explanatory legends in figures. We sincerely appreciate your insightful suggestions and detailed observations, which have helped enhance the scientific rigor and readability of our article.
. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
2. Questions for General Evaluation |
Reviewer’s Evaluation |
Response and Revisions |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Does the introduction provide sufficient background and include all relevant references? |
Yes/Can be improved/Must be improved/Not applicable |
[Please give your response if necessary. Or you can also give your corresponding response in the point-by-point response letter. The same as below] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Are all the cited references relevant to the research? |
Yes/Can be improved/Must be improved/Not applicable |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Is the research design appropriate? |
Yes/Can be improved/Must be improved/Not applicable |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Are the methods adequately described? |
Yes/Can be improved/Must be improved/Not applicable |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Are the results clearly presented? |
Yes/Can be improved/Must be improved/Not applicable |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Are the conclusions supported by the results? |
Yes/Can be improved/Must be improved/Not applicable |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 1: [1. This manuscript does not have the line number for each row, which make it very hard to make revised comments on it.] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 1 : We apologize for the inconvenience.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 2: [The title of this manuscript can be simplified and the current title is too complicated. “Evolution and impacts of land-use and land-cover spatiotemporal dynamics on the landscape patterns of Colombian Coffee Cultural (2014-2034).] Response 2 : Thank you for your suggestions. The title has been revised considering both complexity and clarity. Redundant phrases such as "Evolution of the Spatiotemporal Dynamics" were removed in favor of the simplified and precise expression "Spatiotemporal Land Use and Land Cover Changes," preserving the study's spatiotemporal focus without compromising conceptual accuracy. The improved title maintains the fundamental aspects of the research: Spatiotemporal changes in Land Use and Land Cover (LULC). Their impact on landscape patterns (assessed through landscape metrics). The specific geographic context of the Colombian Coffee Cultural Landscape (CCLC). These modifications optimize both the clarity and accuracy of the title, aligning with the provided recommendations. Thank you for your contribution; your comment is acknowledged and agreed upon. Accordingly, the text has been modified as follows: [Text to be corrected: “Evolution of the Spatiotemporal Dynamics of Land Use and Land Cover and Their Impact on the Landscape Patterns of the Colombian Coffee Cultural Landscape (2014–2034). ” – page number 1 , and line 2-4. ] [Updated text in the manuscript: “Spatiotemporal Land Use and Land Cover Changes and Their Impact on Landscape Patterns in the Colombian Coffee Cultural Landscape (2014–2034).” – page number 1 , and line 2-4. ]
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 3: [Introduction line 2, what is UNESCO? Should not use the abbr.] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 3 : Thank you for your comments. We apologize for not clarifying that "UNESCO" stands for the United Nations Educational, Scientific and Cultural Organization. We have made the corrections. [Text to be corrected: “Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by UNESCO [1]” – page number 2 , paragraph 1, and line 39-40.] [Updated text in the manuscript: “Cultural landscapes, recognized for their Outstanding Universal Value (OUV) by the United Nations Educational, Scientific and Cultural Organization UNESCO [1],” page number 2 , paragraph 1, and line 39-40. ]
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 4: [Equation (4) has mistakes.] Response 4 : We appreciate the thoroughness of your review and thank you for drawing our attention to these issues. We apologize for any oversight and have now corrected the equation as suggested. If you do not see the equations in the updated manuscript, it is because we decided to remove them based on additional feedback from other reviewers. Since this methodology has already been widely applied, we will simply reference published articles where it has been employed.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
[Text to be corrected: “Kappa Coefficient = 100 ” – page number 11 ,equation 4.] [Updated text in the manuscript:
Kappa Coefficient = ” page number 11 , equation 4. ]
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 5: [2.3.2.1 the coordinate system is UTM_ZONE_18N? I think Colombia should be 18 S?] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 5 : Part of Colombia's territory is located north of the equator (mainly between latitudes 12°N and 0°), while certain regions in the southwest of the country may be located south of the equator (up to approximately 4°S). Consequently, depending on the specific geographic location, UTM zones may be used north or south of the equator. However, the Coffee Cultural Landscape of Colombia is located north of the equator, approximately between latitudes 4°N and 6°N. Consequently, it falls within UTM Zone 18N. page number 11 , paragraph 1, and line 273.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 6: [2.3.3.1 the equation has typo in it, and this equation does not have a number] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 6 : We have implemented the necessary changes as requested, as outlined below.
[Text to be corrected: “To analyze LULC changes, transition matrices were employed as a fundamental tool [87]. In the matrix, rows and columns correspond to the land use types in two distinct periods (T₁ and T₂). The values within the matrix (Pₙₙ, where n = 1,2,…,n) represent the areas that have experienced changes among the different land use types during both periods. Based on previous studies [88,89], and with the objective of simplifying its interpretation and application, the matrix is expressed by the following equation:
M??= (?,?=1,2,3…?)
where: Mᵢⱼ represents the land area transformed from type i to type j. i and j denote the land use types before and after the transition, respectively. When, i = j, it represents the unchanged land area. n is the total number of land use types (n = 7 in this study). ” – page number 13 , paragraph 4, and line 312-330.] [Updated text in the manuscript: “To analyze LULC changes, transition matrices were employed as a fundamental tool [87]. In the matrix, rows and columns correspond to the land use types in two distinct periods (T₁ and T₂). The values within the matrix (Mij, where i, j =1,2,…,n) represent the areas that have experienced changes among the different land use types during both periods. Based on previous studies [88,89]. ”– page number 4 , paragraph 13, and line 312-316. ]
M= (5)
The equation has been corrected, as has the text; however, the equation does not appear in the updated manuscript because it is too long. Instead, we have chosen to cite the methodology documented in previous studies.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 7: [7.Table 6, some equations have typos and mistakes] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 7 : We appreciate your corrections, The errors identified in Table 6 have been carefully reviewed and corrected accordingly.
[Text to be corrected: “Table 6. Landscape Pattern Indices and Their Meaning Used in the Study.
