Systematic and exhaustive wood dissection: a standardized method for studying saproxylic insect communities in tropical rainforests

Abstract

Research on tropical saproxylic insects (insects associated with dead wood) commonly uses flight interception traps (FITs). However, because these traps are a passive method of collection, the beetle communities captured by FITs may not reflect the microhabitat associations of the beetles. A standardized sampling method is therefore needed that enables the direct affiliation of the saproxylic insects with the deadwood resources. Here we describe a systematic and exhaustive wood dissection (SEWD) method and compare the saproxylic beetle assemblages captured by this method with the communities collected with FITs from the same study plots. Saproxylic beetles were collected from 63 naturally occurring logs and 24 ethanol-baited FITs across eight plots in old-growth and restored forests in Singapore. The FITs captured a higher abundance and number of species per trap compared to per log sampled through SEWD. Interestingly, the communities captured by the two methods differed significantly, with only a 13.4% overlap in species between the two methods. Specifically, the beetle communities collected with SEWD were dominated by predaceous beetles, with Pselaphinae (Staphylinidae) being the most abundant and specious group after excluding the parthenogenetic paedogenic Micromalthus cf. debilis (Family Micromalthidae). Samples collected with FITs were dominated by xylophagous beetles appearing at the pioneer stage of wood decomposition - mainly Scolytinae (Curculionidae), followed by pioneer predaceous beetles associated with Scolytinae. The distinct beetle communities collected by the two collection methods highlights the high complementarity of FITs and SEWD in capturing beetles with different mobilities and microhabitat preferences and suggest that using FITs alone will not capture the majority of tropical saproxylic beetles living in decaying logs, particularly those on the forest floor. Importantly, the more comprehensive SEWD method gives greater fine-scale ecological data, leading to increased specialist species discovery, and enables direct associations between saproxylic communities and the deadwood to be established.

1 Introduction

Saproxylic insects are insects that rely on dead or dying wood, or the associated products from dead wood such as fungi and slime moulds, during a part of their lifecycle (Speight, 1989). Saproxylic insects carry out various ecosystem functions and services. While wood decomposition and nutrient cycling are the most well recognized functions of saproxylic insects (Ulyshen, 2016), many saproxylic insects are also important for pollination (Gutowski, 1990; Sugiura et al., 2021; Bao et al., 2019), top-down control of specific prey taxa (Weslien and Regnander, 1992), as a source of food for other animals (Laurie and Seidensticker, 1977; Bull et al., 1992; Lee et al., 2017), biomimicry inspiration (Zhang et al., 2020), and beneficial symbionts and chemicals (Yang et al., 2015; Díaz-Navarro et al., 2021; Kundungal et al., 2021; Di Liberto et al., 2024). Thus, it is important to conserve healthy communities of saproxylic insects, both to maintain the functionality of the forest ecosystem and to aid the advancement of human technologies.

Despite their importance, saproxylic insects are understudied in the world’s most biodiverse tropical regions (Grove and Stork, 1999; Slade and Ong, 2023). Limited local funding and a lack of scientific training in many tropical nations have impeded ecological research on most tropical insect groups (Ong et al., 2025). As such, tropical saproxylic insect research remains at an early stage of species discovery (Grimm, 2011; Ando et al., 2024), and there is little research that has attempted to study saproxylic insects at a community level (Grove, 2002b; Ramos-Robles et al., 2020; Araújo et al., 2024), with termites being the exception (Eggleton, 2000; Ashton et al., 2019; Bourguignon et al., 2018). Thus, most saproxylic insects are known only from presence-absence records, with their natural histories and ecologies unknown. Furthermore, big and charismatic taxa have been selectively collected and studied (Dunn, 2005; Troudet et al., 2017; Díaz-Calafat et al., 2024), leaving the majority of small yet abundant taxa with little data available. As such, we have a minimal understanding of tropical saproxylic community compositions and the underlying mechanisms that shape such communities.

Deforestation for urban and agricultural expansion continues to be a key driver of biodiversity decline in tropical regions (Grove, 2002a; Brook et al., 2003; Giam, 2017). Saproxylic insects rely on forest habitats to provide suitable deadwood environments, making them one of the most vulnerable groups to land use change (Grove, 2002b; Giam, 2017). While pristine forest habitats are disappearing, secondary and restored forests regenerated from degraded landscapes are becoming increasingly common (Meli et al., 2017). The conservation value of secondary and restored forests as refuges will therefore determine the survival of many tropical saproxylic insects in the future (Grove and Stork, 1999).

Flight interception traps (FITs, also known as window-flight traps) are the most commonly used sampling method for saproxylic insects in tropical regions due to their high degree of standardization and replicability, easy installation and sample retrieval, and good coverage of flying species (Grove, 2002b; Bouget et al., 2008; Weiss et al., 2019). Adding ethanol baits to the FITs is thought to act as a kairomone, signaling the presence of wood decomposition or tree sap fermentation by microbial activities, and can therefore increase the numbers of beetles caught (Jonsell et al., 2003; Bouget et al., 2009). However, FITs are criticized for including transient species with good dispersal abilities, which can lead to misleading associations between the captured saproxylic communities and the local-scale resources available (Saint-Germain et al., 2006; Bouget et al., 2008). Furthermore, FITs do not provide information on the preferred substrates or microhabitats of the captured species (Bouget et al., 2008). In light of the shortcomings of using FITs for saproxylic studies, a systematic survey method which (i) ensures direct affiliation of the sampled saproxylic assemblages with the local dead and dying wood resources, (ii) allows inferences on the microhabitats and ecology of sampled species to be drawn, and (iii) gives a high degree of standardization is needed to complement the use of FITs, especially when studying the hyper- diverse yet understudied tropical saproxylic insects (Grove and Stork, 1999).

