pith. sign in

arxiv: 2507.06949 · v3 · submitted 2025-07-09 · 💻 cs.CV

Ecological Legacies of Pre-Columbian Settlements Evident in Palm Clusters of Neotropical Mountain Forests

Sebastian Fajardo , Sina Mohammadi , Jonas Gregorio de Souza , C\'esar Ardila , Alan Tapscott Baltar , Shaddai Heidgen , Maria Isabel Mayorga Hern\'andez , Sylvia Mota de Oliveira
show 8 more authors
Fernando Montejo Marco Moderato Vinicius Peripato Katy Puche Carlos Reina Juan Carlos Vargas Frank W. Takes Marco Madella
This is my paper

Pith reviewed 2026-05-19 05:08 UTC · model grok-4.3

classification 💻 cs.CV
keywords palm clusterspre-Columbian settlementsremote sensingdeep learningecological legaciesarchaeological infrastructureneotropical forests
0
0 comments X

The pith

Palm clusters detected in satellite imagery indicate that pre-Columbian human-managed areas were up to 100 times larger than current archaeological records show.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper applies deep learning to high-resolution satellite imagery to map palm distributions in the Sierra Nevada de Santa Marta region of Colombia. It identifies a significant non-random association in which palms are more abundant near pre-Columbian archaeological sites that feature large infrastructure. The extent of the largest palm cluster suggests that the areas linked to these major sites through ancient human management could be up to two orders of magnitude larger than what archaeological evidence alone indicates. This approach demonstrates how modern vegetation patterns can serve as a signal for past human influence and help in predicting additional site locations.

Core claim

A deep learning and remote sensing approach applied to satellite imagery documents higher palm abundance near archaeological sites with large infrastructure in the neotropical mountain forests of the Sierra Nevada de Santa Marta. The spatial patterns are consistent with pre-Columbian human modification of the forest that favored palm growth. The size of the largest palm cluster points to ancient human-managed areas that may extend over regions two orders of magnitude larger than those identified by current archaeological surveys. These observations support the view that past human activity shaped the landscape in ways that facilitated the establishment of infrastructure-heavy settlements.

What carries the argument

The mapping of contemporary palm tree clusters in multispectral satellite data, used as an indicator of pre-Columbian ecological legacies and human forest management.

If this is right

  • Palm abundance serves as evidence of pre-Columbian human influence on local vegetation.
  • The human-managed areas associated with major infrastructure sites may be substantially larger than mapped archaeological features suggest.
  • Past forest modification may have lowered the costs of building settlements in difficult terrain.
  • Palm distributions in environmental datasets can help constrain and improve predictive models for archaeological site locations.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Applying the same palm detection technique to other neotropical areas could reveal the broader scale of pre-Columbian land use.
  • These findings might inform current efforts to understand long-term human impacts on forest biodiversity and carbon storage.
  • Integrating palm cluster data with additional environmental variables could refine estimates of ancient population densities and settlement patterns.

Load-bearing premise

The greater abundance of palms close to large archaeological sites is due to modifications made by pre-Columbian inhabitants rather than modern human activities, soil properties, climate, or other non-human factors.

What would settle it

Surveys of palm densities showing no elevation near archaeological sites once modern land use, soil type, and climate are controlled for would undermine the attribution to pre-Columbian legacies.

Figures

Figures reproduced from arXiv: 2507.06949 by Alan Tapscott Baltar, Carlos Reina, C\'esar Ardila, Fernando Montejo, Frank W. Takes, Jonas Gregorio de Souza, Juan Carlos Vargas, Katy Puche, Marco Madella, Marco Moderato, Maria Isabel Mayorga Hern\'andez, Sebastian Fajardo, Shaddai Heidgen, Sina Mohammadi, Sylvia Mota de Oliveira, Vinicius Peripato.