[updated text in the manuscript: Table A1. Landscape Pattern Indices and Their Meaning Used in the Study
Table 6 has been corrected and will be included as Appendix (Table A1). – page number 9, and line 335.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 8: [Table 7, the table title should be more specific] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 8: We agree with your suggestion, and it has been modified as follows:
[Text to be corrected: Table 7. Classification accuracy – page number 18, and line 446-447.] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 9: [Figure 8, what does each color represent?] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 9: We sincerely apologize for the oversight in not including the legend in Figure 8 (Chord Diagram for 5- and 10-Year Intervals [hectares]). The figure has now been updated to include the appropriate land cover class conventions, as follows: Red: Building Bright Green: Coffee Yellow-Orange: Bareland Dark Green: Forest Light Purple: Grasslands Cyan: Bamboo Blue: Water We appreciate your careful review and thank you for bringing this to our attention. – page number 20] Figure 9. Chord diagrams illustrating Land Use Land Cover (LULC) transition flows (in hectares) for different time intervals within the CCLC: (a) 2014-2019, (b) 2019-2024, (c) 2024-2034, and (d) the cumulative period 2014-2034 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Comments 10: [Table 10, for the transition matrix, which data from 2014 and 2019 are not clear enough?] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Response 10: We appreciate your valuable comments. The data presented in Table 10 have been verified and corrected. Climate variables were added to the forecast, giving us the opportunity to verify all the data. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe submitted paper presents a spatiotemporal assessment of land use and land cover (LULC) changes across the Colombian Coffee Cultural Landscape (CCLC) as relevant and highly dynamic countryside combining multi-sensor remote sensing datasets and hybrid modeling to forecast prospective LULC changes. Although the methodological framework is sound, some important aspects should be addressed to enhance the validity and clarity of the findings.
Potential Biases and Methodological Considerations
Omission of Climatic Variables (Page 35, Limitations Section): The lack of precipitation and temperature data may bias the conclusions presented in the article, as they directly impact agricultural productivity and forest dynamics. While the authors do acknowledge this limitation, they do not consider how climatic variability could affect LULC projections. Drought or extreme rainfall events, for example, may hasten grassland degradation or losses of coffee crops. A sensitivity analysis based on climatic scenarios would improve the robustness of the model.
Buffer Zone Neglect (Page 35): The study's emphasis is on the CCLC and its buffer zone is neglected. This restricts knowledge of spillover effects such as urban spatial growth pushing agricultural activities into edge regions. Future assessment with buffer would also give us a holistic view of landscape fragmentation in our landscapes.
Although train/validation samples do not show class imbalance; validation samples themselves may have class imbalance (Page 10, Table 3): Weakly supervised detection models are known to struggle with class imbalance due to the extent of minority classes (water bodies can be ~0.3–1.57% coverage) in search region. Example: In general accuracy is good (87.88%) but it is essential to report class-specific metrics (e.g. producer accuracy of water) in order to make sure that minority classes are not misclassfied
Results Interpretation and Projections
Urban Growth Estimations (Page 19–20): The estimated 16.95% growth to urbanization by 2034 seems farfetched. The model allows for linear trends, but does not consider potential policy interventions or conservation efforts (e.g., UNESCO-mandated protections). Discussing how regulatory frameworks can curb this trend would have provided some context for the projections.
Forest Loss Dynamics (page 20): Reporting 80.40% forest loss by 2034 contradicts the temporary forest area increase between 2019 and 2024. The authors need to clarify whether this reported decline results from direct deforestation, or edge effects from fragmentation, which, as the authors state, do not equal wholesale removal of forest.
Bamboo Expansion (Page 20): The 64.48% increase in bamboo cover is depicted as a secondary trend, but the detrimental ecological context is missing. Is this an expansion natural or anthropogenic? The role of bamboo in erosion control and its invasive nature should be covered to glean its impact on biodiversity.
Research Bias
Anthropocentric Focus (Page 34–35, Discussion): The study emphasizes anthropogenic pressures while ignoring adaptative strategies taken by local communities. One example is even traditional agroforestry practices in coffee cultivation could reduce fragmentation. Incorporating socio-cultural resilience into the discourse would redress the balance.
Main Broad Hypothesis and Negative Impact Emphasis (Page 36, Conclusions): Urbanization and forest loss are presented as threats, but the potential socio-economic benefits of urbanization (improved infrastructure) are ignored. All the talk about trade-offs seems conspicuous by its absence.
Grammar and Language Edits
Page 17, Table 8 Caption: “Coffee crops represents” → “Coffee crops represent” (subject-verb agreement).
Page 20, Line 5: “corresponded to an additional increase” → “corresponding to an additional increase” (tense consistency).
Page 35 Limitations Section: “climatic variables—precipitation and temperature—that influence agricultural productivity and the natural dynamics of the landscape were omitted” → “climatic variables (e.g., precipitation and temperature), which influence agricultural productivity and natural landscape dynamics, were omitted” (improved phrasing).
Page 36, Conclusions: “driven by territorial transformation dynamics and anthropogenic pressures” → “driven by territorial transformation and anthropogenic pressures” (removal of redundancy).
Recommendations for Revision
Add Climatic Variables Identify Climatic Variables To include climatic data as a secondary sensitivity analysis (Pg. 35).
Clarify the Mechanisms of Forest Loss: Separate the effects of direct deforestation from fragmentation (Page 20).
Consider Socio-Economic Context: Help balance your discussion by recognizing also the potential positive impacts of urbanization and community adaptations (Page 34–36).
Report Class-Specific Accuracy Metrics: Report for minority classes user/producer accuracy (Page 10–11)
Refine Projected Assumptions: Explain how assumptions around policy interventions can change urban growth pathways (Page 19–20)
Author Response
|
We sincerely thank you for the time and dedication devoted to reviewing our manuscript, as well as for the thoughtful and constructive feedback, which has greatly enhanced the clarity, scientific rigor, and contextual depth of our work. Your observations led to substantial improvements in both content and structure. In particular, we incorporated climatic variables (precipitation, temperature, NDVI) into the predictive model, clarified the rationale for focusing on the core zone of the CCCL, and revised the urban growth projections to better reflect governance realities under a Business as Usual (BAU) scenario. We also addressed ecological questions such as forest loss dynamics and bamboo expansion, integrated discussions on socio-cultural resilience, and balanced the conclusions by acknowledging both the risks and potential benefits of urban development. All changes are clearly highlighted in the revised version and demonstrate our commitment to addressing each of your comments with the care and attention they deserve. We are grateful for your valuable insights, which have undoubtedly strengthened the overall quality and relevance of this research. . |
||
|
2. Questions for General Evaluation |
Reviewer’s Evaluation |
Response and Revisions |
|
Does the introduction provide sufficient background and include all relevant references? |
Yes/Can be improved/Must be improved/Not applicable |
[Please give your response if necessary. Or you can also give your corresponding response in the point-by-point response letter. The same as below] |
|
Are all the cited references relevant to the research? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Is the research design appropriate? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the methods adequately described? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the results clearly presented? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the conclusions supported by the results? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
|
||
|
Comments 1: [Omission of Climatic Variables (Page 35, Limitations Section): The lack of precipitation and temperature data may bias the conclusions presented in the article, as they directly impact agricultural productivity and forest dynamics. While the authors do acknowledge this limitation, they do not consider how climatic variability could affect LULC projections. Drought or extreme rainfall events, for example, may hasten grassland degradation or losses of coffee crops. A sensitivity analysis based on climatic scenarios would improve the robustness of the model. |
||
|
Response 1 : We appreciate your suggestions and observations. We regret the initial omission of relevant climatic variables. In response to your comment regarding the possible bias in the conclusions due to the lack of precipitation and temperature data—variables that directly impact agricultural productivity and forest dynamics—the precipitation, temperature, and NDVI variables have been incorporated, which has improved the quality and accuracy of the predictions presented. Nevertheless, we have maintained a Business as Usual (BAU) scenario, which assumes continuity in current land-use trends.