Here, we propose a workflow for the systematic collection of saproxylic insects through wood dissection. Unlike in timed wood dissection surveys in which the search efficiency, and hence species coverage, are heavily influenced by the surveyors’ experience (Müller et al., 2008; Bourguignon et al., 2018), our proposed SEWD method is not restricted by search time. The SEWD method has a standardized protocol with randomized selection of log segments within a fixed survey area in the field, followed by a laboratory protocol that ensures complete extraction of all insects from the sampled log segments. The proposed method is similar to the insect collection method used by Ci et al. (2025), but is modified for sampling saproxylic insects in pre-existing deadwood rather than in experimental deadwood fragments. To our knowledge, no studies have compared the capturing efficiencies and species and functional compositions of the samples from wood dissection methods and FITs. We therefore compared the species pool and the capturing efficiency of both the SEWD method and ethanol-baited FITs placed in old-growth and restored forests in Singapore. For the scope of this paper, we focused on saproxylic coleopteran communities as FITs are most effective in capturing Coleoptera (Knuff et al., 2019). Specifically, we hypothesized that:

• As FITs passively trap flying insects and the SEWD method uses active sampling of the logs themselves, we hypothesized that the collection efficacy between the two trapping methods will differ, resulting in differences in saproxylic beetle abundance, species richness, diversity, and composition between the two trapping methods.

• Related to changes in species composition between the two trapping methods, we hypothesized that the functional composition, in terms of feeding guild compositions, of saproxylic beetles captured with SEWD and ethanol-baited FITs will also differ.

• Based on previous studies (Chung et al., 2000; Grove, 2002b), it is hypothesized that old-growth forests will contain a higher diversity of saproxylic beetles, and both the community and feeding guild compositions will differ between old-growth and restored forests.

2 Materials and methods

2.1 Study site

Field measurements and insect collection were carried out from September 2022 to February 2025 in Singapore (1.35° N, 103.82° E). Singapore has a relatively aseasonal tropical climate, in which it remains hot, rainy, and humid all year round, with an annual average rainfall of 2113.3 mm and a daily mean temperature of 25.0 °C (Meteorological Service Singapore, 2025). Eight study plots representing two forest types (restored forest (Figure 1A) and old-growth forest (Figure 1B) were established (Figures 1C, D). These plots were arranged in four clusters, in which each cluster contained one plot representing each forest type. The four site clusters were located in Bukit Timah, Upper Seletar Reservoir Park, Upper Pierce, and Lower Pierce Reservoir. These plots are connected through forested areas of varying qualities, mostly secondary forests, within the Central Catchment Nature Reserve (except for the Bukit Timah cluster which has been separated from the rest of the site clusters by the Bukit Timah Expressway since the 1980s) (National Parks Board (NParks), 2025a). The Central Catchment Nature Reserve is 3440 ha, with over 2000 ha covered by forests (National Parks Board (NParks), 2025a). The Bukit Timah Nature Reserve is 163 ha and harbors the largest continuous stretch of mixed dipterocarp old-growth forest in Singapore (National Parks Board (NParks), 2025a). The restored forest plots are around 30 years old and consist of young forests that have naturally regrown after removal of Dicranopteris Bernh. fern thickets (Shono et al., 2006). The canopy of the restored forests is dominated by pioneer trees such as Macaranga gigantea (Rchb.f. & Zoll.) Müll.Arg., Prunus polystachya (Hook.f.) Kalkman, and Cyrtophyllum fragrans (Roxb.) DC. The old-growth forest plots, in contrast, consist of pristine forest dominated by a mixture of late successional tree species such as members of the Dipterocarpaceae Blume (Yee et al., 2019; see Supplementary Table S1 for more details on the habitat information of each plot).

Figure 1

2.2 Sampling of deadwood parameters

In each of the eight study plots, a 20 x 20 m subplot was placed in a relatively flat area. Since large forest gaps were absent in restored forest plots, large forest gaps were avoided during placement of subplots in all of the eight study plots. All sizable dead wood, both ground (logs) and standing (snags), with diameter > 10cm and length > 60 cm was marked and measured following Bastrup-Birk et al., 2007. For logs, the diameters at both ends and at the mid-length were measured. For snags, the basal diameter was measured just above the buttress roots. If the height exceeded 2.3 m, the upper diameters and heights of the snag were measured using the tangent method (Baker et al., 2007). The volume of the wood was then calculated using the ‘frustum of a cone’ formula following Baker et al. (2007). The decay class of each wood was classified based on the decay class scheme described in Baker et al. (2007).

A 30 cm long section from each of the logs was then removed using a chainsaw or an axe, depending on the hardness of the wood. The position of the segment along the piece of deadwood was randomized with a random number generator. A maximum of ten logs per study plot were sampled due to limited time and manpower. When more than ten sizable logs were present in the plot, selection of logs was stratified by decay class and wood diameter. To do this, logs originating from the same tree were first grouped together, and one log within each group was selected using a random number generator. Then, logs were selected using stratified randomization based on the rough proportions of decay classes within the plot, ensuring that each decay class and diameter size found in each plot was included. The diameters of these 30 cm log segments were measured at their mid-length, and the decay classes of the log segments were determined after examination of the cross-section of the segment. The log segments were then transported back to the lab for subsequent insect collection. For logs bigger than 40 cm in diameter, the log segments were first split into four equal wedges, and only two of the segments (one of the top wedges and the diagonal wedge to it to account for the effects of exposure and direction of incidence) were taken back to the lab. Wood segments were not taken from standing deadwood due to safety concerns, administrative difficulties, and concerns for killing nearby trees during the dead tree felling operation.

