Microclimates affect species assemblages and functions of tropical forest ecosystems (Chen et al., 1999) and microclimatic conditions experienced by several organisms may be quite different from the macroclimatic conditions (De Frenne et al., 2013; von Arx et al., 2013). With global warming predicted to exceed 1.5°C and 2°C in the 21^st century (IPCC, 2021), non-climatic anthropogenic pressures, that is, activities directly resulting in adverse impacts on natural ecosystems, especially forest degradation and fragmentation may worsen microclimatic conditions by reducing the buffering effect (Ewers and Banks-Leite, 2013). Forest degradation has shown to increase the severity of understory microclimatic conditions (Blonder et al., 2018; Jucker et al., 2018). For instance, local air temperatures have been found to be as much as 13.6°C greater in logged forests as compared to primary forests (Senior et al., 2017). Further, the role of intact canopy cover, a feature of old-growth forests (primary forests, henceforth) has a buffering effect on the understory microclimate, offsetting and reducing the severity of macroclimatic variations (De Frenne et al., 2019; Zellweger et al., 2019). Such effects in the canopy microclimate are not as well explored.

Canopies host a high diversity of arthropods (Erwin, 1982; Basset et al., 2008, 2012), particularly beetles (Stork and Grimbacher, 2006). Species unique to the canopy have been observed in the range of 20–30% (Stork and Grimbacher, 2006; Ulyshen and Hanula, 2007; Schroeder et al., 2009) and display very high microhabitat specialization (Wardhaugh et al., 2013). A characteristic feature of tropical rainforests is the presence of vascular epiphytes, which represent approximately 9% of the extant vascular plant diversity (Zotz, 2013). Larger trees, an indicator of older trees, support complex epiphytic communities (Woods et al., 2015; Woods, 2017). Hence, epiphyte communities in old-growth forests are often more diverse than disturbed and secondary forests (Barthlott et al., 2001; Woods and DeWalt, 2013). Consequently, it is no surprise that the presence of epiphytes positively contributes to arthropod diversity in tropical forests (Cruz-Angón et al., 2009; Díaz et al., 2012). The reasoning for the positive association between epiphytes and arthropod abundance and diversity may lie in the fact that epiphytes may offer refuge and resources to arthropods (Nadkarni, 1994). Thus, old-growth trees with greater diversity, abundance, and even size (biomass) of epiphytes may attract a greater abundance and diversity of arthropods. Stuntz et al. (2002a) report substrates with cooler temperatures in the proximity of certain epiphyte species as compared to exposed branch surfaces of host trees, and indicate that this factor, along with the size (biomass) of canopy epiphytes could contribute to the positive association of epiphytes and arthropods (Stuntz et al., 2002b).

Increased forest degradation and fragmentation are well-known to adversely impacted insect communities (Klein, 1989; Feer and Hingrat, 2005; Nichols et al., 2007) through direct exposure to higher ambient temperatures (Piyaphongkul et al., 2012; Woods, 2013; Zhang et al., 2014) and indirectly, such as through changes in host plant qualities. For instance, the leaf surface temperature increases the body temperature of small arthropods on the leaf (Caillon et al., 2014), and higher temperatures can affect several adult phenotypic aspects both within generations and even between generations (Crill et al., 1996). Further, the light intensity has been found to potentially impact the vertical distribution of insects in forests. For instance, Grossner (2009) found light intensity as one of the potential factors affecting the diversity of Heteroptera across vertical strata in beech and oak dominated forests. The understanding of impacts of forest degradation on beetles remains mixed, as some have highlighted the negative effects of fragmentation on beetle species richness and community composition (Feer and Hingrat, 2005; Jung et al., 2018; Salomão et al., 2019), however, some do not demonstrate a clear direction of the impact of fragmentation (Davies and Margules, 1998).

In tropical forests, secondary stands support fewer epiphytes as compared to primary stands (Barthlott et al., 2001; Nadkarni et al., 2004; Padmawathe et al., 2004; Woods and DeWalt, 2013). Forest clearances also adversely affect vascular epiphytes through changes in the microclimate (Zotz and Bader, 2009), such as drier microclimates in secondary forest (Gradstein et al., 2008) or loss of shade-adapted species (Benítez et al., 2012). For instance, Psygmorchis pusilla Dodson and Dressler (Orchidaceae) showed reduced biomass and floral spikes at just 3°C above its optimum temperature (Vaz et al., 2004). Increases in light intensity have shown to reduce relative growth rates of a vascular epiphyte (Laube and Zotz, 2003) as well as photoinhibition (Stancato et al., 2002) which could eventually lead to dieback and reduced numbers in environments with higher light intensity.