[Text to be corrected: “To generate the LULC maps, a multi-sensor database was created for a 10-year period (2014–2024). Image collections available in the United States Geological Survey (USGS) catalog on Google Earth Engine (GEE) (https://developers.google.com/earth-engine/datasets) were utilized. Data processing was conducted on the GEE platform (https://earthengine.google.com/) using cloud computing. For the prediction of LULC for 2024–2034, variables such as elevation, slope, aspect, population, distance to roads, and distance to rivers were included. As shown in Table 1, population data were obtained from the WorldPop Global Project (https://www.worldpop.org/), distance to roads was derived from OpenStreetMap (https://www.openstreetmap.org), and distance to rivers was obtained from the New Global Hydrography derived from spaceborne elevation data (https://developers.google.com/earth-engine/datasets/catalog/USGS_SRTMGL1_003?hl=es-419). Elevation and slope were obtained from the 30 m resolution digital elevation model (DEM) of the Shuttle Radar Topography Mission (SRTM). All these factors were resampled to a 30 m resolution.” – page number 5 , paragraph 1, and line 152-166. – page number 6 , Table 1. ] [Updated text in the manuscript: To generate the LULC maps, a multi-sensor database was compiled for the 2014–2024 period using image collections from the Google Earth Engine (GEE) catalog, managed by the United States Geological Survey (USGS). Data processing was performed entirely on the GEE platform using cloud-based computing. For the prediction of LULC from 2024 to 2034, additional variables were included: elevation, slope, aspect, population, distance to roads, and distance to rivers. Population data were obtained from the WorldPop Global Project, road networks from OpenStreetMap, and river networks from the New Global Hydrography dataset. Elevation and slope were derived from the 30-meter resolution Digital Elevation Model of the Shuttle Radar Topography Mission (SRTM). Temperature data were extracted from the ERA5-Land Daily Aggregated product, precipitation data from the CHIRPS Daily dataset, and NDVI values from the Landsat 8 NDVI 8-Day Composite. All environmental variables were sourced directly from GEE, ensuring consistency in spatial resolution (30 m) and temporal coverage across the study period.
The temperature and climate variables are also added in Table 2 Flowchart of the methodology. – page number 6 , Table 4. line 139-156]
|
||
|
Comments 2: [Buffer Zone Neglect (Page 35): The study's emphasis is on the CCLC and its buffer zone is neglected. This restricts knowledge of spillover effects such as urban spatial growth pushing agricultural activities into edge regions. Future assessment with buffer would also give us a holistic view of landscape fragmentation in our landscapes.] |
||
|
Response 2 : We deeply appreciate the attention given to the buffer zone. Although we share the same interest in studying buffer areas, we have decided to begin our analysis with the core area and to address the buffer zone in future investigations, where the evolution of land-use dynamics and their impacts on landscape patterns will be examined in greater detail. Consequently, the current study focuses exclusively on the core area of the CCCL, where the attributes of Outstanding Universal Value (OUV) that underpinned its designation as a heritage site are located. This decision is primarily based on conservation considerations, as land-use changes in the core zone have a direct impact on the authenticity and integrity of the cultural landscape. Secondly, from a methodological perspective, the core zone represents a more coherent and well-defined unit of analysis, which allows for the application of spatial metrics and predictive models with greater precision.
We sincerely appreciate your valuable suggestions, which have helped to specify and justify the decision to focus this study solely on the core area. Below is the modified text for the study area section. [Updated text in the manuscript: “ The CCCL is divided into two zones. The primary zone corresponds to the core area, which spans 141,120 hectares—of which 140,046 hectares are rural and 1,074 hectares urban—distributed across 47 municipalities, 411 rural settlements (veredas), and 14 urban centers. The buffer zone covers 207,000 hectares, comprising 204,542 hectares of rural land and 2,458 hectares of urban land, distributed across 51 municipalities, 447 veredas, and 17 urban centers. This study focuses on the core area of the CCCL, where the attributes of Outstanding Universal Value (OUV) that justified its designation as a World Heritage Site are located [33]. According to UNESCO guidelines, the boundary of the inscribed property must ensure that all attributes contributing to its OUV are contained within its limits, while the buffer zone plays a complementary role in protection and management but is not considered part of the designated property [34,35]. The decision to focus on the core area is based primarily on conservation considerations: it is within this zone that land use changes have a direct impact on the authenticity and integrity of the cultural landscape. Secondly, from a methodological perspective, the core area offers a more coherent and clearly defined unit of analysis, in which the application of spatial metrics and predictive modeling can be conducted with greater accuracy [31]. ” – page number 3 y 4 , and line 119- 136. ] |
||
|
Comments 3: Although train/validation samples do not show class imbalance; validation samples themselves may have class imbalance (Page 10, Table 3): Weakly supervised detection models are known to struggle with class imbalance due to the extent of minority classes (water bodies can be ~0.3-1.57% coverage) in search region. Example: In general accuracy is good (87.88%) but it is essential to report class-specific metrics (e.g. producer accuracy of water) in order to make sure that minority classes are not misclassfied Response 3: We fully agree that relying solely on Overall Accuracy (OA) may obscure the model’s behavior toward underrepresented classes. To address this, we calculated class-specific metrics including Producer's Accuracy (PA / Recall) and User's Accuracy (UA / Precision)—for all LULC classes across the validation periods: [updated text in the manuscript: “Supplementary Table Included: A detailed table summarizing PA and UA for all LULC classes 2014, 2019, and 2024 has been added as Table S5 in the Supplementary Material.”–] page number 9, paragraph 2, and line 229-236. ]
|
||
|
Comments 4: [Urban Growth Estimations (Page 19-20): The estimated 16.95% growth to urbanization by 2034 seems farfetched. The model allows for linear trends, but does not consider potential policy interventions or conservation efforts (e.g., UNESCO-mandated protections). Discussing how regulatory frameworks can curb this trend would have provided some context for the projections.] |