2.3 Saproxylic insect collection: SEWD

Saproxylic insect collection using the SEWD method (Figure 2A) was carried out from 21^st Jan 2023 to 4^th Feb 2025. Log segments were brought back to the lab in batches of one to four logs, depending on the sizes of the logs and hence the estimated handling time required for dissecting the logs. Each batch of logs were exhaustively searched through for a maximum of 14 days to ensure most saproxylic insects were still alive within the log segments. Each log segment was dissected to manually collect any visible macroinvertebrates. For each individual that was collected, the microhabitat (e.g. “in bark”, “under bark”, “fibrous white rot”, “pocket white rot”, “brown rot”, “wood mold”, “termitaries”, “formicaries”) in which the individual was collected was recorded. For xylophagous (i.e. wood-feeding) coleopteran larvae, up to five larvae of each morphospecies were reared with the substrates immediately next to where the larvae were collected. Rearing containers were constructed by grinding the substrates into fine sawdust using a pestle and mortar and pressing it into containers to mimic the deadwood environment. The size of the containers and the amount of sawdust used depended on the size of the larvae and the expected length of the larval period. Rearing these larvae to adulthood was necessary for species identification given that the genetic library of tropical beetles in Southeast Asia is inadequate. Collected individuals were stored in 80% ethanol and were frozen at -20 °C until identification.

Figure 2

After manually searching through the log segment, the residual sawdust was packed and transferred to mesh bags in Winkler extractors. A collection jar containing 80% ethanol was placed at the bottom of each Winkler extractor. The sawdust from each log segment was dried within the Winkler extractor for seven days to extract any remaining small and/or slow moving saproxylic insects.

2.4 Saproxylic insect collection: ethanol-baited FITs

In each of the eight plots, three FITs (Figure 2B) were deployed for seven days, during three different months - late February-early March, September, and December 2023 respectively. Prior field trials showed substantial specimen degradation beyond 7 days due to the hot and rainy climate in Singapore. The FITs were built with two 40 x 60 cm transparent panels crossing and interlocking with each other. A conical cap was fitted on top to shade the traps from rain and falling leaves. A funnel connected to a collection bottle was attached at the bottom of the trap. A 300mL solution of salt, water, and detergent mixture was used as the killing agent. Direct use of ethanol as a killing agent was not possible due to the high evaporation rate in the hot tropical environment. Instead, 3mL of 70% ethanol was placed into a ziplock bag containing a piece of cotton pad. A pin was used to create ten tiny holes (~0.5mm) in the ziplock bag to allow slow evaporation of the ethanol bait. The baits were still giving off an ethanol scent during trap retrieval, suggesting adequate bait load. Two preliminary non-baited FITs showed very low beetle capture rates (6 beetles in both traps), and thus the use of ethanol as a general chemical attractant was used to increase the number of saproxylic beetles captured (Bouget et al., 2009). Each FIT was hung at a height of 2 m between two healthy trees. We ensured the FITs were hung at least 1 m away from the nearest tree to minimize the attraction effect of the trees on the beetles being sampled.

2.5 Identification and feeding guild delimitation

The collected samples were sorted to order level and all individuals in the order Coleoptera were identified to at least family level and placed into morphospecies using information and keys from various sources, including Marshall (2018); Slipinski (1990), and Legalov (2018). We followed the definition of ‘saproxylic’ given in Speight (1989) and Graf et al. (2022) in which both obligative and facultative saproxylic species were considered ‘saproxylic’. For functionally diverse families such as Staphylinidae and Curculionidae, morphospecies were identified to at least subfamily level to enhance certainty in their respective feeding guilds. Each morphospecies was assigned into one of the four main saproxylic feeding guilds following Ulyshen and Šobotník (2018): (1) Xylophagous: defined as species that directly consume wood as food source; (2) Predaceous: species that consume living parts of another organism. Ectoparasites and social parasites were also placed in the predaceous category; (3) Mycophagous: species that predominately feed at least in part on fungi or slime mold; (4) Saprophagous: species that consume dead or waste materials of other fauna. The feeding guild of each morphospecies was determined by the feeding habit of the larvae (Müller et al., 2008). Ambrosia fungus-feeding scolytines, such as the abundant Xylosandrus spp. Reitter, 1913, were also classified as xylophagous as the mutualistic ambrosia fungi consume the wood around the tunnels excavated by their scolytine hosts (Kirkendall et al., 2015), and thus the presence of these scolytines functionally causes digestion of woody tissue. Species that were unlikely to be saproxylic, such as folivorous chrysomelids, typical hydrophilids, coprophagous scarabaeines, granivorous curculionids and litter-feeding elaterids (e.g. the locally common Agraeus mannerheimii Candèze, 1857) were excluded from the analysis. Voucher specimens are deposited at the Tropical Ecology and Entomology Lab, Asian School of the Environment, Nanyang Technological University and a reference collection will be housed at the Lee Kong Chian Natural History Museum, Singapore.

2.6 Data analysis

2.6.1 Comparison of collection efficiencies between collection methods

Abundance and species richness were first standardized to numbers collected per sampling effort (i.e. per log segment for SEWD or per trap for FIT). To test the differences in efficiencies in capturing saproxylic beetles using the SEWD and ethanol-baited FITs, Generalized Linear Mixed Models (GLMMs) with negative binomial error distributions were fitted with the lme4 package version 1.1-37 (Bates et al., 2015) using the glmer.nb() function. Abundance and species richness were set as the response variables. A negative binomial error distribution was used as the variances of both the beetle abundance and species richness data were overdispersed (Lindén and Mäntyniemi, 2011). The collection method (categorical: SEWD or FITs), forest type (categorical; old-growth or restored forest), and the interaction between collection method and forest type were set as the explanatory variables. Site clusters (Bukit Timah, Upper Seletar, Upper Pierce and Lower Pierce) were set as a random effect to account for any spatial autocorrelation.

