Archaeological prospectivity and terrace morphometry using open geospatial data: a reproducible GIS–machine-learning workflow for the cuyes river valley, ecuador

It may state two separate claims as one.

This form of compressed verticality is characteristic of the eastern-Andean foothills, where strong environmental contrasts occur across relatively short horizontal distances.

“This form of compressed verticality is characteristic of the eastern-Andean foothills, where strong environmental contrasts occur across relatively short horizontal distances (Rostain 2012;Valdez 2023).”

• cites Rostain 2012Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Valdez 2023 sourceCould not verify

  The discussion of the Huancabamba Depression describes low mountain passes and access between the Andes and lowland forest; the Upano section describes the eastern-Andean foothills and movement among ecological niches. Neither passage says that strong environmental contrasts occur over short horizontal distances or characterizes the foothills as having compressed verticality. I also searched the paper for altitudinal gradients, ecological zones, and short-distance contrasts; no passage addressed that specific claim. The available passages have no page numbers. (LLM, conf=0.72)

  Experimental — accuracy not yet measured

Probably not a substantial claim. The paper’s conclusions may not depend on it.

This distinction follows broader cautions in archaeological predictive modelling concerning incomplete survey coverage and the rarity of reliable absence data.

“This distinction follows broader cautions in archaeological predictive modelling concerning incomplete survey coverage and the rarity of reliable absence data (Yaworsky et al. 2020).”

• cites Yaworsky et al. 2020 sourceSupported

Probably not a substantial claim. The paper’s conclusions may not depend on it.

The generalized distribution (Figure 2) preserves the separation between upper valley and lower valley terrace systems without exposing exact archaeological coordinates.

“The generalized distribution (Figure 2) preserves the separation between upper valley and lower valley terrace systems without exposing exact archaeological coordinates.”

It may state two separate claims as one.

The workflow produced a harmonized inventory of 78 terrace records: 17 TEP, 35 TSM and 26 TNZ.

“The workflow produced a harmonized inventory of 78 terrace records: 17 TEP, 35 TSM and 26 TNZ.”

It may state two separate claims as one.

Area plotted against reported elevation further separated TNZ from the upper valley systems: TNZ occupied approximately 1290-1508 m a.s.l., whereas most TSM and TEP records occurred above 2150 m a.s.l.

“Area plotted against reported elevation further separated TNZ from the upper valley systems: TNZ occupied approximately 1290-1508 m a.s.l., whereas most TSM and TEP records occurred above 2150 m a.s.l.(Figure 4).”

It may state two separate claims as one.

Across the inventory the median was 60.48 m2 and the mean 79.29 m2, with values ranging from 5.60 to 263.00 m2.

“Across the inventory the median was 60.48 m 2 and the mean 79.29 m 2 , with values ranging from 5.60 to 263.00 m 2 (Figure 3).”

It may state two separate claims as one.

Espíritu Playa represents the highest of the three analysed terrace systems, San Miguel de Cuyes occupies an intermediate montane zone, and Nueva Zaruma lies considerably lower toward the Amazonian sector.

“Espíritu Playa consequently represents the highest of the three analysed terrace systems, San Miguel de Cuyes occupies an intermediate montane zone, and Nueva Zaruma lies considerably lower toward the Amazonian sector.”

• cites Lara 2009aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

It may state two separate claims as one.

After harmonization, 78 terrace records were retained: 17 TEP, 35 TSM and 26 TNZ records with usable quantitative information.

“After harmonization, 78 terrace records were retained: 17 TEP, 35 TSM and 26 TNZ records with usable quantitative information.”

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

It may state two separate claims as one.

Most terraces were below 100 m2, whereas a smaller upper tail comprised larger construction units.

“Most terraces were below 100 m 2 , whereas a smaller upper tail comprised larger construction units.”

It may state two separate claims as one.

TEP contained both very small terraces and the largest record, TSM was dominated by moderate-sized forms with several larger examples, and TNZ showed a compact lower valley distribution with some medium-to-large terraces.