||
|
Response 4: We sincerely appreciate your valuable observation regarding the projected magnitude of urban growth in the CCCL. Your comment enabled us to refine and better contextualize our findings. Following the proposed model adjustments and the inclusion of climatic variables, the projected urban coverage for the 2014–2034 period was revised from an initial 1.48% to 15.64%. Although this figure is slightly lower than previously reported, it still represents a significant increase, which validates your concern regarding the scale of this transformation. Regarding the consideration of political or legislative restrictions, such as those linked to the UNESCO designation, we opted not to incorporate them explicitly into the predictive model. This decision was based on conditions observed in the field: much of the recent land transformation appears to result from widespread unawareness of planning regulations and limited enforcement capacity by local authorities. Although land-use planning instruments do exist, their effective implementation and local awareness are minimal. Therefore, we considered that a business-as-usual scenario—reflecting this de facto dynamic of limited practical constraints—more accurately represents the current trajectory and associated risks in the absence of improved governance. To explicitly address this issue and clarify the scope of our results, the following modifications were made in the manuscript: In the Discussion section, we added the following paragraph, intended to emphasize that our results should not be interpreted as inevitable outcomes, but as potential scenarios that underscore the urgent need for proactive land-use management. [Updated text in the manuscript: “Similar trajectories are observed in other traditional landscapes recognized by UNESCO [116,120], where urbanization, shifts in agricultural practices, and habitat loss represent common challenges. In this context, the future of the CCCL will depend on the capacity of its stakeholders to manage land use conflicts and strengthen sustainable practices. While the 2034 projection warns of potentially undesirable trajectories, it also helps identify priority areas for intervention. ”– page number 30, paragraph 2, and line 675-680. ] In the Limitations section (4.5), we included the following clarification regarding the scope of the model and the rationale for not simulating alternative policy-based scenarios: [Updated text in the manuscript: “This study provides relevant findings; however, it is essential to acknowledge its limitations and suggest potential avenues for future research. First, the predictive model is based on a business-as-usual scenario that does not account for the implementation of protection policies promoted by UNESCO [121]. This choice is justified by the limited application of land-use planning instruments in the municipalities analyzed and by gaps in the enforcement of existing regulations [114,122]. As a result, the model does not include alternative scenarios for urban containment. ”– page number 30 , paragraph 3, and line 683-689. ] |
||
|
Comments 5: [Forest Loss Dynamics (page 20): Reporting 80.40% forest loss by 2034 contradicts the temporary forest area increase between 2019 and 2024. The authors need to clarify whether this reported decline results from direct deforestation, or edge effects from fragmentation, which, as the authors state, do not equal wholesale removal of forest.] |
||
|
Response 5: We thank the editor for highlighting the need to clarify the dynamics of forest cover change. In response, we have refined our simulation model and updated the manuscript to address this important issue. The revised version clarifies the calculation of percentage change, provides an explanation for the non-linear temporal trend (including a potential COVID-19-related effect), and explicitly distinguishes the nature of the projected forest loss.
Part 1: Clarification of the Projected Forest Loss Percentage [Text to be corrected: “Forest cover displayed a non-linear decreasing trend, declining from 6,999.22 ha in 2014 to 6,537.58 ha in 2019—a decrease of 6.60%—then experiencing a temporary increase to 7,598.64 ha in 2024—an increase of 16.23%. However, the projection for 2034 indicates a drastic reduction to 1,489.63 ha, representing a loss of 80.40% relative to 2024. ” – page number 20 , paragraph 2, and line 40-41.] [Updated text in the manuscript: “The findings indicate an increasingly anthropized scenario, characterized by rapid urban expansion from 1.48% in 2014 to 15.64% in 2034, a projected 77.8% decline in forest cover by 2034, and a gradual reduction of coffee cultivation by 1% over two decades. ”– page number 29 , paragraph 1, and line 607-610. ] Part 2: Addition of Explanatory Text in the Discussion We added a brief explanation of the possible causes behind the temporary forest increase and clarified that the projected loss is mainly caused by direct deforestation, not just fragmentation.
[Updated text in the manuscript: “ This trend aligns with critical deforestation patterns reported in other South American landscapes [118], underscoring the intense anthropogenic pressure on both ecological and cultural values within the CCLC. Notably, forest cover experienced a temporary increase between 2019 and 2024, which may be partially attributed to reduced anthropogenic activity during the COVID-19 pandemic, likely slowing land conversion processes. However, this short-term recovery appears to be an anomaly, as model projections indicate a substantial reversal: a 79.55% relative loss of forest cover between 2024 and 2034. This anticipated decline emphasizes the continued influence of urban expansion and confirms that the reduction in forest extent—amounting to an overall 77.80% loss since 2014—is predominantly driven by direct land-use conversion (deforestation), rather than by fragmentation processes alone, as discussed in Section 3.4.”– page number 29 , paragraph 2, and line 626-641. ]
|
||
|
Comments 6: [Bamboo Expansion (Page 20): The 64.48% increase in bamboo cover is depicted as a secondary trend, but the detrimental ecological context is missing. Is this an expansion natural or anthropogenic? The role of bamboo in erosion control and its invasive nature should be covered to glean its impact on biodiversity.] |
||
|
Response 6: We appreciate your observation regarding the need to provide ecological context for the trends observed in bamboo cover. Upon revisiting this point, we noted that our most recent simulation—now reflected in Table 9, presents a different dynamic for bamboo compared to the version of the manuscript you previously reviewed [Text to be corrected: “Secondary covers also exhibited significant changes. Bamboo showed an expansion trend, increasing from 8787.77 ha in 2014 to 14454.67 ha in 2034, an increase of 64.48%.” – page number 19, paragraph 5, and line 501-506.]
