| 17 Jul 2026
The geomorphology of burrowing bioturbation in the Central French Pyrenees
Abstract. Burrowing mammals can alter mountain hillslopes by excavating subsurface voids, producing mounds of loose sediment, and creating local weaknesses in banks and slopes. Their geomorphic role remains poorly documented in alpine and subalpine catchments, where steep slopes, shallow soils, seasonal snowmelt, and grazing can rapidly rework small disturbances. We quantified the geomorphic imprint of marmots, voles, and moles in the eastern Bastan catchment, Central French Pyrenees, using field surveys, burrow and mound measurements, and soil analyses. We recorded 439 vole burrow entrances, 131 marmot burrow entrances, and 357 molehills. Vole activity was especially associated with gully banks, where 274 entrances were counted along 50 surveyed gully segments and local densities reached 3 entrances m−1. Marmot burrows were less frequent but larger: 42 measured internal voids had a mean volume of 4.68 × 10−2 m3 and a total measured volume of 1.97 m3. Marmot mounds averaged 1.95 × 10−2 m3, whereas molehills were smaller (5.26 × 10−3 m3) but more numerous. Soil pH and median grain size did not differ significantly between mounds and adjacent soil. The clearest soil-property effect was the higher bulk density of marmot mounds relative to adjacent soil, suggesting post-excavation compaction. Overall, burrowing mammals in the Bastan create a visible microtopographic and subsurface signature: molehills roughen fine-grained grassland surfaces, marmot burrows generate large local voids and mounds, and vole galleries are concentrated in gully banks where they may favour bank collapse and gully widening.
This paper examines the geomorphic effects of marmots, voles, and moles in the Central French Pyrenees. The comparative approach is interesting, and the field observations provide useful information on burrowing activity in mountain environments. However, several methodological issues remain. My main concerns relate to species identification, excavation-volume estimates, sampling design, and the interpretation of geomorphic processes. Most of these issues directly affect the reliability of the current results and conclusions.
Major comments:
Species identification and attribution of biogenic landforms(Sections 2.2.1, 3.1 and 3.2; Figures 1g, 1j and 1k)
Identification of molehills: Are the mounds shown in Figures 1g, 1j and 1k really molehills? They appear to have an opening near their summit, which raises questions about their origin (molehills usually do not have openings). In addition, molehills usually follow a linear pattern/alignment associated with underground galleries, which does not seem to be the case in Figure 1k. None of these observations rules out mole activity, but they make the identification really uncertain. What criteria were used to distinguish molehills from vole mounds?
Species identification: How were the burrowing mammals identified in the field? The authors mention entrance size, mound morphology, and field observations, but provide no detail. Were animals directly observed? These details are essential because the species-specific results depend on reliable identification, considering also that different vole species might be responsible for what you identified as molehills.
Absence of vole mounds: Why did you not record vole mounds despite the 439 recorded burrow entrances? Voles may produce few visible mounds, depending on the species and burrowing behaviour. However, the contrast with the 357 mounds attributed to moles is striking. Could vole mounds have been removed by runoff, wind, or trampling? If yes, why were molehills preserved? Differences in topographic setting might explain part of this contrast, but this needs to be discussed. The possibility of misidentification should also be considered, as in my previous comments.
Molehill density: The authors reported 326 molehills in only 14 plots, including 56 in a single 100 m² plot. They also report a mean density of 0.34 molehills m⁻² in four activity zones. How were these zones defined? How was this density calculated? Could the observed concentrations include mounds accumulated over several seasons or years? These densities are quite high for moles, but their interpretation depends on mound age and reliable species identification.
Excavation-volume estimates(Sections 2.2.3 and 3.2; Tables 1 and 3)
Marmot burrow volumes: How were the internal volumes of the 42 marmot burrows measured? The authors do not explain how underground dimensions were obtained. Were only visible or accessible sections measured? Were complete galleries explored? This distinction matters because a measured tunnel section does not represent the total excavated volume of a burrow system.