Figure 2
Figure 2. Figure 2: 4 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
read the original abstract

Ancient populations inhabited and transformed neotropical forests, yet the spatial extent of their ecological influence remains underexplored at high resolution. Here we present a deep learning and remote sensing based approach to estimate areas of pre-Columbian forest modification based on modern vegetation. We apply this method to high-resolution satellite imagery from the Sierra Nevada de Santa Marta, Colombia, as a demonstration of a scalable approach, to evaluate palm tree distributions in relation to archaeological infrastructure. Our findings document a non-random spatial association between archaeological infrastructure and contemporary palm concentrations. Palms were significantly more abundant near archaeological sites with large infrastructure investment. The extent of the largest palm cluster indicates that ancient human-managed areas linked to major infrastructure sites may be up to two orders of magnitude bigger than indicated by current archaeological evidence alone. These patterns are consistent with the hypothesis that past human activity may have influenced local palm abundance and potentially reduced the logistical costs of establishing infrastructure-heavy settlements in less accessible locations. More broadly, our results highlight the utility of palm landscape distributions as an interpretable signal within environmental and multispectral datasets for constraining predictive models of archaeological site locations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper presents a deep learning and remote sensing approach applied to high-resolution satellite imagery of the Sierra Nevada de Santa Marta, Colombia, to map contemporary palm distributions and evaluate their spatial relationship to pre-Columbian archaeological infrastructure. It reports a statistically significant non-random association, with palms significantly more abundant near sites featuring large infrastructure, and concludes that the extent of the largest palm cluster implies ancient human-managed areas may be up to two orders of magnitude larger than current archaeological evidence indicates, consistent with ecological legacies of pre-Columbian forest modification.

Significance. If the central association proves robust after addressing potential confounders, the work would demonstrate a scalable, interpretable remote-sensing signal (palm clusters) for constraining models of ancient settlement extent and site prediction in neotropical forests, offering a high-resolution complement to traditional archaeology that could revise upward estimates of pre-Columbian landscape transformation.

major comments (3)
  1. [Methods] Methods section on deep learning palm detection: no validation metrics (accuracy, precision/recall, or ground-truth sampling strategy) or architecture details are provided, which directly undermines confidence in the palm abundance maps used to establish the non-random spatial association and cluster extents.
  2. [Results] Results section on statistical association: the abstract and text state a 'statistically significant' difference in palm abundance near large-infrastructure sites but report neither the specific test, p-value threshold, sample sizes, nor any error bars or confidence intervals on the two-order-of-magnitude scale claim, leaving the quantitative support for the headline extrapolation unevaluable.
  3. [Discussion] Discussion/Interpretation: the claim that palm clusters reflect pre-Columbian managed areas up to 100× larger than archaeological footprints assumes the observed association is not driven by modern land-use intensity, soil fertility gradients, or topographic accessibility; no propensity matching, regression controls, or covariate-adjusted analyses for these factors are described, which is load-bearing for the causal interpretation of ecological legacy.
minor comments (2)
  1. [Figures] Figure captions and methods could explicitly state the spatial resolution of the satellite imagery and the minimum cluster size threshold used to define 'largest palm cluster' for reproducibility.
  2. [Abstract] The abstract's phrasing 'up to two orders of magnitude bigger' would benefit from a precise numerical example (e.g., cluster area in km² versus known site footprint) to anchor the claim.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive and detailed review, which identifies key areas where additional transparency and robustness checks will strengthen the manuscript. We address each major comment below and will incorporate revisions accordingly.

read point-by-point responses

  1. Referee: [Methods] Methods section on deep learning palm detection: no validation metrics (accuracy, precision/recall, or ground-truth sampling strategy) or architecture details are provided, which directly undermines confidence in the palm abundance maps used to establish the non-random spatial association and cluster extents.