[updated text in the manuscript:– page number 17. ] |
||
|
Comments 7: [Anthropocentric Focus (Page 34-35, Discussion): The study emphasizes anthropogenic pressures while ignoring adaptative strategies taken by local communities. One example is even traditional agroforestry practices in coffee cultivation could reduce fragmentation. Incorporating socio-cultural resilience into the discourse would redress the balance.] |
||
|
Response 7: We sincerely thank the reviewer for the comment regarding the need to balance the discussion by incorporating the adaptive strategies and socio-cultural resilience of the local communities within the CCLC. To address this, we have restructured the Discussion section of the manuscript. Specifically, we have created a new dedicated subsection 4.3 Persistence of the Coffee Landscape and Resilience Processes and revised other parts of the discussion to explicitly incorporate these crucial aspects. [Updated text in the manuscript: “4.3 Persistence of the Coffee Landscape and Resilience Processes The coexistence of transformation processes and adaptation strategies confirms the active character of the CCCL as a living landscape. Local initiatives such as shade-grown coffee, specialty coffee production, rural tourism, and community organization—developed in municipalities such as Filandia, Salento, and Salamina [114] enhance the response to land use change. Although these actions are not directly captured through remote sensing results, they contribute to maintaining the ecological and functional heterogeneity of the territory. Similar trajectories are observed in other traditional landscapes recognized by UNESCO [116,120], where urbanization, shifts in agricultural practices, and habitat loss represent common challenges. In this context, the future of the CCCL will depend on the capacity of its stakeholders to manage land use conflicts and strengthen sustainable practices. While the 2034 projection warns of potentially undesirable trajectories, it also helps identify priority areas for intervention.” – page number 30, paragraph 1, and line 663-680.]
|
||
|
Comments 8: [Main Broad Hypothesis and Negative Impact Emphasis (Page 36, Conclusions): Urbanization and forest loss are presented as threats, but the potential socio-economic benefits of urbanization (improved infrastructure) are ignored. All the talk about trade-offs seems conspicuous by its absence.] |
||
|
Response 8: We sincerely thank the reviewer for comment regarding the emphasis and balance in the Conclusions section. [Text to be corrected: “This study integrated multisensor remote sensing, hybrid predictive CA-ANN models, and a detailed analysis of landscape metrics to evaluate the evolution of LULC and the impacts of landscape patterns in the CCLC during the period 2014–2034. [...] providing scientific bases to help mitigate the identified negative impacts while promoting ecological and socio-cultural resilience in this valuable heritage region. ” – page number 36 , paragraph 1, and line 824-859.] [Updated text in the manuscript: “This study integrated multisensor remote sensing, hybrid predictive CA-ANN models, and a detailed analysis of landscape metrics to evaluate the evolution of LULC and the impacts of landscape patterns in the Coffee Cultural Landscape of Colombia (CCLC) during the period 2014–2034. The results demonstrate significant changes in landscape structure driven by territorial transformation dynamics and anthropogenic pressures. The main conclusions are as follows: |
||
|
Comments 9: [Grammar and Language Edits Page 17, Table 8 Caption: “Coffee crops represents” → “Coffee crops represent” (subject-verb agreement). Page 20, Line 5: “corresponded to an additional increase” → “corresponding to an additional increase” (tense consistency). Page 35 Limitations Section: “climatic variables—precipitation and temperature—that influence agricultural productivity and the natural dynamics of the landscape were omitted” → “climatic variables (e.g., precipitation and temperature), which influence agricultural productivity and natural landscape dynamics, were omitted” (improved phrasing). Page 36, Conclusions: “driven by territorial transformation dynamics and anthropogenic pressures” → “driven by territorial transformation and anthropogenic pressures” (removal of redundancy).]
|
||
|
Response 9: Due to substantial revisions and the integration of new analyses, the previously indicated sentences on pages 17, 20, and 35 have been removed or replaced entirely. Consequently, those grammatical issues no longer appear in the current version of the manuscript. [Text to be corrected: “driven by territorial transformation dynamics and anthropogenic pressures” – page number 36, paragraph 1, and line 832.] [Updated text in the manuscript: “driven by territorial transformation and anthropogenic pressures”– page number 31, paragraph 1, and line 721-722.]
|
||
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsI would like to congratulate the authors on the improvements made to the manuscript and the effort invested. I believe the article is now ready for publication. However, I recommend that the authors include clear references within the text to what can be found in the appendix and the supplementary materials. Additionally, they should ensure that these documents contain the correct data and information.
Author Response
For research article
|
Response to Reviewer 1 Comments
|
||
|
|
|
|
|
We are grateful for the opportunity to contribute to Land and deeply appreciate the thoughtful and rigorous feedback received during the review process. Your comments were essential in strengthening the manuscript’s organization and clarity. In particular, we revised and updated all references to the appendices and supplementary materials, ensuring that tables and figures are accurately cited and fully consistent with the corresponding sections of the main text.
|
||
|
2. Questions for General Evaluation |
Reviewer’s Evaluation |
Response and Revisions |
|
Does the introduction provide sufficient background and include all relevant references? |
Yes/Can be improved/Must be improved/Not applicable |
[Please give your response if necessary. Or you can also give your corresponding response in the point-by-point response letter. The same as below] |
|
Are all the cited references relevant to the research? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Is the research design appropriate? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the methods adequately described? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the results clearly presented? |
Yes/Can be improved/Must be improved/Not applicable
|
|
|
3. Point-by-point response to Comments and Suggestions for Authors
|
||
|
We are pleased to know that our manuscript is in the final stages of preparation for publication. It is an honor for us to have the opportunity to publish in Land Journal, and we deeply appreciate the contributions made during the review process. The feedback received was meticulous, consistent, and extremely valuable, significantly enhancing the quality of our manuscript.
Comments 1: [“I believe the article is now ready for publication. However, I recommend that the authors include clear references within the text to what can be found in the appendix and the supplementary materials. Additionally, they should ensure that these documents contain the correct data and information.”]
Response 26: We are pleased to know that our manuscript is in the final stages of preparation for publication. It is an honor for us to have the opportunity to publish in Land Journal, and we deeply appreciate the contributions made during the review process. The feedback received was meticulous, consistent, and extremely valuable, significantly enhancing the quality of our manuscript.
For a detailed description of the formulas and definitions, please refer to the Appendix A (Table A1). – page number 12, paragraph 1 and line 340-341. The full formulas and references are provided in (Table S1). – page number 8, paragraph 2 and line 202-203. 3.3.1 Land Use Transfer Analysis The transition matrix enabled the quantification of conversions among Land Use and Land Cover (LULC) categories for the periods 2014–2019, 2019–2024, 2024–2034, and the cumulative period 2014–2034 (Table S3). Chord diagrams were employed for visualization (Figure 8), where the thickness of the links represents the magnitude of transitions) – page number 19, paragraph 2 and line 444-446.