The need for prioritizing research on microclimate and its impact on forest biodiversity has been indicated (De Frenne et al., 2021). The importance of the microclimate has been recognized for decades, but our understanding of biotic responses to microclimates in the context of human land-use change is still in its infancy (Santos and Benítez-Malvido, 2012; De Frenne et al., 2021). Further, such work is lacking in Indian rainforests in spite of the existence of two biodiversity hotspots (Myers et al., 2000)—the Western Ghats and the Indo-Burma region. Epiphytes form an important part of the rainforest canopy in the southern Western Ghats including several endemics (Parthasarathy, 1988; Ganesan and Livingstone, 2001). Similarly, beetles also form a major component of the insect taxa (Devy and Davidar, 2003; Mohanraj et al., 2014). The responses of these taxa to environmental change have been little quantified. To provide further insights into the impacts of disturbance on microclimatic variables, beetle and epiphyte communities, we focused on the following two questions:

Throughout this paper, we use the word “canopy” or “forest canopy” referring to the overstorey defined as the “stratum of trees that have outgrown the other vegetation in a forest to have their uppermost crown foliage largely of fully in direct sunlight, usually as a relatively continuous layer (excluding gaps)” (Moffett, 2006). We use the word “crown” or “crown-tree” when referring to the top limbs and leaves of individual trees (from the lowermost limb, excluding the trunk). We chose an evergreen wet tropical forest site, with a known history of varying gradients of logging. Hence, there exist intact, old-growth stands, in close proximity to logged stands. The sites were chosen with the expectation of markedly distinct light and temperature levels in the primary and secondary forests, while the elevation and climatic conditions experienced were the largely very similar. To the best of our knowledge, this is the first study in Indian tropical moist forests to study beetle and vascular epiphyte community responses to microclimatic differences in primary and secondary forest canopies.

We obtained a total of 29,972 (Primary: 14,781, Secondary: 15,191) temperature readings from 36 trees and (Primary: 7,257, Secondary: 8,267) daytime light intensity readings from 35 trees (one HOBO was dislodged from its position from a tree in the disturbed forest, presumably by a Lion-Tailed Macaque). Further, the C. exarillata trees we sampled were larger in the primary forest (mean ± SD, tree DBH: 91.58 cm ± 26.93 cm in the primary forest, 73.96 cm ± 21.62 cm in the secondary forest; mean ± SD, tree height: 22.88 m ± 2.98 m in the primary forest; 18.56 m ± 2.56 m in the secondary forest).

Research on microclimate in ecology has been gaining considerable attention recently (Bramer et al., 2018). Higher temperatures and more intense microclimates in the understorey of logged or secondary forests as compared to primary forests have been well-quantified (Blonder et al., 2018; Jucker et al., 2018; De Frenne et al., 2019). Our work was entirely focused on the canopy and extends earlier findings by showing that the canopy microclimate in selectively logged secondary forests is generally warmer than in primary forests. Our results are consistent with Padmawathe et al. (2004), the only other study focused on Indian forests which reported higher temperature and light intensities at heights of 15–16 m. Sampled trees in the primary forest were surrounded by similar, large neighboring trees. In contrast, sampled trees in the logged sites had a greater inter-tree distance with other trees of comparable size, which is a plausible explanation for our findings.

Our beetle data and findings are consistent with those of Wagner (2000), who also reported significantly higher beetle abundances in a secondary forest from a tropical rainforest site in Uganda with some similarities as our site in terms of forest dynamics and characteristics. Thorat et al. (2016) report higher herbivory levels in the leaves of Macaranga peltata trees in secondary forests in Kalakad (our study region), suggesting higher numbers of insect herbivores such as Aphthona macarangae, a beetle in the Chrysomelidae family. We found eight Chrysomelidae beetles in the secondary forest, whereas only one individual in the primary forest. An explanation for the higher abundance of beetles in the secondary forest could be that more light penetrates the canopy, resulting in higher leaf density and more productivity, which are more palatable to phytophagous beetles (Aide and Zimmerman, 1990). Beetle abundance has also been found to be affected by canopy or crown structure and may be driven by dense and moist foliage (Dial et al., 2006), which is contradictory to our findings. Although we did not account for canopy structure or relative humidity, one would expect higher temperatures and light in secondary forests to result in lower humidity and drier foliage.