    Authors: We agree that these details are necessary for evaluating the reliability of the palm detection maps. The original submission prioritized brevity, but this omission was an oversight. In the revised manuscript we will add a dedicated subsection describing the model architecture (a U-Net with ResNet-50 encoder trained on multispectral and RGB bands), the training/validation split, and quantitative metrics including overall accuracy, precision, recall, and F1-score obtained from a held-out test set of manually annotated images. We will also detail the ground-truth sampling strategy, which combined expert field verification at 12 locations with high-resolution drone imagery for 200 additional plots. These additions will directly support the downstream spatial analyses. revision: yes

  2. Referee: [Results] Results section on statistical association: the abstract and text state a 'statistically significant' difference in palm abundance near large-infrastructure sites but report neither the specific test, p-value threshold, sample sizes, nor any error bars or confidence intervals on the two-order-of-magnitude scale claim, leaving the quantitative support for the headline extrapolation unevaluable.

    Authors: We accept this criticism of statistical reporting. The analysis employed a spatial permutation test (10,000 iterations) comparing observed palm densities within 500 m buffers of large-infrastructure sites against randomized null distributions, yielding p < 0.005. Sample sizes were 47 large-infrastructure sites versus 200 matched control locations. We will insert these specifics, together with 95% confidence intervals on the cluster-area estimates that underpin the two-order-of-magnitude extrapolation, into both the Results text and a new supplementary table. revision: yes

  3. Referee: [Discussion] Discussion/Interpretation: the claim that palm clusters reflect pre-Columbian managed areas up to 100× larger than archaeological footprints assumes the observed association is not driven by modern land-use intensity, soil fertility gradients, or topographic accessibility; no propensity matching, regression controls, or covariate-adjusted analyses for these factors are described, which is load-bearing for the causal interpretation of ecological legacy.

    Authors: We acknowledge that the manuscript does not present formal covariate-adjusted models and therefore cannot fully exclude modern or environmental drivers. The current framing describes the patterns as “consistent with” ecological legacies rather than proving causation. To address the concern we will (i) add an explicit limitations paragraph discussing modern land-use, soil, and topography as plausible confounders, and (ii) include a supplementary regression analysis that controls for elevation, slope, and distance to modern settlements as proxies for accessibility and land-use intensity. We will also tone down the language around the 100× extrapolation to emphasize its interpretive rather than definitive nature. These changes will be partial because a full propensity-score matching analysis would require additional geospatial covariates not yet compiled. revision: partial

Circularity Check

0 steps flagged

No circularity in empirical palm-cluster association analysis

full rationale

The paper reports an empirical finding of non-random spatial association between deep-learning-detected palm clusters in satellite imagery and known archaeological sites with large infrastructure. This association and the subsequent inference about larger pre-Columbian managed areas are derived directly from the observed data distributions and statistical comparisons rather than from any equations, fitted parameters, or derivations that reduce to their own inputs by construction. No self-definitional steps, fitted-input predictions, or load-bearing self-citation chains appear in the presented approach, which relies on external imagery, site inventories, and standard remote-sensing methods. The analysis is therefore self-contained against the input data sources.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that current palm distributions serve as interpretable signals of pre-Columbian forest modification, with the abstract providing limited information on how alternative explanations were ruled out.

free parameters (1)
  • Abundance significance threshold
    The determination of palms being 'significantly more abundant' near sites likely involves a statistical cutoff that may be data-dependent.
axioms (1)
  • domain assumption Palm tree distributions reflect pre-Columbian human activity
    Invoked to interpret the spatial association as evidence of ancient forest management.

pith-pipeline@v0.9.0 · 5801 in / 1403 out tokens · 49822 ms · 2026-05-19T05:08:43.674733+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

52 extracted references · 52 canonical work pages

  1. [1]

    Levis, et al., Persistent effects of pre-Columbian plant domestication on Amazonian forest composition

    C. Levis, et al., Persistent effects of pre-Columbian plant domestication on Amazonian forest composition. Science 355 (6328), 925–931 (2017), publisher: American Association for the Advancement of Science, https://www.jstor.org/stable/24918447

    work page arXiv 2017

  2. [2]