(It has been verified that all information related to the data, figures, as well as every comma and period, is consistent between the supplementary materials and the manuscript text, thus ensuring clarity and coherence.)
[Updated text in the manuscript
The described procedure was implemented in a script on Google Earth Engine (GEE), the details of which are presented in Appendix A. – page number 10, paragraph 1 and line 337-338. The described procedure was implemented using a script in Google Earth Engine (GEE), available at: https://code.earthengine.google.com/be158215b47827977de85525784ae34f – page number 33, paragraph 1 and line 787-788.
Indices and their meaning of landscape pattern used in this study are compiled in Appendix B (Table B1). – page number 14, paragraph 1 and line 346-347.
Appendix B
[The spectral indices, formulas, and references used in the classification are listed in Table S1. – page number 9, paragraph 2 and line 205-206.
Table S1. Spectral Indices. – page number 1, supplementary material]
[The Kappa coefficients reflected strong agreement in the classification, with values exceeding 0.80 throughout the period, reaching a maximum of 0.91 in 2014 and a minimum of 0.81 in 2019, with an average of 0.84, For a more detailed understanding of the class-specific validation metrics, Table S2 provides a summary of the Producer’s Accuracy (PA) and User’s Accuracy (UA) for all LULC classes in 2014, 2019, and 2024. – page number 16-17, paragraph 1 and line 369-373.
Table S3. Area (ha) and percentage (%) of each LULC class during the period 2014–2024. – page number 2,3, supplementary material] [3.3 LULC Changes 2014–2034 The analysis of LULC dynamics between 2014 and 2034 reveals a continuous process of territorial transformation, characterized by urban expansion, the loss of natural areas, and the conversion of agricultural land cover. Figure 7 displays the LULC change maps corresponding to the periods 2014–2019, 2019–2024, and 2024–2034, respectively Table S4 details the analysis of changes (in hectares) between categories for each interval. – page number 18, paragraph 2 and line 403-408.
3.3.1 Land Use Transfer Analysis The transition matrix enabled the quantification of conversions among Land Use and Land Cover (LULC) categories for the periods 2014–2019, 2019–2024, 2024–2034, and the cumulative period 2014–2034. Detailed counts for each transition are provided in Table S4 and are visualized in Figure 9 as chord diagrams, where the thickness of the links represents the magnitude of transitions. – page number 20, paragraph 2 and line 451-455.]
|
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThank you for providing the reviewed version of the manuscript. Here I am sharing some minor comments:
Spell out all acronyms (like CA-ANN) on first reference. NA: The first time an abbreviation appears in the text (including the Abstract), please ensure it is that abbreviation's full term followed by the abbreviation itself (e.g. Cellular Automata-Artificial Neural Network (CA-ANN)).
Use the full phrase at least once in the abstract or introduction, as not all readers will know its abbreviation (LULC).
Improve flow by adding a short sentence which potentially links global examples (e.g., Sagarmatha National Park) to the Colombian Coffee Cultural Landscape (CCLC). For example, each OUV includes selected pressuring factors listed in parentheses: “Similar pressures are evident in the CCLC, a unesco site under threat from urbanization and agricultural expansion.”
Explain why 30m population data (WorldPop) will be used in tandem with 30m Landsat/Sentinel data. Discuss potential limitations of this resolution for modeling localized urban expansion.
Explain selection of core zone (vs. buffer zone) for analysis (in Section 2.1) To maintain coherence, repeat this rationale briefly — in the limitations paragraph.
Comment on the smaller Kappa coefficient (0.81) found over the area in 2019 Reasons why you believe there are missing, omitted or otherwise excluded data (i.e., due to cloud cover, temporal anomalies and the like) should be documented briefly for transparency.
Discuss the mentioned “trade-offs” in the conclusion (i.e., socio-economic benefits of urbanization) in the discussion section so that the conclusions are cross-referenced to earlier analysis.
In the limitations, you could discuss how UNESCO policies help to integrate into future models (i.e., scenario modeling with policy constraints).
References to use consistent italicisation for journal names (E.g. “Remote Sens. Environ." vs."Landscape and Urban Planning").
Verify formatting for in-text citations (e.g., commas/semicolons between multiple references: [1,2] vs. [1;2] according to journal specifics.
Figures and Tables
Double-check that all figure captions (e.g., Figure 1) correspond to the content. It does not refer to zones as A–F in figure 1B, as the caption indicates, and not to “six distinct landscape zones,” as mentioned in the text. Ensure consistency.
Add and cite relevant supplemental tables (e.g., Table S1) in the appendix or supplemental files.
Check if references are all formatted the same (e.g. some have full first names, some have initials). Follow MDPI standards for quality assurance.
Author Response
For research article
|
Response to Reviewer 3 Comments |
||
|
|
|
|
|
We are grateful for your careful review and deep expertise. The suggestions provided have significantly enhanced the clarity and scientific robustness of the manuscript. In response, we conducted thorough revisions across all sections: we defined the key abbreviations (CA‑ANN, LULC), strengthened the links between global and local cases, and justified our methodological choices (30 m resolution, focus on the core zone). We also expanded the discussion of model limitations and the balance between urban growth and landscape conservation. References to supplementary tables (S1‑S5) were verified, figure captions were refined, and formatting was corrected throughout the document.
|
||
|
2. Questions for General Evaluation |
Reviewer’s Evaluation |
Response and Revisions |
|
Does the introduction provide sufficient background and include all relevant references? |
Yes/Can be improved/Must be improved/Not applicable |
[Please give your response if necessary. Or you can also give your corresponding response in the point-by-point response letter. The same as below] |
|
Are all the cited references relevant to the research? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Is the research design appropriate? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the methods adequately described? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the results clearly presented? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
Are the conclusions supported by the results? |
Yes/Can be improved/Must be improved/Not applicable |
|
|
3. Point-by-point response to Comments and Suggestions for Authors
|
||
|
Comments 1: [Spell out all acronyms (like CA-ANN) on first reference. NA: The first time an abbreviation appears in the text (including the Abstract), please ensure it is that abbreviation's full term followed by the abbreviation itself (e.g. Cellular Automata-Artificial Neural Network (CA-ANN)).] |
||
|
Response 1 : Thank you for your observation. As recommended, we have revised the abstract to ensure that the full term Cellular Automata–Artificial Neural Network (CA-ANN) is spelled out upon its first mention (Abstract, line 17). This now reads: “...using a hybrid Cellular Automata–Artificial Neural Network (CA-ANN) model...”.