“TEP contained both very small terraces and the largest record, TSM was dominated by moderate-sized forms with several larger examples, and TNZ showed a compact lower valley distribution with some mediumto-large terraces.”

It may state two separate claims as one.

TNZ was also wetter, with mean annual precipitation of 1595.9 mm compared with approximately 840 mm for TEP and 822.6 mm for TSM.

“TNZ was also wetter, with mean annual precipitation of 1595.9 mm compared with approximately 840 mm for TEP and 822.6 mm for TSM (Table 1).”

Probably not a substantial claim. The paper’s conclusions may not depend on it.

The Cuyes Valley terrace systems extend approximately from 1290 to 2520 m a.s.l.

“The terrace systems analysed here extend approximately from 1290 to 2520 m a.s.l.”

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

It may state two separate claims as one.

Reported elevation, DEM elevation, slope, roughness and WorldClim variables were complete for all records; terrace area was available for 77 records and height for 60.

“Reported elevation, DEM elevation, slope, roughness and WorldClim variables were complete for all records; terrace area was available for 77 records and height for 60.”

It may state two separate claims as one.

Across valid spatial folds, mean ROC-AUC was 0.753, and mean average precision was 0.604.

“Across valid spatial folds, mean ROC-AUC was 0.753, and mean average precision was 0.604 (Table 3).”

It may state two separate claims as one.

TEP and TSM formed an upper valley group with mean reported elevations of 2473.4 and 2315.7 m a.s.l., respectively, whereas TNZ occupied a substantially lower sector with a mean elevation of 1340.9 m a.s.l.

“TEP and TSM formed an upper valley group with mean reported elevations of 2473.4 and 2315.7 m a.s.l., respectively, whereas TNZ occupied a substantially lower sector with a mean elevation of 1340.9 m a.s.l.”

It may state two separate claims as one.

Cluster 2 was the largest (n = 32) and represented moderate, high-elevation terraces concentrated in TEP and TSM.

“Cluster 2 was the largest (n = 32) and represented moderate, high-elevation terraces concentrated in TEP and TSM.”

It may state two separate claims as one.

The presence-background random forest model achieved a mean spatial ROC-AUC of 0.753, indicating useful ranking performance for exploratory survey prioritization while also showing sufficient spatial variability to require conservative interpretation and independent field verification.

“The presence-background random forest model achieved a mean spatial ROC-AUC of 0.753, indicating useful ranking performance for exploratory survey prioritization while also showing sufficient spatial variability to require conservative interpretation and independent field verification.”

It may state two separate claims as one.

Higher scoring cells occurred around both upper valley terrace environments and selected lower valley sectors, reflecting the combined effect of precipitation, elevation and temperature.

“Higher scoring cells occurred around both upper valley terrace environments and selected lower valley sectors, reflecting the combined effect of precipitation, elevation and temperature.”

It may state two separate claims as one.

TEP and TSM occur at higher, cooler and substantially drier elevations, whereas TNZ represents a lower, warmer and wetter terrace context.

“TEP and TSM occur at higher, cooler and substantially drier elevations, whereas TNZ represents a lower, warmer and wetter terrace context.”

It may state two separate claims as one.

In the Cuyes River Valley, the integration of 78 documented terrace records with openly accessible environmental geospatial data revealed a marked eco-topographic differentiation between upper and lower valley systems and identified recurrent morphometric patterns.

“In the Cuyes River Valley, the integration of 78 documented terrace records with openly accessible environmental geospatial data revealed a marked eco-topographic differentiation between upper and lower valley systems and identified recurrent morphometric patterns.”

It may state two separate claims as one.

Annual precipitation had the greatest relative feature importance, followed by DEM elevation and annual mean temperature; precipitation seasonality contributed moderately, whereas roughness and slope were less influential.