    S. D. Coelho, et al. , Eighty-four per cent of all Amazonian arboreal plant individuals are useful to humans. PLOS ONE 16 (10), e0257875 (2021), publisher: Public Library of Science, doi:10.1371/journal.pone.0257875, https://journals.plos.org/plosone/ article?id=10.1371/journal.pone.0257875

    work page doi:10.1371/journal.pone.0257875 2021

  3. [3]

    A. Leal, R. Gass ´on, H. Behling, F. S ´anchez, Human-made fires and forest clearance as evi- dence for late Holocene landscape domestication in the Orinoco Llanos (Venezuela). Vegeta- tion History and Archaeobotany 28 (5), 545–557 (2019), doi:10.1007/s00334-019-00713-w, https://doi.org/10.1007/s00334-019-00713-w

    work page doi:10.1007/s00334-019-00713-w 2019

  4. [4]

    S. Y. Maezumi, et al. , Fire-human-climate interactions in the Bolivian Amazon rainforest ecotone from the Last Glacial Maximum to late Holocene. Frontiers in Environmental Archaeology 2 (2023), publisher: Frontiers, doi:10.3389/fearc.2023. 1208985, https://www.frontiersin.org/journals/environmental-archaeology/ articles/10.3389/fearc.2023.1208985/full

    work page doi:10.3389/fearc.2023 2023

  5. [5]

    Peripato, et al., More than 10,000 pre-Columbian earthworks are still hidden throughout Amazonia

    V. Peripato, et al., More than 10,000 pre-Columbian earthworks are still hidden throughout Amazonia. Science 382 (6666), 103–109 (2023), publisher: American Association for the Advancement of Science, doi:10.1126/science.ade2541, https://www.science.org/doi/ 10.1126/science.ade2541

    work page doi:10.1126/science.ade2541 2023

  6. [6]

    R. K. Sales, et al. , Potential distributions of pre-Columbian people in Tropical An- dean landscapes. Philosophical Transactions of the Royal Society B: Biological Sciences 377 (1849), 20200502 (2022), publisher: Royal Society, doi:10.1098/rstb.2020.0502, https: //royalsocietypublishing.org/doi/10.1098/rstb.2020.0502

    work page doi:10.1098/rstb.2020.0502 2022

  7. [7]

    Persis- tent effects of pre-Columbian plant domestication on Amazonian forest composition

    C. H. McMichael, K. J. Feeley, C. W. Dick, D. R. Piperno, M. B. Bush, Comment on “Persis- tent effects of pre-Columbian plant domestication on Amazonian forest composition”.Science 19 358 (6361), eaan8347 (2017), publisher: American Association for the Advancement of Sci- ence, doi:10.1126/science.aan8347, https://www-science-org.ezproxy.leidenuniv. nl/do...

    work page doi:10.1126/science.aan8347 2017

  8. [8]

    C. N. H. McMichael, F. Matthews-Bird, W. Farfan-Rios, K. J. Feeley, Ancient human dis- turbances may be skewing our understanding of Amazonian forests. Proceedings of the National Academy of Sciences 114 (3), 522–527 (2017), publisher: Proceedings of the Na- tional Academy of Sciences, doi:10.1073/pnas.1614577114, https://www.pnas.org/doi/ 10.1073/pnas.1614577114

    work page doi:10.1073/pnas.1614577114 2017

  9. [9]

    C. R. Clement, et al., The domestication of Amazonia before European conquest.Proceedings of the Royal Society B: Biological Sciences 282 (1812), 20150813 (2015), publisher: Royal Society, doi:10.1098/rspb.2015.0813, https://royalsocietypublishing.org/doi/10. 1098/rspb.2015.0813

    work page doi:10.1098/rspb.2015.0813 2015

  10. [10]