To investigate the sampling completeness of the two collection methods, sample coverage curves were built. The endpoint was set as double the observed number of individuals collected by the collection method with the higher total beetle abundance. The sample coverage estimations were computed every 50 individuals. The observed number of individuals of the collection method with lower total beetle abundance was also calculated so that the sampling completeness of the two collection methods could be compared at the same sampling effort through individual-based rarefaction. Similarly, individual-based rarefactions were performed to compare beetle diversity at the same sampling effort. Hill numbers (q= 0,1 & 2, which are equivalent to species richness, Shannon index and Simpson index respectively) were estimated at the same intervals as in the above sample completeness curves, and at the observed number of individuals of the collection method with the lower total beetle abundance. Both sampling completeness and rarefaction analyses were conducted using the iNEXT() function from the iNEXT package version 3.0.2 (Hsieh et al., 2016).

2.6.2 Comparison of beetle communities captured by the two collection methods

Within each of the eight study plots all the beetles collected by either the SEWD method or the FITs were pooled together to form plot-level beetle communities. To visualize the differences in species communities between the SEWD method and ethanol-baited FITs, a Bray-Curtis distance Non-metric Multidimensional Scaling (NMDS) plot was constructed using the metaMDS() function from the vegan package version 2.6-10 (Oksanen et al., 2025). The number of ordination dimensions k was set at two and was shown to return a good model with stress below 0.1.

To test the distinctiveness of the species communities collected with two methods, a Permutational Analysis of Variance (PERMANOVA) with 999 permutations was run with the adonis2() function from the vegan package version 2.6-10 (Oksanen et al., 2025). In the PERMANOVA model, the Bray-Curtis distance in species communities between the SEWD method and FITs was the response variable, while the collection methods (SEWD or FITs), forest types (old-growth or restored forest), and the interaction between collection method and forest type were the explanatory variables. Morphospecies collected with SEWD whose identities were identified solely based on larval morphologies and were identified to a taxonomic level courser than genus level were excluded from the community analysis as without full identification these larvae could not be compared to the adult beetles collected with FITs. Permutational Multivariate Analysis of Dispersion (PERMDISP) was carried out to test the differences in community dispersion between collection method and forest type. The dispersions between collection method and between forest type in the same Bray-Curtis distance matrix used for the PERMANOVA model were calculated with the betadisper() function from vegan package version 2.6-10 (Oksanen et al., 2025).

To provide more in-depth analysis on the difference in community composition captured by the SEWD method and FITs, the species communities of the top five most specious families collected with SEWD were compared with those collected with FITs. The species overlap between the two collection methods was visualized with Euler diagrams constructed with the euler() function from the eulerr package 7.0.2 (Larsson and Gustafsson, 2018). Out of the top five most specious families, Elateridae Leach, 1815 was excluded from the analysis as the majority of species remained unidentified due to low rearing success or long larval periods.

2.6.3 Comparison of beetle feeding guild compositions between the two collection methods

The differences in feeding guild compositions for beetles collected by the SEWD method and FITs were examined by comparing the total and proportional abundance and species richness in each of the three major feeding guilds. Only the top three most abundant and specious feeding guilds (predaceous beetles, xylophagous beetles, and mycophagous beetles) were included in the analysis as together they made up 99.7% of individuals and 98.2% of species in the samples. The remaining feeding guilds were too rare (10 individuals, < 0.3% of total abundance; 6 species, < 2% of total species richness) to enable any statistically informative analysis. GLMMs fitted with negative binomial error distributions were conducted with the lme4 package version 1.1-37 (Bates et al., 2015). Either the mean abundance or mean species richness of the particular feeding guild collected per log segment for the SEWD method, or per trap for FIT, was set as the response variable, with the collection method, forest type, and the interaction between collection method and forest type set as the explanatory variables. Site was set as a random variable. To assess the relative importance of each of the three major feeding guilds, GLMMs fitted with beta-binomial error distribution were constructed with the glmmTMB package version 1.1.11 using the glmmTMB() function (Brooks et al., 2017). The proportion of either the total abundance or species richness of each of the three major feeding guilds collected per log segment for the SEWD method, or per trap for FIT, was set as the response variable. For predaceous beetles, the collection method, forest type, and the interaction between collection method and forest type were set as the explanatory variables. In the case of xylophagous and mycophagous beetles, zero inflation was detected, since several logs collected with the SEWD method yielded no xylophagous and/or mycophagous beetles, and so the interaction term between collection method and forest type could not be included in the model.

Due to the uncertain status of nativeness and the unusual parthenogenetic paedogenic reproductive system of the most abundant xylophagous species, Micromalthus cf. debilis LeConte, 1878 (Philips, 2001; Pollock and Normark, 2002; Perotti et al., 2016; Barclay and Bouchard, 2023a), additional analyses with the individual count of this species excluded were conducted for analyses that were sensitive to the total abundance of xylophagous beetles or total abundance of all saproxylic beetles.

All analyses were conducted with R version 4.5.0 (R Core Team, 2025) on RStudio version 2025.5.0.496 (Posit team, 2025).