“Annual precipitation had the greatest relative feature importance, followed by DEM elevation and annual mean temperature; precipitation seasonality contributed moderately, whereas roughness and slope were less influential (Figure 6).”

It may state two separate claims as one.

The upper terrace systems occupy eastern-Andean montane environments above 2300 m a.s.l., whereas Nueva Zaruma lies more than 1000 m lower under markedly wetter conditions, although all three systems belong to the same river corridor draining toward the Amazon basin.

“The upper terrace systems occupy eastern-Andean montane environments above 2300 m a.s.l., whereas Nueva Zaruma lies more than 1000 m lower under markedly wetter conditions, although all three systems belong to the same river corridor draining toward the Amazon basin.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Nueva Zaruma I produced a determination of 1270-1400 ad together with a later dated context of 1450-1650 ad.

“Nueva Zaruma I produced a determination of 1270-1400 ad together with a later dated context of 1450-1650 ad.”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

It may state two separate claims as one.

These determinations indicate occupation during the late second millennium bc, extending into the early first millennium bc.

“These determinations indicate occupation during the late second millennium bc, extending into the early first millennium bc”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

Probably not a substantial claim. The paper’s conclusions may not depend on it.

Exact archaeological coordinates were treated as restricted information throughout the workflow.

“Exact archaeological coordinates were treated as restricted information throughout the workflow.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

The reports described competing interpretations of cultural affiliation-including Inca, Cañari, local and mixed possibilities-and documented continuing threats from mining, roads, timber extraction, erosion, livestock, vegetation and looting.

“The reports also described competing interpretations of cultural affiliation-including Inca, Cañari, local and mixed possibilities-and documented continuing threats from mining, roads, timber extraction, erosion, livestock, vegetation and looting.”

• cites Lara 2009aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2009bNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

Probably not a substantial claim. It may state two separate claims as one.

This field-verification stage was not undertaken in the present study and is therefore treated as a required next step rather than as completed validation.

“This field-verification stage was not undertaken in the present study and is therefore treated as a required next step rather than as completed validation.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

The original environmental survey distinguished lower montane humid forest around Espíritu Playa, very humid lower montane forest around San Miguel and Ganazhuma and progressively lower premontane environments toward Nueva Tarqui and the lower valley.

“The original environmental survey distinguished lower montane humid forest around Espíritu Playa, very humid lower montane forest around San Miguel and Ganazhuma and progressively lower premontane environments toward Nueva Tarqui and the lower valley”

• cites Lara 2009aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

The second report supplied most terrace coordinates and measurements, whereas the final report provided broader archaeological context.

“The second report supplied most terrace coordinates and measurements, whereas the final report provided broader archaeological context.”

• cites Lara 2009bNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

Probably not a substantial claim. It may be too vague to check against evidence.

A fourth stage-independent field ground-truthing of predicted high-, intermediate- and low-prospectivity sectors-is considered necessary for external validation of the digital results.

“A fourth stage-independent field ground-truthing of predicted high-, intermediate-and low-prospectivity sectors-is considered necessary for external validation of the digital results.”

It may state two separate claims as one.

Spatial dependence presents a challenge because random train-test partitioning may place environmentally similar neighbouring observations in both subsets, inflating apparent predictive performance.

“Spatial dependence presents an additional challenge because random train-test partitioning may place environmentally similar neighbouring observations in both subsets, inflating apparent predictive performance.”

• cites Roberts et al. 2017 sourceNot checked — no full text of the source

Probably not a substantial claim. The paper’s conclusions may not depend on it.

Sitio Playa produced calibrated ranges of 1310-1040 bc and 1210-970/960-940 bc.

“Sitio Playa produced calibrated ranges of 1310-1040 bc and 1210-970/960-940 bc”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

Probably not a substantial claim. The paper’s conclusions may not depend on it.

The classification was used as exploratory morphometry rather than as evidence of cultural affiliation.

“The classification was used as exploratory morphometry rather than as evidence of cultural affiliation.”