    M. G. Letts, M. Mulligan, The impact of light quality and leaf wetness on photosynthesis in north-west Andean tropical montane cloud forest.Journal of Tropical Ecology21 (5), 549–557 (2005), doi:10.1017/S0266467405002488

    work page doi:10.1017/s0266467405002488 2005

  11. [11]

    T. J. Fahey, R. E. Sherman, E. V. Tanner, Tropical montane cloud forest: environmental drivers of vegetation structure and ecosystem function. Journal of Tropical Ecology 32 (5), 355–367 (2016), doi:10.1017/S0266467415000176

    work page doi:10.1017/s0266467415000176 2016

  12. [12]

    Instituto Geogr ´afico Agust´ın Codazzi (IGAC), Estudio general de suelos y zonificaci ´on de tierras del Departamento de Magdalena (Imprenta Nacional de Colombia Bogot´a, Colombia) (2009)

    work page 2009

  13. [13]

    S. D. Crausbay, P. H. Martin, Natural disturbance, vegetation patterns and ecological dynamics in tropical montane forests. Journal of Tropical Ecology32 (5), 384–403 (2016), doi:10.1017/ S0266467416000328

    work page 2016

  14. [14]

    C. H. Langebaek, ¿ Cu´antos eran?¿ D´onde estaban?¿ Qu´e les pas´o? Poblamiento ind´ıgena en la Colombia prehisp´anica y su transformaci´on despu´es de la Conquista, in Colombia. Preguntas y respuestas sobre su pasado y su presente (Ediciones Uniandes), pp. 27–52 (2010). 20

    work page 2010

  15. [15]

    Botiva Contreras, et al., Colombia prehisp´anica: regiones arqueol´ogicas (Colcultura, Insti- tuto Colombiano de Antropolog´ıa) (1989)

    A. Botiva Contreras, et al., Colombia prehisp´anica: regiones arqueol´ogicas (Colcultura, Insti- tuto Colombiano de Antropolog´ıa) (1989)

    work page 1989

  16. [16]

    Iriarte, et al., The origins of Amazonian landscapes: Plant cultivation, domestication and the spread of food production in tropical South America

    J. Iriarte, et al., The origins of Amazonian landscapes: Plant cultivation, domestication and the spread of food production in tropical South America. Quaternary Science Reviews 248, 106582 (2020), doi:10.1016/j.quascirev.2020.106582,https://www.sciencedirect.com/ science/article/pii/S0277379120305448

    work page doi:10.1016/j.quascirev.2020.106582 2020

  17. [17]

    N. Y. Paniagua-Zambrana, R. W. Bussmann, Dictyocaryum lamarckianum (Mart.) H. Wendl., in Ethnobotany of the Andes , N. Y. Paniagua-Zambrana, R. W. Bussmann, Eds. (Springer International Publishing, Cham), pp. 1–17 (2020), doi:10.1007/978-3-319-77093-2 99-1, https://doi.org/10.1007/978-3-319-77093-2_99-1

    work page doi:10.1007/978-3-319-77093-2 2020

  18. [18]

    Moraes, N

    M. Moraes, N. Paniagua-Zambrana, R. C ´amara-Leret, H. Balslev, M. Macia, Palmas ´ utiles de Bolivia, Colombia, Ecuador y Per´ u, inCosecha Palmas en el Noroeste Suram´erica Bases Cient´ıficas Para Su Manejo Conservaci´on, H. Balslev, M. Macia, H. Navarrete, Eds. (Pontificia Universidad Cat´olica del Ecuador, Quito, Ecuador), pp. 87–102 (2015)

    work page 2015

  19. [19]

    Rodr´ıguez Osorio, S

    D. Rodr´ıguez Osorio, S. Giraldo, E. Mazuera, A. Burbano, E. Figueredo, Beyond Visualization: Remote Sensing Applications in Prehispanic Settlements to Understand Ancient Anthropogenic Land Use and Occupation in the Sierra Nevada de Santa Marta, Colombia. Latin American Antiquity 34 (4), 821–841 (2023), doi:10.1017/laq.2022.91

    work page doi:10.1017/laq.2022.91 2023

  20. [20]