3 Results

3.1 Summary

In total, 2362 saproxylic beetles from 170 morphospecies in at least 21 families were collected from 63 log segments through the SEWD method, while 1030 saproxylic beetles from 176 species in at least 35 families were collected from 24 ethanol-baited FITs (sampling effort = 7 days x 24 traps = 168 trap-days). The most abundant beetle family collected with the SEWD method was Micromalthidae LeConte, 1878 (1476 individuals, 62.5% of the total catch), followed by Staphylinidae Latreille, 1802(282 individuals, 11.9%), Carabidae Latreille, 1802 (145 individuals, 6.14%), Elateridae (121 individuals, 5.12%), Tenebriondae Latreille, 1802 (88 individuals, 3.73%), and Zopheridae Solier, 1834 (61 individuals, 2.58%) (Figures 3A, C). For the FITs, Curculionidae Latreille, 1802 was the most common family (636 individuals, 61.7%), with the majority in the subfamily Scolytinae Latreille, 1804 (611 individuals, 96.1% of curculionids). Laemophloeidae Ganglbauer, 1899 (81 individuals, 7.86%) was the second most common family, followed by Cleridae Latreille, 1802 (51 individuals, 4.95%), Staphylinidae (44 individuals, 4.27%), and Biphyllidae LeConte, 1861 (30 individuals, 2.91%) (Figure 3B). In terms of species richness, the most dominant family collected by the SEWD method was Staphylinidae (69 species), accounting for 40.6% of total species richness. Out of the 69 staphylinid species, 39 species were in the subfamily Pselaphinae Latreille, 1802, accounting for over half (56.5%) of the staphylinid species pool. Tenebrionidae ranked the second most speciose family (28 species, 16.5%), then Elateridae (12 species, 7.06%), Curculionidae (9 species, 5.29%, of which 4 species were Scolytinae), and Eucnemidae Eschscholtz, 1829 (8 species, 4.71%) (Figure 3D). For FITs, Curculionidae (50 species, 28.4%) was the most diverse family, of which 36 species were Scolytinae (72% of the curculionids) (Figure 3E). The second most species rich family collected with FITs was Staphylinidae (27 species, 15.3%), of which only 5 species (18.5% of the staphylinids) were pselaphines. Anthribidae Billberg, 1820 (10 species, 5.68%) was the third most species rich family, followed by Laemophoeidae and Biphyllidae, each with 8 species (4.55%).

Figure 3

3.2 Differences in sampling efficiency between the SEWD method and FITs

A mean of 44.33 ± 7.39 (SE) saproxylic beetles were captured per ethanol-baited FIT, while a mean of 38.10 ± 13.61 (SE) beetles were collected from each log segment through SEWD (Figure 4A). FITs collected a significantly higher abundance and species richness of beetles per trap than that collected per log through SEWD regardless of whether Micromalthus cf. debilis was excluded from the abundance count (Table 1; Figures 4A–C). Significantly higher species richness of saproxylic beetles were collected from restored forests than in old-growth forests (Table 1; Figure 4E), and significantly higher abundances, but only when Micromalthus cf. debilis was excluded (Table 1; Figure 4D). The interaction between collection method and forest type had no statistically significant effect on the abundance and species richness of beetles captured (Table 1). Site as a random effect accounted for 35.5% and 15.2% of variance in beetle abundance with and without Micromalthus cf. debilis, but the site random effect was negligible (< 0.001% of the variance) in explaining beetle species richness (Table 1).

Figure 4

The sampling effort using the SEWD method captured 96.7% (95% confidence interval= 96.2% - 97.2%) of the expected species while that for FITs captured 89.6% (95% CI = 88.3% - 90.9%) of the expected species (Supplementary Figure S1A). When the sampling effort for the SEWD method was rarefacted to 1064 individuals, the sample coverage of the SEWD method decreased slightly to 94.1% (95% CI = 93.5% - 94.6%) but was still significantly higher than the sampling completeness of FITs for the observed sampling effort. When the dominating Micromalthus cf. debilis was excluded, the sample coverage of the SEWD method at 886 individuals dropped to 91.1% (95% CI = 89.6% - 92.6%), and the extrapolated sampling completeness at 1064 individuals was 92.5% (95% CI = 90.7% - 94.4%) (Supplementary Figure S1B).

Results from individual-based rarefaction indicated significantly higher species richness, and Shannon and Simpson diversity for FITs at 1064 individuals than for the SEWD method (Supplementary Figures S2A–C). However, these three diversity indexes showed no significant differences between the two collection methods when the parthenogenetic paedogenic Micromalthus cf. debilis was excluded (Supplementary Figures S2D–F).

3.3 Species overlap between ethanol-baited FITs and SEWD

The species composition collected with the SEWD method was significantly distinct from that collected with ethanol-baited FITs (Figure 5A; PERMANOVA: R² = 0.3, F_{1, 13} = 5.815, p < 0.001). Forest type (R² = 0.03, F_{1, 13} = 0.588, p = 0.894), and the interaction between collection method and forest type (R² = 0.028, F_{1, 13} = 0.524, p = 0.942) had no statistically significant effect on the community compositions being captured. The dispersions of community compositions were greater for the SEWD method than that of FITs (Figure 5A; PERMDISP: F_{1, 14} = 21.979, p < 0.001). Out of the 134 species (captured as adults or successfully reared to adulthood) from the SEWD method, only 18 species (13.4%) were also captured by ethanol-baited FITs (Figure 5B). The most abundant species Micromalthus cf. debilis was exclusively captured by the SEWD method. Only nine out of the 67 species of staphylinids (13.4%) collected with the SEWD method were also captured using FITs (Figure 5C), and only two out of the 39 pselaphine species (5.13%) from the SEWD method were found in FITs. There were no tenebrionid species in common between the two methods (Figure 5D). For Curculionidae, half (4 out of 8 species) of curculionids collected with the SEWD method were also found in FITs (Figure 5E), all of which were Scolytinae.