It may state two separate claims as one.

Archaeological investigations conducted between 2009 and 2010 documented a monumental landscape distributed across multiple sectors and identified terraces as a major component of the valley's built environment.

“Archaeological investigations conducted between 2009 and 2010 documented a monumental landscape distributed across multiple sectors and identified terraces as a major component of the valley's built environment”

• cites Lara 2009aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2009bNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

Probably not a substantial claim.

The workflow comprised three completed analytical components and one subsequent validation stage required for operational archaeological prospection.

“The workflow comprised three completed analytical components and one subsequent validation stage required for operational archaeological prospection.”

Probably not a substantial claim. The paper’s conclusions may not depend on it.

Santa Rosa yielded a calibrated range extending approximately from 1250 to 980 bc.

“Santa Rosa yielded a calibrated range extending approximately from 1250 to 980 bc”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

It may state two separate claims as one.

This retained broad spatial patterns while reducing disclosure risk for vulnerable archaeological locations.

“This retained broad spatial patterns while reducing disclosure risk for vulnerable archaeological locations.”

It may state two separate claims as one.

Spatial blocking reduces leakage between environmentally similar neighbouring samples and provides a stricter assessment of transfer to unsampled sectors.

“Spatial blocking reduces leakage between environmentally similar neighbouring samples and provides a stricter assessment of transfer to unsampled sectors (Roberts et al. 2017).”

• cites Roberts et al. 2017 sourceNot checked — no full text of the source

It may state two separate claims as one.

Phytolith analyses from terrace deposits at San Miguel de Cuyes and Nueva Zaruma identified maize (Zea mays), achira (Canna spp.) and lerén or bijao (Calathea spp.), supporting cultivation on at least part of the terraced landscape.

“Phytolith analyses from terrace deposits at San Miguel de Cuyes and Nueva Zaruma identified maize (Zea mays), achira (Canna spp.) and lerén or bijao (Calathea spp.), supporting cultivation on at least part of the terraced landscape”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

Probably not a substantial claim.

The current model output is interpreted as a survey-prioritization surface, not as field-confirmed evidence of undiscovered archaeological sites.

“Accordingly, the current model output is interpreted as a surveyprioritization surface, not as field-confirmed evidence of undiscovered archaeological sites.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

La Florida was dated to 1410-1470 ad and Buenos Aires to 1440-1640 ad.

“La Florida was dated to 1410-1470 ad and Buenos Aires to 1440-1640 ad.”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Feature importance was reported from the fitted random forest as a relative measure of variable use and was not interpreted causally because known-site patterns may also reflect preservation, accessibility, survey history and reporting bias.

“Feature importance was reported from the fitted random forest as a relative measure of variable use and was not interpreted causally because known-site patterns may also reflect preservation, accessibility, survey history and reporting bias (Breiman 2001).”

• cites Breiman 2001 sourceNot checked — no full text of the source

Probably not a substantial claim. The paper’s conclusions may not depend on it.

Exact archaeological coordinates are excluded from the public archive for heritage protection reasons.

“Exact archaeological coordinates are excluded from the public archive for heritageprotection reasons.”

Probably not a substantial claim.

Variables were included in the predictive model only when they were numeric and complete for both known terraces and the prediction grid.

“Variables were included in the predictive model only when they were numeric and complete for both known terraces and the prediction grid.”

Probably not a substantial claim. The paper’s conclusions may not depend on it.

Espíritu Playa yielded calibrated ranges of 1430-1530 and 1560-1630 ad.

“Espíritu Playa yielded calibrated ranges of 1430-1530 and 1560-1630 ad”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

It may state two separate claims as one.

The model's ranking performance was useful but spatially variable.

“indicating useful but spatially variable ranking performance”

Probably not a substantial claim. It may state two separate claims as one.

Exact archaeological coordinates were withheld and public outputs spatially generalized.

“Exact archaeological coordinates were withheld and public outputs spatially generalized.”