    Microsoft, Microsoft Global ML Footprints, https://github.com/microsoft/ GlobalMLBuildingFootprints (2023), accessed: 2025-05-30

    work page 2023

  21. [21]

    J. C. Vargas, W. Londo ˜no, M. Soto, Arqueolog´ıa de la microcuenca El Congo, Sierra Nevada de Santa Marta (Editorial Unimagdalena, Santa Marta, Colombia) (2022)

    work page 2022

  22. [22]

    L. M. Soto Rodr ´ıguez, J. C. Vargas, Network analysis in Tairona chiefdoms of the R ´ıo Fr´ıo basin, Sierra Nevada de Santa Marta, Colombia.Journal of Archaeological Science179, 106230 (2025), doi:https://doi.org/10.1016/j.jas.2025.106230,https://www.sciencedirect.com/ science/article/pii/S0305440325000792. 21

    work page doi:10.1016/j.jas.2025.106230 2025

  23. [23]

    S. Giraldo, Lords of the Snowy Ranges: Politics, place, and landscape transformation in two Tairona towns in the Sierra Nevada de Santa Marta, Colombia , Phd thesis, University of Chicago, Chicago, Illinois (2010)

    work page 2010

  24. [24]

    van der Hammen, G

    T. van der Hammen, G. W. Noldus, Paleoecology of the Ci ´enaga Grande (Great Lagoon) of Santa Marta, in La Sierra Nevada de Santa Marta (Colombia) Transecto Buritaca la Cumbre, T. van der Hammen, P. M. Ruiz, Eds. (J.Cramer, Berlin-Stuggart, Germany), pp. 581–588 (1984)

    work page 1984

  25. [25]

    L. F. Herrera, Buritaca 200: survey of the archaeology and a palynological investigation, inLa Sierra Nevada de Santa Marta (Colombia) Transecto Buritaca la Cumbre, T. van der Hammen, P. M. Ruiz, Eds. (J.Cramer, Berlin-Stuggart, Germany), pp. 531–546 (1984)

    work page 1984

  26. [26]

    M. I. V ´elez, et al., Humans and climate in ritualized landscapes, the case of Lake Tota in the eastern highlands of Colombia. The Holocene 34 (11), 1587–1597 (2024), doi:10.1177/ 09596836241266408

    work page 2024

  27. [27]

    K. N. Laland, J. Odling-Smee, M. W. Feldman, Niche construction, biological evolution, and cultural change. Behavioral and brain sciences 23 (1), 131–146 (2000)

    work page 2000

  28. [28]

    R. P. Fortier, et al., Diversity, composition, and structure of a 1-hectare tree plot in the cloud for- est of the Sierra Nevada de Santa Marta, Colombia.Annals of Forest Science81 (1), 24 (2024), doi:10.1186/s13595-024-01239-6, https://doi.org/10.1186/s13595-024-01239-6

    work page doi:10.1186/s13595-024-01239-6 2024

  29. [29]

    Dechner, M

    A. Dechner, M. Diazgranados, Floristic composition and structure of forested areas along the low basin of the San Salvador river, north side of Sierra Nevada de Santa Marta, Colombia. Univ. Sci. 12 (2), 99–124 (2007)

    work page 2007

  30. [30]

    Norden, et al., Successional dynamics in Neotropical forests are as uncertain as they are pre- dictable

    N. Norden, et al., Successional dynamics in Neotropical forests are as uncertain as they are pre- dictable. Proceedings of the National Academy of Sciences112 (26), 8013–8018 (2015), doi:10. 1073/pnas.1500403112, https://www.pnas.org/doi/abs/10.1073/pnas.1500403112

    work page doi:10.1073/pnas.1500403112 2015

  31. [31]