Molehill volumes: Only 14 molehills were measured to estimate the total volume of 357 mounds. How were these 14 selected? Were they representative of the different sizes (for example, based on four different shapes, Table 1) and preservation states? The extrapolation may be reasonable, but its uncertainty needs to be assessed.
Excavated mass: How was the estimated 2.72 × 10³ kg of excavated material calculated? The authors should specify which bulk-density values were used and how the extrapolated mound volumes were included. Measured mound volumes, underground void volumes, and excavated sediment masses should be clearly distinguished as they relate to different things.
Sampling design and spatial representativeness(Sections 2.2.1, 2.2.2 and 3.2)
How were the 50 sampling plots selected? Along walking transects, but the selection procedure remains unclear. Were they systematically spaced or selected according to accessibility, field conditions, or burrowing activity? A map showing the plots (especially since you have GPS-recorded sampling plots), gully transects, and main geomorphic units would help. Also, some segments of the sampled gullies appear to have been defined according to the presence or absence of burrows. Could this have influenced the reported densities? Please provide more details and complete data.
Soil properties and statistical analyses(Section 3.3; Table 4; Figure 3)
Some interpretations of the grain-size results appear inconsistent with Table 4. For example, the text indicates that marmot mounds contain less coarse sand than adjacent soils. However, Table 4 reports 32.42% in mounds compared with 23.25% in adjacent soils. Similarly, the text suggests that molehills contain less fine sand, whereas Table 4 reports 12.44% in molehills and 8.17% in adjacent soils. These discrepancies need to be checked.
The statistics also require clarification. Several paired t-tests were performed on related soil variables, including grain-size fractions. Were multiple comparisons considered? Please report the sample sizes for each test and, where possible, effect sizes and confidence intervals. Finally, the higher bulk density of marmot mounds is attributed mainly to compaction. Could differences in organic matter, root content, or excavated material also explain this result?
Observed features vs inferred geomorphic processes(Sections 3.4–3.6 and Conclusions)
Several geomorphic processes are discussed without being measured. For example, the authors suggest that vole galleries contribute to gully-bank collapse. The only presence of burrows in unstable banks does not demonstrate that burrowing caused the instability. Could erosion have exposed older galleries?
The same concern applies to infiltration, piping, sediment transport, and sediment connectivity. These are plausible mechanisms, but they were not measured. I suggest distinguishing more clearly between what was observed and measured, what can reasonably be inferred, and what remains hypothetical.
Temporal interpretation of burrowing activity(Sections 3.2, 3.5 and Conclusions)
The field survey was conducted in July 2024, but the observed landforms likely represent different periods of activity. Some mounds are fresh, while others are compacted, eroded, or partly vegetated. How can differences in burrowing intensity be assessed without knowing when these features were produced? The number of visible mounds or entrances reflects accumulated and preserved disturbance, not necessarily current excavation activity. The authors acknowledge this limitation, but it should be more clearly reflected in the interpretation of species-specific geomorphic impacts.
Minor comments:
Is Figure 2 drawn to scale? The relative dimensions of the burrows, galleries, and mounds are unclear. If it is conceptual, this should be stated in the caption and adding dimensions would be helpful. I would also recommend including a map showing the locations of the 50 plots and gully transects. Since these locations were recorded with GPS, such a map should be feasible and would considerably improve the presentation of the sampling design.
Manuscript organization: sections 3.4 to 3.6 are more like a discussion than results. I suggest separating the Results and Discussion sections. This would make it easier to distinguish field observations from their interpretation and comparison with other studies.
Figures: some photographs are too small to assess the geomorphic features clearly, and Figs. 1g, 1j, and 1k are very important because they illustrate the mounds attributed to moles. Figure 3 could also better illustrate the paired differences between mound and adjacent-soil samples.
Some references do not appear to support the statements for which they are cited. For example, Wijnhoven et al. (2007) focus on heavy-metal concentrations in small mammals rather than excavation volumes or sediment redistribution.