[Text to be corrected: “Future projections were conducted using a hybrid Cellular Automata and Artificial Neural Network model, ” – page number 1 , paragraph 1, and line 17-18.]
[Updated text in the manuscript: Future projections were conducted using a hybrid Cellular Automata and Artificial Neural Network model (CA-ANN– page number 1 , paragraph 1, and line 18-19] |
||
|
|
||
|
Comments 2: [Use the full phrase at least once in the abstract or introduction, as not all readers will know its abbreviation (LULC).]
[Updated text in the manuscript: changes in land use and land cover (LULC)” – page number 1, paragraph 1, and line 10.]]
|
||
|
Comments 3: [Improve flow by adding a short sentence which potentially links global examples (e.g., Sagarmatha National Park) to the Colombian Coffee Cultural Landscape (CCLC). For example, each OUV includes selected pressuring factors listed in parentheses: “Similar pressures are evident in the CCLC, a UNESCO site under threat from urbanization and agricultural expansion.]
[Updated text in the manuscript: Similar pressures, notably from urbanization and agricultural expansion, also threaten the integrity of the Colombian Coffee Cultural Landscape (CCLC). – page number 2, paragraph 1, and line 49-51] |
||
|
Comments 4: [Explain why 30 m population data (WorldPop) will be used in tandem with 30 m Landsat/Sentinel data. Discuss potential limitations of this resolution for modeling localized urban expansion.]
(1) Justification of resolution use: We added a sentence in Section 2.3.2.1 (line 250) to explain that the 30 m spatial resolution of WorldPop was selected to ensure consistency with the Landsat and Sentinel datasets, allowing seamless integration of population density with remote sensing layers. (2) Discussion of limitations: We included an additional sentence in Section 4.4 (line 724) under the discussion of study limitations, highlighting that while the resolution enables large-scale analysis, it may limit the detection of fine-scale urban expansion, particularly in dense or informal settlement areas. These additions improve the transparency of our methodology and acknowledge its constraints.
[Updated text in the manuscript: The 30 m WorldPop population dataset was chosen to match the resolution of Landsat and Sentinel imagery, facilitating integration of demographic and environmental layers in the modeling process. – page number 10, paragraph 2, and line 450-452]
In addition, the 30 m spatial resolution may limit the detection of localized urban expansion, especially in areas characterized by small-scale or informal development. Subpixel urban changes may remain undetected, potentially affecting the accuracy of urban growth projections in densely built environments. – page number 31, paragraph 7, and line 724-727]
|
|
Comments 5: [Explain selection of core zone (vs. buffer zone) for analysis (in Section 2.1). To maintain coherence, repeat this rationale briefly — in the limitations paragraph.] In addition, we added a concise reiteration in the limitations section to maintain internal consistency. This brief explanation acknowledges that while the decision was methodologically sound, it also limits the assessment of peripheral effects outside the core boundary.
[[Text to be corrected: 2.1 Study Area “The decision to focus on the core area is based primarily on conservation considerations: it is within this zone that land use changes have a direct impact on the authenticity and integrity of the cultural landscape. Secondly, from a methodological perspective, the core area offers a more coherent and clearly defined unit of analysis, in which the application of spatial metrics and predictive modeling can be conducted with greater accuracy.” – page number 4, paragraph 2, and line 131 -136]
[Updated text in the manuscript: This study focuses exclusively on the core zone of the Colombian Coffee Cultural Landscape (CCLC), where the attributes that justify its inscription under the Outstanding Universal Value (OUV) criteria are officially located [33]. According to UNESCO guidelines, only the core zone is considered part of the designated World Heritage property, while the buffer zone serves a complementary role in protection and management [34,35]. From a conservation standpoint, land use changes within the core area have a direct impact on the authenticity and integrity of the landscape. Methodologically, the core zone provides a clearly defined and spatially coherent unit for analysis, enabling more accurate application of spatial metrics and predictive modeling. Including the buffer zone, which exhibits greater heterogeneity and lacks formal recognition as part of the inscribed site, could introduce analytical inconsistencies and dilute the interpretation of landscape dynamics related to the OUV. – page number 4, paragraph 2, and line 128 -129]
[[Text to be corrected: 4.4 Study Limitations and Future Research Directions
[Updated text in the manuscript: Second, the analysis was restricted to the core zone of the Colombian Coffee Cultural Landscape, following heritage integrity criteria defined by UNESCO [123]. This decision, while methodologically justified, also implies a trade-off. Focusing solely on the core zone ensured analytical precision and alignment with UNESCO’s definition of the inscribed property, but it limited the capacity to evaluate indirect transformations occurring in the buffer zone, such as fringe urbanization or landscape fragmentation beyond the core boundary
|
|
Comments 6: [Comment on the smaller Kappa coefficient (0.81) found over the area in 2019. Reasons why you believe there are missing, omitted or otherwise excluded data (i.e., due to cloud cover, temporal anomalies and the like) should be documented briefly for transparency.]
[[Text to be corrected: “The Kappa coefficients reflected strong agreement in the classification, with values exceeding 0.80 throughout the period, reaching a maximum of 0.91 in 2014 and a minimum of 0.81 in 2019, with an average of 0.84.” – page number 16, paragraph 1, and line 370 -372]
[Updated text in the manuscript: The Kappa coefficients reflected strong agreement in the classification, with values exceeding 0.80 throughout the period, reaching a maximum of 0.91 in 2014 and a minimum of 0.81 in 2019, with an average of 0.84. The slightly lower coefficient in 2019 is attributed to persistent cloud cover during that year, which reduced the availability of high-quality imagery and affected class separability in some transitional areas. Nonetheless, the result remains within acceptable accuracy thresholds – page number 16, paragraph 1, and line 370 -375]
|
|
Comments 7: [“Discuss the mentioned “trade-offs” in the conclusion (i.e., socio-economic benefits of urbanization) in the discussion section so that the conclusions are cross-referenced to earlier analysis."] While Section 4.3 (Persistence of the Coffee Landscape and Resilience Processes) briefly touched upon the potential social benefits of urban expansion , we acknowledge that explicitly framing this as a "trade-off" and elaborating slightly within the Discussion would strengthen the connection to the Conclusion.
