Disturbance-driven changes in tropical dry forests can alter belowground carbon storage and nutrient dynamics, yet the combined effects of seasonality, soil depth, and root traits on fine-root functioning remain poorly understood. We conducted a factorial field experiment across low, medium, and high disturbance levels. Root samples were collected in five replicates during three seasons from two soil depths (0–10 and 10–20 cm). Fine roots were classified by diameter (<1 and 1–5 mm) and status (live or dead). Biomass and nutrients were analyzed using ANOVA and principal component analysis (PCA). Fine-root biomass, carbon, and nitrogen content significantly declined with increasing disturbance, with strongest reductions during winter. Biomass declined from 1.75 g m−2 under low disturbance to 1.10 g m−2 under high disturbance, maximum during the rainy season (2.10 g m−2) and minimum during winter. Carbon content ranged from 250 to 400 mg m−2, whereas nitrogen ranged from 22.5 to 35 mg m−2. Roots <1 mm had higher nutrient concentrations and greater sensitivity to disturbance. Soil depth had a stronger effect on nitrogen than carbon, with higher carbon and nitrogen contents in surface soil. The C:N ratio increased with disturbance in live roots, suggesting reduced nutrient quality under stress, while dead roots showed greater variability and higher C:N ratios. PCA revealed separation of root traits by season and disturbance level, indicating shifts in belowground resource allocation and nutrient status. Overall, increasing disturbance reduced fine-root biomass and nutrient concentrations, potentially weakening belowground carbon sequestration and nutrient cycling in tropical dry forests.

Geophagy at mineral licks is common among wildlife in western Amazonia and is widely interpreted as a behavioural response to chronic sodium deficiency. We conducted the first controlled field experiment on sodium attraction, comparing wildlife visitation at paired salt-enriched and untreated control pits. Five pit pairs were monitored with camera traps over six months at Los Amigos Biological Station, southeastern Peru, alongside two natural mineral licks used as ecological references. Mixed-effects modelling revealed a strong system-wide effect of salt supplementation, corresponding to an approximate twofold increase in visitation relative to control pits. Although estimated effect sizes varied among taxa, support for strong species-specific heterogeneity was limited. Responses were nevertheless broadly consistent across terrestrial herbivores, including rodent and ungulate taxa, with the strongest effects observed in paca (Cuniculus paca), brocket deer (Mazama americana and Passalites nemorivagus), and tapir (Tapirus terrestris). Variation in response strength likely reflects both physiological demand and encounter rates, with larger-bodied species having greater sodium requirements and locally common species more likely to detect these artificial salt licks. Patterns of use at artificial salt licks closely mirrored those observed at natural mineral licks for terrestrial species. Most arboreal mammals, however, were absent from artificial salt licks, suggesting that longer time periods may be required for arboreal taxa to detect and incorporate these sites into their foraging behaviour.

Monitoring mobile, cryptic, and habitat-specialist species in complex mountainous landscapes presents significant challenges. In this survey, we combined passive acoustic monitoring with BirdNET, a free-to-use convolutional neural network algorithm for sound recognition, to describe the species’ vocal activity patterns over a full year. BirdNET showed high precision (0.98) for P. mocinno detections, although recall was moderate (0.38), reflecting a trade-off between maximizing detection and minimizing false positives. We observed a bimodal daily vocal activity pattern, with a pronounced peak at dawn and a smaller peak in the late afternoon. Vocal activity peaked in the early months of the year, particularly in January, coinciding with the breeding season, and showed a tendency to be associated with topographic position (TPI), with higher activity in sheltered terrain such as valleys and depressions rather than across broad elevational gradients. These patterns may reflect underlying differences in habitat use, abundance, or detectability across the landscape. These findings support the use of passive acoustic monitoring (PAM) coupled with deep learning techniques such as BirdNET as an effective approach for monitoring vocal behavior and habitat use in cryptic montane birds, adding to a growing body of evidence highlighting its utility. While further validation remains important, the results provide insights that can help inform future conservation strategies. Overall, this study underscores the potential of bioacoustic approaches to enhance our understanding of species ecology and support evidence-based conservation in tropical montane ecosystems.

Tropical rainforests exhibit some of the highest net primary production (NPP) rates in the world. It is widely hypothesized that soil nutrient availability, particularly phosphorus (P), limits NPP and influences the allocation of productivity between above- and below-ground components in tropical forests. However, these hypotheses have rarely been tested experimentally in extremely wet tropical rainforests. Here, we evaluated the effects of soil fertilization on above- and below-ground NPP in two localities in the Chocó region (Colombia), among the rainiest forests in the world, using five fertilization treatments (N, P, K, NPK, and control) applied in permanent plots. Contrary to expectations, total NPP did not respond to the addition of N, P, or K, despite the very high rainfall and extremely nutrient-poor soils. While total NPP remained unchanged, N fertilization significantly increased canopy NPP, measured as litterfall biomass. In addition, P fertilization induced a redistribution of NPP between above- and below-ground compartments, increasing the relative contribution of fine roots and litterfall while decreasing allocation to woody biomass. These shifts in carbon allocation likely represent an ecological strategy to enhance nutrient acquisition in soils with extremely low phosphorus availability.

Intraspecific trait variation is increasingly recognised as a key component of ecological responses to environmental change in both tropical and subtropical ecosystems, yet sex-specific differences in phenotypic integration remain largely understudied. We investigated sexual dimorphism in the dung beetle Onthophagus atripennis across seven subtropical forest plots in southwestern China, by integrating single-trait analyses with trait network approaches. We measured 18 morphological traits in 299 individuals (150 females and 149 males). At the single-trait level, six morphological traits differed significantly between sexes, with males consistently larger than females; these differences were concentrated in body size and structures associated with digging. Trait–environment associations also diverged between sexes, with females showing more specific responses to herb-layer diversity and litter cover, whereas males exhibited more diffuse responses across altitude, litter thickness, and shrub structure. At the trait network level, although network parameters did not differ significantly between sexes, environmental drivers affected network integration in a sex-specific manner. Females showed increased network integration in response to steeper slopes and thicker litter, whereas males displayed decreased integration at higher elevations. These results indicate that sex differences were primarily reflected in trait values, in trait–environment associations, and in network–environment associations, rather than in intrinsic differences in static network topology. Owing to limited plot-level replication (n = 7 plots available for network analyses), these results should be interpreted as exploratory and hypothesis-generating. Our findings suggest that sex-specific perspectives could improve assessments of functional responses in soil fauna. This is particularly relevant for tropical and subtropical forest restoration efforts.