Results of the community analysis suggest dissimilar beetle assemblages between the primary and secondary forest types and similar vascular epiphyte assemblages. Wagner (2000) also observed a low degree of overlap in certain beetle families in the canopies of the same tree species between primary and secondary forests. Although we found higher beetle abundances, species richness, and diversity in the secondary forest, the multivariate analysis indicate that the beetle communities are more variable in the primary forests than in the secondary forests. Conversely, the epiphyte community analysis indicated similarity in the communities between both forest types, which was also reported by Barthlott et al. (2001). One reason for the similarity in communities could be because we focused on the same tree species, and host-specificity has been widely observed (Wagner et al., 2015). Studies have reported differences in communities when other host trees were considered, for instance, Woods and DeWalt (2013) report an approximately 40% similarity in epiphyte species composition between a 55-year-old secondary forest (similar age as the secondary forests we sampled) and an old-growth forest.

There is not much work on drivers of beetle communities in the canopy at the microscale, making it difficult to provide insights into drivers of canopy beetle communities. Most of our beetles were small, in the size range of 1–5 mm; Pincebourde et al. (2016) suggest that for small ectotherms, thermal properties of the habitat surface may be drivers of temperature variations experienced by the organism. Disturbances in forests increase the types of microhabitats available, and the more diverse beetle communities may be responding to the vegetation recovery in the under- and midstorey (Perry et al., 2018) rather than the absence of large trees. Further, certain beetles such as dung beetles (which includes Scarabaeidae) may have larger microclimatic variation tolerances than thought (Torppa et al., 2020), and this could extend to other beetle families.

The higher epiphyte abundances and richness were in primary forests were consistent with previous studies (Barthlott et al., 2001; Guzmán-Jacob et al., 2020) and can be explained by the larger and older trees in the primary forest which possess greater microhabitat heterogeneity (Woods et al., 2015). Additionally, the secondary site we chose has seen a higher intensity of logging with higher inter-tree distance (Devy and Davidar, 2001). In India, Padmawathe et al. (2004) observed a strong decline in non-orchid angiosperm epiphyte abundance and richness in secondary forests. However, Seshadri et al. (unpublished) demonstrate a lower epiphyte abundance in primary forest in Kalakad, as compared to selectively logged sites (which were different from the secondary forest plots in this study). Research in the montane tropical rainforests in the Andes has also shown an increase in Orchid abundance and richness and no overall significance in differences in epiphyte abundances and richness in tree canopies with an increase in managed and remnant trees relative to unmanaged mature forest (Larrea and Werner, 2010).

The higher abundances of Bulbophyllum sp.1 in the secondary forest trees indicates that it is a colonizing species, and the “bulbs” are a water-conserving adaptation which could contribute to their establishment in secondary forest trees (Seshadri et al., 2021). Previous explanations for lesser epiphytes in disturbed forests (Barthlott et al., 2001) or differences in communities (Larrea and Werner, 2010) have been attributed to drier and more sunlit microhabitats. Our results suggest that light and temperature may not be the primary factors influencing epiphyte communities. However, the effects of relative humidity are yet to be observed.

Our research design employed a “staggered” deployment of HOBOs, i.e., HOBOs in the primary and secondary forest trees were deployed on different days. This was the most practical way to make the best use of our time, with limited manpower as the process of rigging the tree, accessing the crown, and collecting samples is laborious. An ideal research design would have deployed an equal number of HOBOs and funnel traps in the primary and secondary forest canopies with a complete overlap in the deployment duration. Methods in microclimate ecology vary, as sensors and data loggers used for measurements differ in aspects, making comparisons between datasets difficult (Bramer et al., 2018). We used Onset HOBOs (HOBO pendant logger), a commonly used economical sensor with precisions comparable to more sophisticated sensors (Long et al., 2012). Hence, our method can be easily replicated for spatially (both horizontally and vertically) denser measurements. Future research could deploy sensors at a higher density, with multiple sensors on each tree. Research could also examine coupling between air temperature and leaf surface temperature, and their impacts on arthropods (Pincebourde and Woods, 2012).

Our short-term study provides an insight into the microclimatic differences between primary and secondary tropical wet forest canopies, a first for the southern Western Ghats. Further, we provide an important baseline for future research on beetles in the region. Long-term studies on seasonal fluctuations in beetle communities, comparisons with trends in canopy tree phenology, and the role of crown structure should reveal much more about the ecology of beetles in tropical wet forests.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://github.com/atreebangalore/rubinsagar.

RS and MD: conceptualization, funding acquisition, project administration, and writing—review and editing. RS: formal analysis, investigation, visualization, and writing the original draft. Both authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.