• AssumptionRests on an assumption: Spatial generalization of public outputs sufficiently reduces the risk of disclosing vulnerable archaeological locations.

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Eighteen archaeological sites were documented and organized into six sectors extending from the upper to the lower valley.

“Eighteen archaeological sites were documented and organized into six sectors extending from the upper to the lower valley.”

• cites Lara 2009aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2009bNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: The legacy archaeological reports accurately record terrace locations, dimensions, elevations and chronology.

It may state two separate claims as one.

The valley preserves evidence for at least two broad occupational components: an early component dating to the late second millennium bc and extending into the early first millennium bc, concentrated in the upper valley, and a later sequence of intensive occupation and monumental construction beginning during the late pre-Hispanic period and extending into the early colonial era.

“The valley therefore preserves evidence for at least two broad occupational components: an early component dating to the late second millennium bc and extending into the early first millennium bc, concentrated in the upper valley, and a later sequence of intensive occupation and monumental construction beginning during the late pre-Hispanic period and extending into the early colonial era.”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Folds containing only one response class could not yield ROC-AUC and were excluded from metric reporting; three folds contained both presence and background observations and therefore contributed to the reported cross-validation statistics.

“Folds containing only one response class could not yield ROC-AUC and were excluded from metric reporting; three folds contained both presence and background observations and therefore contributed to the reported cross-validation statistics.”

It may state two separate claims as one.

The clustering identifies recurrent construction forms while also demonstrating that environmental setting and morphology do not map one-to-one: Lower valley TNZ terraces can resemble moderate upper valley terraces in area despite occupying very different eco-climatic conditions.

“The clustering therefore identifies recurrent construction forms while also demonstrating that environmental setting and morphology do not map one-to-one: Lower valley TNZ terraces can resemble moderate upper valley terraces in area despite occupying very different eco-climatic conditions.”

Probably not a substantial claim. It may state two separate claims as one.

Model outputs are interpreted as relative archaeological prospectivity: a ranking of environmental similarity to known terrace locations that can be used to prioritize subsequent survey.

“Because background points represent available landscape rather than verified archaeological absence, model outputs are interpreted as relative archaeological prospectivity: a ranking of environmental similarity to known terrace locations that can be used to prioritize subsequent survey.”

• AssumptionRests on an assumption: Background points represent available landscape rather than verified archaeological absence.

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Records lacking coordinates were retained in the documentation layer but excluded from spatial modelling.

“Records lacking coordinates were retained in the documentation layer but excluded from spatial modelling.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Consequently, the repository supports inspection and execution of the analytical workflow and reproduction using synthetic data, whereas exact regeneration of coordinate-dependent Cuyes spatial outputs requires authorized access to the restricted archaeological input locations.

“Consequently, the repository supports inspection and execution of the analytical workflow and reproduction using synthetic data, whereas exact regeneration of coordinate-dependent Cuyes spatial outputs requires authorized access to the restricted archaeological input locations.”

It may state two separate claims as one.

Prospectivity scores are not interpreted as probabilities of archaeological presence, and low values do not constitute evidence of archaeological absence.

“Prospectivity scores are not interpreted as probabilities of archaeological presence, and low values do not constitute evidence of archaeological absence.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Radiocarbon results indicate that occupation of the valley was temporally complex and substantially predates the better-known late pre-Hispanic monumental phase.

“Radiocarbon results indicate that occupation of the valley was temporally complex and substantially predates the betterknown late pre-Hispanic monumental phase.”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionRests on an assumption: Radiocarbon determinations reported in the legacy documentation are reliable.

Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.

These characteristics make the workflow relevant beyond the Cuyes River Valley.

“These characteristics make the workflow relevant beyond the Cuyes River Valley.”

It may state two separate claims as one.

The fold-specific evaluation reinforces the spatial sensitivity of the classifier: Performance was high in Folds 2 and 5 but poor in Fold 4, a pattern that would have been obscured by a conventional random split.

