Trends of ECVs and their association with low-clouds

Beyond low-cloud trends, our results also indicate that most TMCFs exhibit shifts in their climate to warmer environments. Such changes are described mainly by increases in the surface temperature (i.e., average, minimum, and maximum), dew point, and pressure. On a decadal basis, the observed increases in average temperature (0.31 K per decade) might suggest that these ecosystems are warming at higher rates when compared with tropical forest regions (0.26 K per decade) (Malhi and Wright, 2004) or global estimations (0.2 K per decade) (Allen et al., 2018). Warming rates could be even stronger than global when TMCFs are detailed at the macro-ecological level as occurs in Neotropic and Indomalayan TMCFs (0.35 and 0.31 K per decade, respectively). However, we acknowledge that the comparison among other studies should be performed with caution, given the differences in methods.
Although it is expected that the warming of TMCFs also leads to increases in evapotranspiration and presumably PET (Still et al., 1999), its association with ΔCF was not as important as VSWC. It may be considered that increases in PET driven by lowland warming correlate with cloud formation upwind (Still et al., 1999). However, we evaluated the ΔPET - ΔCF association as an in-situ factor without considering the potential movement of moist masses. On the other hand, it could be expected that the variability in VSWC would be more consistent with changes in CF, as changes in clouds have the potential to drive mist interception and water availability. Previous studies have shown similar patterns where high soil moisture in lowland and mountainous areas is related to cloud base heights and their regimes (Lawton et al., 2001; Nair et al., 2008; Ray et al., 2006). At some Neotropical TMCFs (e.g., Monteverde, Costa Rica), negative trends in VSWC and precipitation could be tied together with increases in the number of dry days (Pounds et al., 2006, 1999). An increasing number of dry days in a year is likely to reduce annual averages of VSWC; thus, increases in their frequency over time may also lead to VSWC declines. The coherent variation of VSWC with dry days, low-clouds, or precipitation may suggest the former is an indirect indicator of cloud dynamics which could be used to for the local monitoring the of TMCFs forest health. The reliability of VSWC as an indicator of forest health at the TMCF’s should be evaluated in detail by future studies.
Overall, our results appear congruent with the climatic mechanisms associated with changes in clouds at the TMCFs. Our observed increases in surface temperature and its implication in negatives ΔCF could exemplify the warming effects on cloud formation as well as the rising of cloudbanks. Likewise, our observed increases in temperature are likely to lead increases evapotranspiration which may implies decreases in cloud formation. Despite this, there is an important variance that our PLSR model does not explain according to the coefficient of determination. Differences among TMCFs associated with orographic effects, proximity to water bodies, El Niño Southern Oscillation (ENSO) effects, lowlands land use changes can also to contribute to the observed ΔCF (Lawton et al., 2001; Nair et al., 2011; Still et al., 1999). For instance, ENSO effects — that have shown to drive the TMCFs dry seasonal moisture (Anchukaitis and Evans, 2010) — are likely to have a higher impact on the TMCF’s cloudiness that strictly dependent on water masses from the ocean than those that strictly depend on lowland forest conditions. As such, future studies should disentangle which drivers are closely related to local cloud trends.

Broad implications and future directions

The interconnected web of life in TMCFs is intricately tied to cloud formation, making it imperative to address global changes to preserve these vital ecosystems and safeguard their biodiversity and invaluable services. Our results reveal that the fingerprints of global change are already having a profound impact on the cloudiness of TMCFs. Altered cloud patterns are likely to disrupt the delicate balance of these ecosystems, which may lead to significant consequences for biodiversity and ecosystem services. With a changing climate affecting the frequency of clouds and the ecosystem balance, species that rely on constant mist and humidity may need to adapt to warmer conditions or migrate to higher elevations (Feeley et al., 2020, 2011). The previous could be particularly important for species in those sites where declines in low-clouds occur at higher rates. Therefore, it is crucial for countries and conservation agencies to consider the rate of change in cloudiness as a key climatic factor in their preservation efforts and to develop appropriate management strategies. For instance, South American countries such as Colombia, Venezuela, and Bolivia, as well as Central American countries like Honduras and Costa Rica, should recognize the reduction in low-clouds as a significant threat to their TMCFs, given the pronounced declines observed in these regions and the number of sites affected (Fig. 5). However, we need to highlight that our findings are restricted to the evaluated sites described as TMCFs (Aldrich et al., 1997a), which may not represent the much broader TMCF distribution.