Figure 5

3.4 Differences in feeding guild compositions

Overall, FITs resulted in significantly higher beetle abundance and species richness for xylophagous and mycophagous feeding guilds than the SEWD method (Table 2; Figures 6B–D, F), but there was no significant difference in the abundance and species richness of predatory beetles between the methods (Table 2; Figures 6A, E). In terms of abundance, xylophagous beetles dominated both the communities captured with SEWD and ethanol-baited FITs (Figures 6C, D, 7A). In total, 1676 xylophagous beetles (71.0% of total saproxylic beetle abundance) were collected with SEWD, while 657 xylophagous beetles (63.8%) were collected with FITs (Figure 7A). However, xylophagous beetles were only the third most common in the species pool collected with the SEWD method if the parthenogenetic paedogenic Micromalthus cf. debilis was excluded, dropping to 202 individuals (22.8% of the remaining 886 saproxylic beetles) (Figure 7B). Instead, predaceous beetles (437 individuals, 49.3%) became the most abundant feeding guild when using the SEWD method after excluding Micromalthus cf. debilis (Figure 7B). Mycophagous beetles were the second most abundant for both methods when Micromalthus was excluded, with the SEWD method capturing 242 mycophagous beetles (27.3%) and FITs capturing 215 mycophagous beetles (20.9%) (Figures 7A, B). Predaceous beetles (155 individuals, 15.0%) were the third most common among the catch from FITs (Figures 7A, B).

Figure 6

Figure 7

Regarding species richness, predaceous beetles (84 species, 49.4% of the 170 species in total) dominated the SEWD species pool and xylophagous beetles (66 species, 37.5% of the 176 species in total) composed the majority of the FIT species pool (Figure 7C). Xylophagous beetles (56 species, 32.9%) ranked second for the SEWD method, while mycophagous beetles (63 species, 35.8%) were the second most speciose for FITs (Figure 7C). Mycophagous beetles (30 species, 17.6%) were the third most specious group in the species pool collected with SEWD (Figure 7C), whereas for the FITs it was the predaceous beetles (46 species, 26.1%; Figure 7C).

Restored forests had significantly higher abundance of mycophagous and xylophagous beetles, and higher species richness of xylophagous beetles than old-growth forests, although the difference in abundance of xylophagous beetles was significant only when Micromalthus cf. debilis was included (Table 2; Figures 6H–J). The interaction between collection method and forest type had no statistically significant effect on the abundance and species richness of all three of the major feeding guilds (Table 2).

Proportionally, the SEWD method collected significantly more predaceous beetles than FITs both in terms of abundance and species richness (Table 3; Figures 8A, D, G). FITs collected proportionally more xylophagous and mycophagous beetles than the SEWD method (Table 3; Figures 8B, C, E, F, H, I). Old-growth forests had a significantly higher proportion of predatory beetles than restored forests, although the differences were significant only when Micromalthus cf. debilis was included (Table 3; Figure 8J). The interaction between collection method and forest type had no significant effects on the proportions of any of the three major feeding guilds in terms of both the abundance and species richness (Table 3).

Figure 8

3.5 Undescribed species, new national records for Singapore, and rediscoveries

Two undescribed species, Echinoaesalus CL sp.1 (Family Lucanidae Latreille, 1804) and Loebliorylon CL sp.1 (Family Cerylonidae Billberg, 1820) were collected in this study - both were collected through the SEWD method (Supplementary Table S2). Three genera/species were newly recorded from Singapore: Litargus elegantulus Háva, 2024 (Family Mycetophagidae Leach, 1815) and Lioglyptoxenus CL sp.1 (Family Scarabaeidae Latreille, 1802) were collected with FITs, and Micromalthus cf. debilis was captured with the SEWD method and represents the first member of the suborder Archostemata Kolbe, 1908 from Singapore (Supplementary Table S2). The type specimen of Stereomera pusilla Arrow, 1905 (Family Scarabaeidae) was collected in Singapore in 1889 and was rediscovered in the type locality through the SEWD method (Supplemntary Supplementary Table S2).

4 Discussion

To facilitate the study of the ecology of tropical saproxylic insects, a standardized collection method is needed which allows direct affiliations between the saproxylic assemblages and the dead and dying wood resources and their specific microhabitats. As predicted, the collection efficiency and thus the community and feeding guild composition differed between the SEWD method and the commonly used flight interception traps (FITs). However, contrary to our hypothesis and previous studies, saproxylic beetle diversity was higher in restored forest sites than old-growth forest sites.

4.1 Effectiveness of sampling methods to assess saproxylic communities across tropical forest ecosystems

Our results show that using different sampling methods yields different saproxylic communities across tropical forest ecosystems. The FITs had a higher capturing efficiency than the SEWD method. However, the SEWD method captured distinct saproxylic communities that showed little overlap with the species pool collected with FITs, regardless of forest type. Although ethanol baited FITs are believed to be effective in capturing a wide range of saproxylic beetles across the spectrum of wood decay stages (Jonsell et al., 2003; Bouget et al., 2009), our findings showed a bias toward pioneer saproxylic beetles such as Scolytinae that prefer dying or fresh deadwood (Kirkendall, 2007; Kirkendall et al., 2015). Moreover, a considerable proportion of other beetles from FITs, such as Leptophloeus spp. (most of the Laemophloeidae from the FIT catch and the second most abundant family from the FITs) and Stigmatium spp. (majority of the Cleridae and the third most abundant family from FITs) are known predators of Scolytinae (Kulhavy and Miller, 1989; Brill et al., 2021), and so may also have a preference for early stage deadwood. The dominance of pioneer taxa in our study agrees with the results of numerous previous studies (Chénier and Philogène, 1989; Bouget et al., 2009; Miller, 2025). However, fresh deadwood is a temporally and spatially sporadic resource (Cheong, 2019). In our plots only a small proportion (3.17%) of course wood debris (CWD) on the forest floor was classified as decay stage I, and most CWD were in the moderate to late stage of decomposition (decay stage III or IV together accounting for 65.1%), and yet the beetles associated with these later stages of decay were notably absent from the FITs.