    B. M. Pavlik, et al., Plant species richness at archaeological sites suggests ecological legacy of Indigenous subsistence on the Colorado Plateau. Proceedings of the National Academy 22 of Sciences 118 (21), e2025047118 (2021), doi:10.1073/pnas.2025047118, https://www. pnas.org/doi/abs/10.1073/pnas.2025047118

    work page doi:10.1073/pnas.2025047118 2021

  32. [32]

    Sapir, Y

    Y. Sapir, Y. Sapir, A. Faust, Using floristic characteristics of contemporary vegetation for identifying archaeological sites: Tel ‘Eton archaeological site as a test case. Israel Journal of Plant Sciences 66 (1–2), 60 – 68 (2019), doi:10.1163/22238980-00001029,https://brill. com/view/journals/ijps/66/1-2/article-p60_60.xml

    work page doi:10.1163/22238980-00001029 2019

  33. [33]

    Hambrecht, et al., Archaeological sites as Distributed Long-term Observing Networks of the Past (DONOP)

    G. Hambrecht, et al., Archaeological sites as Distributed Long-term Observing Networks of the Past (DONOP). Quaternary International 549, 218–226 (2020), long-term perspectives on circumpolar social-ecological systems, doi:https://doi.org/10.1016/j.quaint.2018.04.016, https://www.sciencedirect.com/science/article/pii/S1040618217311746

    work page doi:10.1016/j.quaint.2018.04.016 2020

  34. [34]

    M. P. Ferreira, et al. , Individual tree detection and species classification of Amazonian palms using UA V images and deep learning. Forest Ecology and Management 475, 118397 (2020), doi:https://doi.org/10.1016/j.foreco.2020.118397, https://www.sciencedirect. com/science/article/pii/S037811272031166X

    work page doi:10.1016/j.foreco.2020.118397 2020

  35. [35]

    H. Zhou, F. Jiang, H. Lu, SSDA-YOLO: Semi-supervised domain adaptive YOLO for cross- domain object detection. Computer Vision and Image Understanding229, 103649 (2023)

    work page 2023

  36. [36]

    Zhang, H

    S. Zhang, H. Tuo, J. Hu, Z. Jing, Domain adaptive YOLO for one-stage cross-domain detection, in Asian conference on machine learning (PMLR) (2021), pp. 785–797

    work page 2021

  37. [37]

    A. C. Rodr´ıguez, S. D’ Aronco, K. Schindler, J. D. Wegner, Mapping oil palm density at country scale: An active learning approach. Remote Sensing of Environment 261, 112479 (2021)

    work page 2021

  38. [38]

    Serje, Arquitectura y urbanismo en la cultura Tairona

    M. Serje, Arquitectura y urbanismo en la cultura Tairona. Bol. Mus. Oro 19 (1), 87–96 (1987)

    work page 1987

  39. [39]

    Dever, Social and economic development of a specialized community in Chengue, Parque Tairona, Colombia, Ph.D

    A. Dever, Social and economic development of a specialized community in Chengue, Parque Tairona, Colombia, Ph.D. thesis, University of Pittsburgh, Pittsburgh, Pennsylvania (2007)

    work page 2007

  40. [40]

    N. Guti ´errez Montoya, Las sociedades originarias de la Sierra Nevada de Santa Marta Ocu- paci´on espacial y desarrollo arquitect ´onico Siglos x al xvi (ediciones doce calles, Madrid) (2022). 23

    work page 2022

  41. [41]

    Y ´epez, Y

    Y. Y ´epez, Y. Guti´errez, Relato del espacio, territorio y lugar de los asentamientos Kogui dentro de la L´ınea Negra: Desarrollo cartogr´afico de ocupaci´on e integraci´on, Ba thesis, Univerisidad de la Costa C.U.C., Barranquilla, Atl´antico (2020)

    work page 2020

  42. [42]

    G. Reichel-Dolmatoff, Cultural Change and Environmental Awareness: A Case Study of the Sierra Nevada de Santa Marta, Colombia.Mountain Research and Development2 (3), 289–298 (1982), http://www.jstor.org/stable/3673093

    work page arXiv 1982

  43. [43]