[[Text to be corrected: “It is important to note that urban expansion does not necessarily entail degradation, as in some cases it improves access to infrastructure and services, generating social benefits. The coexistence of transformation processes and adaptation strategies confirms the active character of the CCLC as a living landscape.” – page number 31, paragraph 2,3 and line 697 -698]
[Updated text in the manuscript: Revised text for Section 4.3 (Lines 684-686 approx.): It is important to note that urban expansion does not necessarily entail degradation, as in some cases it improves access to infrastructure and services, generating social benefits. This dynamic presents inherent trade-offs, particularly within a valued cultural landscape where development pressures must be carefully weighed against heritage conservation. Effectively navigating these transformations requires acknowledging both potential socio-economic gains and the impacts on the landscape's integrity. The coexistence of transformation processes and adaptation strategies confirms the active character of the CCLC as a living landscape. – page number 31, paragraph 2, and line 694 -701]
In Section 5 (Conclusion)
[Updated text in the manuscript: recognizing these trade-offs is crucial for effective planning. Sustainable territorial governance must balance the preservation of ecosystems and cultural heritage (which underpin the CCLC's Outstanding Universal Value) with the development needs of its inhabitants. – page number 32, paragraph 7, and line 771 -774]
|
|
Comments 8: [In the limitations, you could discuss how UNESCO policies help to integrate into future models (i.e., scenario modeling with policy constraints).]
[[Text to be corrected: “This choice is justified by the limited application of land-use planning instruments in the municipalities analyzed and by gaps in the enforcement of existing regulations.” – page number 31 , paragraph 5, and line 718 -720]
[Updated text in the manuscript: This choice is justified by the limited application of land-use planning instruments in the municipalities analyzed and by gaps in the enforcement of existing regulations. However, future models could incorporate policy-based scenarios aligned with UNESCO guidelines, allowing simulation of land-use change under conservation-oriented constraints. – page number 31, paragraph 7, and line 718 -722]
|
|
Comments 9: [References to use consistent italicisation for journal names (E.g. “Remote Sens. Environ." vs. "Landscape and Urban Planning").]
|
|
Comments 10: [Verify formatting for in-text citations (e.g., commas/semicolons between multiple references: [1,2] vs. [1;2] according to journal specifics.]
|
|
Comments 11: [Figures and Tables
Double-check that all figure captions (e.g., Figure 1) correspond to the content. It does not refer to zones as A–F in figure 1B, as the caption indicates, and not to “six distinct landscape zones,” as mentioned in the text. Ensure consistency.] [[Text to be corrected: “The territory is composed of six distinct landscape zones (Figure1).” – page number 3 , paragraph 5, and line 111 -112].
[Updated text in the manuscript: The territory is composed of six distinct landscape zones—Zone A through Zone F, as shown in Figure 1 (b), – page number 3, paragraph 5, and line 111 -112]
Figure 1. Maps of the study area: (a) Location of the Colombian Coffee Cultural Landscape (CCLC) within Colombia and South America. (b) Core area of the CCLC, showing the internal division into six landscape zones (Zone A through Zone F), distributed across the departments of Caldas, Quindío, Risaralda, and Valle del Cauca. |
|
Comments 12: [Add and cite relevant supplemental tables (e.g., Table S1) in the appendix or supplemental files.]
[Updated text in the manuscript
The described procedure was implemented in a script on Google Earth Engine (GEE), the details of which are presented in Appendix A. – page number 10, paragraph 1 and line 337-338. Appendix A The described procedure was implemented using a script in Google Earth Engine (GEE), available at: https://code.earthengine.google.com/be158215b47827977de85525784ae34f – page number 33, paragraph 1 and line 787-788.
Indices and their meaning of landscape pattern used in this study are compiled in Appendix B (Table B1). – page number 14, paragraph 1 and line 346-347.
Appendix B Table B1 describes the indices used to assess fragmentation, connectivity, and landscape structure in the study. Each metric is presented with its equation, variables, and interpretive meaning) – page number 33, paragraph 2 and line 791-793.
[The spectral indices, formulas, and references used in the classification are listed in Table S1. – page number 9, paragraph 2 and line 205-206.
Table S1. Spectral Indices. – page number 1, supplementary material]
[The Kappa coefficients reflected strong agreement in the classification, with values exceeding 0.80 throughout the period, reaching a maximum of 0.91 in 2014 and a minimum of 0.81 in 2019, with an average of 0.84, For a more detailed understanding of the class-specific validation metrics, Table S2 provides a summary of the Producer’s Accuracy (PA) and User’s Accuracy (UA) for all LULC classes in 2014, 2019, and 2024. – page number 16-17, paragraph 1 and line 369-373.
Table S2. Class validation matrices. – page number 1,2 supplementary material]
[The quantitative analysis of spatiotemporal patterns, expressed in hectares (ha) and percentages (%), is presented in Figure 5 and Table S3, corresponding to each LULC class identified during the 2014-2024 period. – page number 17, paragraph 1 and line 377-379.
Table S3. Area (ha) and percentage (%) of each LULC class during the period 2014–2024. – page number 2,3, supplementary material]
[3.3 LULC Changes 2014–2034 The analysis of LULC dynamics between 2014 and 2034 reveals a continuous process of territorial transformation, characterized by urban expansion, the loss of natural areas, and the conversion of agricultural land cover. Figure 7 displays the LULC change maps corresponding to the periods 2014–2019, 2019–2024, and 2024–2034, respectively Table S4 details the analysis of changes (in hectares) between categories for each interval. – page number 18, paragraph 2 and line 403-408.
3.3.1 Land Use Transfer Analysis The transition matrix enabled the quantification of conversions among Land Use and Land Cover (LULC) categories for the periods 2014–2019, 2019–2024, 2024–2034, and the cumulative period 2014–2034. Detailed counts for each transition are provided in Table S4 and are visualized in Figure 9 as chord diagrams, where the thickness of the links represents the magnitude of transitions. – page number 20, paragraph 2 and line 451-455.]
[Table S4 Land Use Change Analysis and LULC Transitions (2014–2034). – page number 2-4, supplementary material]
[The analysis of the gains and losses in LULC is presented in Figure 8 and Table S5, where more detailed information can be found, showing the areas (hectares) for each land use and land cover class. – page number 19, paragraph 4 and line 431-433.
Table S5. LULC Gains and Losses (2014 -2034) – page number 5, supplementary material]
|
|
Comments 12: [Check if references are all formatted the same (e.g. some have full first names, some have initials). Follow MDPI standards for quality assurance.] |
|
|
|
|
|
|
|
|
|
|
|
|
Author Response File: Author Response.pdf