In contrast, the saproxylic beetle communities collected with the SEWD method were dominated by predatory beetles. Staphylinids (mainly pselaphines) dominated the beetle communities collected through SEWD both in terms of abundance and species richness. Similarly, staphylinids were the most diverse and abundant taxa collected in the cloud forests in Mexico through timed wood dissection (Ramírez-Hernández et al., 2019). Such dominance by predatory beetles might be due to the permeability of the moderate to heavily decomposed logs. With the all-year-round high temperature and high rainfall provided by the aseasonal climate of Singapore, wood decomposition rates are very rapid (Seibold et al., 2021), leading to a steady supply of moderate to heavily decaying logs that are soft and partially disintegrated. This allows predatory beetles that are often not adapted for boring through intact wood to enter the logs from softened areas or pre-exiting cavities and crevices (Ulyshen, 2016). The higher abundances of saprophagous prey, such as mites, collembola, and dipteran larvae in heavily decayed wood may also support more predatory beetles (Hövemeyer and Schauermann, 2003; Skubala and Maslak, 2010; Ulyshen, 2016). In contrast, the rapid wood decomposition rates in the tropics mean that there is little early-stage deadwood on the forest floor. This could explain the low abundance of Scolytinae and Cerambycidae Latreille, 1802 in our SEWD samples. Instead, these early-stage wood decomposers, such as Xylosandrus spp., the most abundant genus collected with FITs (Kirkendall, 2007; Kirkendall et al., 2015), are likely utilizing dying or freshly dead standing deadwood.

Noticeably, many of the pselaphines exhibited myrmecophilous morphological or behavioral traits (Parker, 2016). The abundance and diversity of saproxylic ants is much higher in the tropics than in temperate regions (Kass et al., 2022), and ants are thought to be important in shaping tropical saproxylic ant communities (Ulyshen and Šobotník, 2018). The large numbers of myrmecophilous beetles recorded in the deadwood in Singapore’s forests therefore suggests a strong influence of ants in shaping saproxylic communities in Oriental tropical forests.

The distinct saproxylic beetle communities collected between the FITs and the SEWD method may also suggest a low mobility of many small and specialized tropical saproxylic beetles living in decaying logs. Flightless species, such as the very common Argiloborus spp. Jeannel, 1937 (Family Carabidae), could not be sampled by the FITs. Moreover, the myrmecophilous or termitophilous lifestyles of some species, in which the miniature body sizes and defensive forms have evolved for intruding or assimilating into ant or termite nests, means that they have low dispersal abilities (Parker, 2016). Our findings highlight the prevalence of fossorial species with low mobility on the forest floor of Oriental tropical forests, and thus the need for inclusion of standardized exhaustive searching like the SEWD method proposed here for comprehensive sampling of tropical saproxylic communities.

Tenebrionidae was represented by only two individuals in the FIT species pool compared to 88 individuals from 28 species in the SEWD method. Several temperate studies have found that ethanol-baited flight interception traps are largely ineffective for capturing tenebrionids (Bouget et al., 2009; Quinto et al., 2013, but see Jonsell et al., 2003 for certain fungal fruiting body-eating species that are attracted to ethanol baits). Most adult saproxylic tenebrionids in Singapore either scrape fungi or lichen off the surface of deadwood or feed on hyphae infested decaying wood rather than feeding directly on the fruiting fungi bodies (Barclay and Bouchard, 2023b; Choeng L.F., per. comm.). It is therefore likely that ethanol is ineffective in capturing most local species of tenebrionids. Given the high diversity and abundance of tenebrionids in tropical saproxylic communities (Ando et al., 2024), our results demonstrate the need for SEWD to systematically record fossorial tenebrioinids and provide a complete representation of the tropical saproxylic communities at a local scale.

Given the distinct saproxylic beetle communities collected by ethanol-baited FITs and the SEWD method, our results suggest strong complementarity in the two collection methods for surveying biodiverse tropical saproxylic communities: ethanol-baited FITs are effective in capturing beetles with good flight abilities, especially those beetles that are attracted to dying wood or early stage standing deadwood, whereas the SEWD method was most effective for collection of fossorial and/or slow-moving species residing in decaying logs on the forest floor. Although the SEWD method is labor intensive and destructive, the inclusion of SEWD along with FITs in future tropical saproxylic surveys is recommended to significantly broaden the species coverage and for a better understanding of the community composition of tropical saproxylic beetles and their responses to disturbance.

4.2 The importance of restored forests for the conservation of tropical saproxylic beetles

As deforestation continues to increase, tropical saproxylic insects are increasingly relying on secondary and restored forests for suitable deadwood habitats (Grove, 2002b). The higher abundance and species richness of saproxylic beetles collected from restored forests compared to old-growth forests with both methods suggests a high conservation value of restored forests for tropical saproxylic beetles. The higher diversity of tropical saproxylic beetles in restored forests could be due to the rapid and steady supply of deadwood from fast-growing pioneer trees with high mortality rates that dominate restored forests (Chave et al., 2009; Falconí-López et al., 2024). The frequent and steady supply of deadwood is particularly important under the aseasonal hot and wet climate in the tropics, as the fast deadwood decomposition rates limit the availability of deadwood resources for tropical saproxylic insects (Grove and Stork, 1999; Seibold et al., 2021). However, the apparent higher diversity in restored forests in our study may also be due to the shifting baseline in Singapore’s fragmented and disturbed old-growth forests. Extensive deforestation since the late 19^th century for agricultural and urban expansion has left Singapore with less than 0.3% of old-growth mixed dipterocarp forests (Yee et al., 2011). As a result of fragmentation and a limited supply of large sized deadwood from small old-growth forest fragments, many large saproxylic beetles with long life cycles such as Chalcosoma atlas Linnaeus, 1758 (Dynastinae MacLeay, 1819, Family Scarabaeidae), Serrognathus bucepahlus Perty, 1831 (Family Lucanidae), and many Glena spp. Hulst, 1896 (Family Cerambycidae), are extirpated in Singapore (TKC Leung, pers. comm.; Cheong, 2019). This deforestation history could have selected for species that are resilient to drier and hotter microclimates induced by edge effects, and that rely on smaller sized deadwood. Such species would be more adapted to live in restored forests than in old-growth forests (Leung and Bonebrake, 2021). Nonetheless, our results demonstrate the high importance of restored forests in conserving tropical saproxylic beetles in places with extensive deforestation histories.