“The fold-specific evaluation (Figure 7) reinforces the spatial sensitivity of the classifier: Performance was high in Folds 2 and 5 but poor in Fold 4, a pattern that would have been obscured by a conventional random split.”

Probably not a substantial claim. It may be too vague to check against evidence.

The value of the present analysis is to narrow environmental and spatial hypotheses that can be tested by future archaeological work rather than to replace those lines of evidence.

“The value of the present analysis is to narrow environmental and spatial hypotheses that can be tested by future archaeological work rather than to replace those lines of evidence.”

It may state two separate claims as one.

Finally, the small and spatially clustered presence sample limits model complexity and makes fold-level uncertainty important.

“Finally, the small and spatially clustered presence sample limits model complexity and makes foldlevel uncertainty important.”

Probably not a substantial claim. It may be too vague to check against evidence.

This iterative relationship between prediction and field verification is essential if prospectivity modelling is to move from exploratory landscape analysis toward an operational archaeological survey tool.

“This iterative relationship between prediction and field verification is essential if prospectivity modelling is to move from exploratory landscape analysis toward an operational archaeological survey tool.”

Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.

It can help prioritize pedestrian survey, UAV or LiDAR acquisition, geoarchaeological sampling, monitoring of vulnerable sectors and decisions concerning where intensive archaeological investigation is most likely to improve existing knowledge.

“It can help prioritize pedestrian survey, UAV or LiDAR acquisition, geoarchaeological sampling, monitoring of vulnerable sectors and decisions concerning where intensive archaeological investigation is most likely to improve existing knowledge.”

It may state two separate claims as one.

More broadly, it shows that data scarcity does not necessarily prevent quantitative archaeological landscape analysis when legacy records are critically evaluated, environmental datasets are openly accessible, uncertainty is made explicit, and digital predictions remain connected to archaeological field verification.

“More broadly, it shows that data scarcity does not necessarily prevent quantitative archaeological landscape analysis when legacy records are critically evaluated, environmental datasets are openly accessible, uncertainty is made explicit, and digital predictions remain connected to archaeological field verification.”

Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.

The same logic can be adapted to other terrace systems and, with appropriate archaeological reformulation, to other spatially distributed heritage features.

“The same logic can be adapted to other terrace systems and, with appropriate archaeological reformulation, to other spatially distributed heritage features.”

It may state two separate claims as one.

The dominance of precipitation, elevation and temperature shows how the fitted classifier distinguishes known terraces from sampled background, not that these variables caused ancient settlement decisions.

“Feature importance must be read similarly: The dominance of precipitation, elevation and temperature shows how the fitted classifier distinguishes known terraces from sampled background, not that these variables caused ancient settlement decisions.”

It may state two separate claims as one.

The morphometric groups show repeated forms but do not support a single functional assignment.

“The morphometric groups show repeated forms but do not support a single functional assignment.”

It may state two separate claims as one.

Terrace geometry alone cannot distinguish agricultural, domestic, stabilizing, connective or ceremonial functions, and the original Cuyes reports already emphasized the need for further stratigraphic and archaeobotanical study.

“Terrace geometry alone cannot distinguish agricultural, domestic, stabilizing, connective or ceremonial functions, and the original Cuyes reports already emphasized the need for further stratigraphic and archaeobotanical study (Lara 2010).”

• cites Lara 2010Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.

The workflow has particular potential for other Andean terrace landscapes because many of the required inputs are already common to archaeological terrace documentation.

“The workflow has particular potential for other Andean terrace landscapes because many of the required inputs are already common to archaeological terrace documentation:”

It may state two separate claims as one.

Background points are not true absences, and the model cannot determine that low-scoring sectors lack archaeology.

“Background points are not true absences, and the model cannot determine that low-scoring sectors lack archaeology.”

Probably not a substantial claim.

First, the inventory inherits uneven survey intensity, visibility, preservation and reporting from legacy investigations.