    Cadavid, L

    G. Cadavid, L. F. Herrera, Manifestaciones culturales en el ´area Tairona. Inform. Antropol. 1, 5–54 (1985)

    work page 1985

  44. [44]

    A. M. Groot, Arqueolog ´ıa y conservaci´on de la localidad precolombina de Buritaca 200 en la Sierra Nevada de Santa Marta. Inform. Antropol. 1, 55–102 (1985)

    work page 1985

  45. [45]

    L. F. Herrera, Agricultural activity in the Sierra Nevada de Santa Marta (Colombia): historical perspective, in La Sierra Nevada de Santa Marta (Colombia) Transecto Buritaca la Cumbre, T. van der Hammen, P. M. Ruiz, Eds. (J.Cramer, Berlin-Stuggart, Germany), pp. 501–530 (1984)

    work page 1984

  46. [46]

    M. I. Mayorga Hern ´andez, El agua como elemento constructor de territorio y arquitectura en Colombia. Caso de estudio Teyuna, Sierra Nevada de Santa Marta, Phd thesis, Universidad de Granada, Granada, Spain (2024), https://hdl.handle.net/10481/89459

    work page 2024

  47. [47]

    P. M. van der Hammen, Thomas; Ruiz, La Sierra Nevada de Santa Marta (Colom- bia) Transecto Buritaca – La Cumbre , vol. 2 (Schweizerbart Science Publishers, Stuttgart, Germany) (1986), http://www.schweizerbart.de//publications/detail/ isbn/9783443650018/Studies\_on\_Tropical\_Andean\_Ecosystems\_

    work page arXiv 1986

  48. [48]

    Henderson, G

    A. Henderson, G. Galeano, R. Bernal, Field guide to the palms of the Americas (Princeton University Press) (1995)

    work page 1995

  49. [49]

    Wang, et al., YOLOv10: Real-Time End-to-End Object Detection, in Advances in Neural Information Processing Systems, A

    A. Wang, et al., YOLOv10: Real-Time End-to-End Object Detection, in Advances in Neural Information Processing Systems, A. Globerson, et al., Eds. (Curran Associates, Inc.), vol. 37 (2024), pp. 107984–108011, https://proceedings.neurips.cc/paper_files/paper/ 2024/file/c34ddd05eb089991f06f3c5dc36836e0-Paper-Conference.pdf. 24

    work page 2024

  50. [50]

    R. J. G. B. Campello, D. Moulavi, J. Sander, Density-Based Clustering Based on Hierarchical Density Estimates, in Advances in Knowledge Discovery and Data Mining, J. Pei, V. S. Tseng, L. Cao, H. Motoda, G. Xu, Eds. (Springer Berlin Heidelberg, Berlin, Heidelberg) (2013), pp. 160–172

    work page 2013

  51. [51]

    GBIF, Dictyocaryum H.Wendl. Occurrence Download, https://www.gbif.org/occurrence/download/0025145-250310093411724 (2025), doi:10.15468/DL.ND44PT, https://www.gbif.org/occurrence/download/ 0025145-250310093411724, global Biodiversity Information Facility. Accessed: 2025-03- 24

    work page doi:10.15468/dl.nd44pt 2025

  52. [52]

    NASA Shuttle Radar Topography Mission Global 1 arc sec- ond [Data set]

    NASA JPL, NASA Shuttle Radar Topography Mission Global 1 arc second [Data set],https: //doi.org/10.5067/MEASURES/SRTM/SRTMGL1.003 (2013), nASA Land Processes Dis- tributed Active Archive Center. Accessed: 2025-06-20. Acknowledgments Funding: This research is part of the ”Mapping the Archaeological Pre-Columbian Heritage in South America – MAPHSA” Project ...

    work page doi:10.5067/measures/srtm/srtmgl1.003 2013