“First, the inventory inherits uneven survey intensity, visibility, preservation and reporting from legacy investigations.”

Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.

This balance is especially relevant for archaeological prospection, where better spatial knowledge can simultaneously improve conservation and increase exposure.

“This balance is especially relevant for archaeological prospection, where better spatial knowledge can simultaneously improve conservation and increase exposure.”

Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.

Fourth, spatially structured validation provides a more conservative assessment of transferability than conventional random data splitting and helps identify sectors where additional observations would be most informative.

“Fourth, spatially structured validation provides a more conservative assessment of transferability than conventional random data splitting and helps identify sectors where additional observations would be most informative.”

Probably not a substantial claim. The paper may give no reason why it holds.

Cultural attribution requires ceramic, architectural, chronological and archaeobotanical evidence.

“Cultural attribution requires ceramic, architectural, chronological and archaeobotanical evidence.”

Probably not a substantial claim. It may be too vague to check against evidence.

Field ground-truthing remains essential before the prospectivity surface can be regarded as an operational archaeological prediction tool.

“Field ground-truthing remains essential before the prospectivity surface can be regarded as an operational archaeological prediction tool.”

Probably not a substantial claim. It may state two separate claims as one.

Prospectivity scores identify where limited survey resources can be allocated strategically, whereas the spatial cross-validation results indicate where model extrapolation is weakest and additional observations are most valuable.

“Prospectivity scores identify where limited survey resources can be allocated strategically, whereas the spatial cross-validation results indicate where model extrapolation is weakest and additional observations are most valuable.”

It may state two separate claims as one.

The public archive therefore provides aggregated nonsensitive derived tables and a synthetic 78-record demonstration dataset that contains no authentic archaeological locations.

“The public archive therefore provides aggregated nonsensitive derived tables and a synthetic 78-record demonstration dataset that contains no authentic archaeological locations.”

Probably not a substantial claim. The paper’s conclusions may not depend on it.

These values indicate model associations rather than causal drivers.

“These values indicate model associations rather than causal drivers.”

Probably not a substantial claim.

First, it extends the scientific life of legacy archaeological reports by converting dispersed coordinates, measurements, plans and contextual descriptions into a structured and analysable spatial inventory.

“First, it extends the scientific life of legacy archaeological reports by converting dispersed coordinates, measurements, plans and contextual descriptions into a structured and analysable spatial inventory.”

Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.

The framework therefore supports both research design and preventive heritage management.

“The framework therefore supports both research design and preventive heritage management.”

Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.

Terraces should therefore be interpreted as flexible landscape infrastructure.

“Terraces should therefore be interpreted as flexible landscape infrastructure.”

Probably not a substantial claim. It may be too vague to check against evidence.

The approach is particularly useful as a bridge between archival information and renewed field research.

“The approach is particularly useful as a bridge between archival information and renewed field research.”

Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.

Second, it provides a comparatively low-cost first analytical stage for regions where comprehensive new survey, LiDAR, UAV mapping or excavation is not immediately feasible.

“Second, it provides a comparatively low-cost first analytical stage for regions where comprehensive new survey, LiDAR, UAV mapping or excavation is not immediately feasible.”

Probably not a substantial claim. It may state two separate claims as one.

The surface is not a map of confirmed archaeological probability: High values indicate relative similarity to the environmental signature represented in the known inventory.

“The surface is not a map of confirmed archaeological probability: High values indicate relative similarity to the environmental signature represented in the known inventory.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

Third, the prediction surface uses topographic and climatic covariates; available SoilGrids descriptors were not propagated through the full prediction grid.

“Third, the prediction surface uses topographic and climatic covariates; available SoilGrids descriptors were not propagated through the full prediction grid.”

Probably not a substantial claim. It may be too vague to check against evidence.

For the Cuyes River Valley, the study provides a reproducible basis for renewed investigation of a poorly documented Andean-Amazonian terrace landscape while retaining appropriate caution regarding chronology, function and cultural affiliation.

“For the Cuyes River Valley, the study provides a reproducible basis for renewed investigation of a poorly documented Andean-Amazonian terrace landscape while retaining appropriate caution regarding chronology, function and cultural affiliation.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.

Second, presence-background modelling cannot substitute for verified absence data.

“Second, presence-background modelling cannot substitute for verified absence data.”

Probably not a substantial claim.

For archaeological prospection, this is preferable to reporting a single optimistic accuracy estimate because it identifies where new field observations would add the most information.

“For archaeological prospection, this is preferable to reporting a single optimistic accuracy estimate because it identifies where new field observations would add the most information (Roberts et al. 2017).”

• cites Roberts et al. 2017 sourceNot checked — no full text of the source

It may state two separate claims as one.

Exact regeneration of coordinate-dependent spatial products requires authorized access to the restricted archaeological input data and remains subject to heritage-protection considerations and permissions from the relevant custodians.

“Exact regeneration of coordinatedependent spatial products requires authorized access to the restricted archaeological input data and remains subject to heritage-protection considerations and permissions from the relevant custodians.”

It may state two separate claims as one.

The mean spatial ROC-AUC of 0.753 indicates useful ranking capacity, but the large between-fold variation is equally important.

“The mean spatial ROC-AUC of 0.753 indicates useful ranking capacity, but the large between-fold variation is equally important.”

Probably not a substantial claim. It may state two separate claims as one.

Fifth, the workflow separates analytical reproducibility from unrestricted disclosure of sensitive archaeological locations, allowing code, environmental covariates, derived variables and generalized outputs to remain transparent, whereas exact heritage coordinates remain protected.

“Fifth, the workflow separates analytical reproducibility from unrestricted disclosure of sensitive archaeological locations, allowing code, environmental covariates, derived variables and generalized outputs to remain transparent, whereas exact heritage coordinates remain protected.”

Probably not a substantial claim.

These locations are treated as restricted heritage information because unrestricted disclosure could increase risks of looting, disturbance or unauthorized visitation.

“These locations are treated as restricted heritage information because unrestricted disclosure could increase risks of looting, disturbance or unauthorized visitation.”

Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.

Fourth, a 30-m DEM is suitable for valley-scale environmental structure but not for terrace-wall or microrelief detection.

“Fourth, a 30-m DEM is suitable for valley-scale environmental structure but not for terrace-wall or microrelief detection.”

Probably not a substantial claim. It may be too vague to check against evidence.

This cycle also provides a pathway for integrating future stratigraphic, ceramic, archaeobotanical, chronological and high-resolution topographic evidence without overextending conclusions derived from environmental modelling alone.

“This cycle also provides a pathway for integrating future stratigraphic, ceramic, archaeobotanical, chronological and high-resolution topographic evidence without overextending conclusions derived from environmental modelling alone.”

It may state two separate claims as one.

The environmental differentiation identified by the present model should therefore be interpreted as structure within a connected foothill landscape, not as evidence for separate 'Andean' and 'Amazonian' populations.

“The environmental differentiation identified by the present model should therefore be interpreted as structure within a connected foothill landscape, not as evidence for separate 'Andean' and 'Amazonian' populations.”

It may state two separate claims as one.

The mean performance supports use as an exploratory ranking tool, but the weak fold demonstrates that the environmental signature of known terraces is not uniformly transferable throughout the valley.

“The mean performance supports use as an exploratory ranking tool, but the weak fold demonstrates that the environmental signature of known terraces is not uniformly transferable throughout the valley.”

It may state two separate claims as one.

The spatial separation of upper and lower valley systems is compatible with vertical articulation and exchange, but it does not identify the builders as Inca, Cañari, local Amazonian or mixed groups.

“The spatial separation of upper and lower valley systems is compatible with vertical articulation and exchange, but it does not identify the builders as Inca, Cañari, local Amazonian or mixed groups.”