4.3 Applications of the fine scale ecological data collected with the SEWD method

4.3.1 New species discoveries

Under the current rapid Anthropocene extinction, species discovery is a race against time, particularly in places with high rates of urban expansion (Brook et al., 2003; Dunn, 2005; Boyle et al., 2024, Boyle et al., 2025). Such urgency is especially applicable to many tropical saproxylic insects that have a strong reliance on forest habitats and very specific microhabitat requirements (Lachat and Müller, 2018). Through the exhaustive search method, many rediscoveries, new local records and species new to science were collected. Many of these species, such as Echinoaesalus CL sp.1 (Family Lucanidae), Stereomera pusilla (Family Scarabaeidae), Micromalthus cf. debilis and many obligatory myrmecophilous pselaphines are known to be extremely hard to collect with other collection methods either due to their low mobilities, larvae-dominated life cycles, or very specific foraging strata (such as in formicaries or termitaries for myrmecophilous or termitophilous beetles) within deadwood (Maruyama and Nomura, 2011; Paulsen, 2018). Importantly, we made the first records of several fossorial species such as Echinoaesalus CL sp.1 and Micromalthus cf. debilis which were widespread in our samples. The SEWD method is therefore important to enable the discovery of many secretive, undescribed species from biodiverse yet understudied tropical rainforests.

4.3.2 Revealing the microhabitat requirements of threatened saproxylic insects

Species-specific conservation strategies targeting saproxylic insects are currently mostly limited to temperate regions (Audisio et al., 2007; Thomaes et al., 2008; Bosso et al., 2018). These conservation strategies are only possible with a good understanding of the natural history and ecological needs of these threatened saproxylic taxa (Audisio et al., 2007; Thomaes, 2009; Bosso et al., 2018). The SEWD method enables the collection of ecological information on the microhabitat requirements of each species collected, such as the rot types of the wood tissues surrounding the specimens, wood diameters, species identities of the tree hosts, and the identities of social hosts. These fine scale ecological data of each sampled species are useful for planning conservation strategies. For example, addition or creation of deadwood environments suitable for certain threatened species, or protection of habitats where biodiverse saproxylic microhabitats are abundant. Such detailed ecological knowledge is necessary to make species-specific conservation in understudied regions like Southeast Asia possible.

4.3.3 The importance of studying interspecific interactions

Under the pressure of deforestation and urbanization, there has been an increasing emphasis on the importance of secondary forests and the restoration of these ecosystems (Dent and Wright, 2009). While numerous forest restoration projects, such as the OneMillionTrees movement in Singapore (National Parks Board (NParks), 2025b) and the United Nations Decade on Ecosystem Restoration projects (United Nations, 2025), have been carried out globally, few of these have studied the effectiveness of these restoration efforts in the recovery of species interactions and network structure (Vander Zanden et al., 2006). More complex and well-connected ecological networks tend to show higher resilience to disturbance (Loreau and De Mazancourt, 2013). Therefore, ecological network analysis that explores shifts in community structures and stabilities and identifies dominant and keystone species is crucial in assessing the effectiveness of restoration efforts (Johnson et al., 2014; Rooney and McCann, 2012). However, ecological information on the majority of tropical saproxylic insects, including inter-specific interactions that shape saproxylic ecological networks is unavailable due to the elusive behaviors of the majority of species and the difficulty in observing most fossorial saproxylic species (Grove and Stork, 1999). The SEWD method allows direct association of species to the host trees and associated fungi and rot types they are collected from, and thus enables the construction of saproxylic interaction networks.

4.3.4 Environmental DNA metabarcoding

Environmental DNA (eDNA) metabarcoding is used for the detection of partially degraded genetic material left from organisms that have been present in the environment of interest (Beng and Corlett, 2020). Due to the high efficiency, relatively low cost, and non-destructive nature of eDNA metabarcoding, this method has been increasingly popular for rapid biodiversity assessment (Beng and Corlett, 2020). In recent years, the use of eDNA metabarcoding has expanded to the study of saproxylic arthropod communities (Winiger et al., 2022; Shuvo et al., 2025). However, eDNA techniques are still under development for saproxylic ecosystems due to difficulties such as rapid DNA degradation, high heterogeneity of microhabitats, and inaccessibility to the interior of deadwood (Winiger et al., 2022; Shuvo et al., 2025). Although destructive sampling is less desirable, building reference collections from specimens collected using the SEWD method allows species to be barcoded along with their ecological and microhabitat information. Eventually, this will allow sawdust collected from the log segments prior to an exhaustive search to be used for eDNA studies.

References

• 1

  AndoK.CheongL. F.HazmiI. R. (2024). Notes on the Tenebrionidae (Coleoptera) in Singapore and its adjacent areas, with ten new species of Cnodalonini. Elytra. -New. Ser.14, 217–238.

  • Google Scholar
  • View reference in article

• 2

  AraújoJ. F.da SilvaP. V.SalomaoR. P.MouraR. C. (2024). Microhabitat and seasonal influence on bess beetle (Coleoptera: Passalidae) assemblage in Altitudinal Atlantic Forest. Austral Entomol.63, 393–405. doi: 10.1111/aen.12706

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AshtonL. A.GriffithsH. M.ParrC. L.EvansT. A.DidhamR. K.HasanF.et al. (2019). Termites mitigate the effects of drought in tropical rainforest. Science363, 174–177. doi: 10.1126/science.aau9565

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  AudisioP.BrustelH.CarpanetoG. M.ColettiG.ManciniE.PiattellaE.et al. (2007). Updating the taxonomy and distribution of the European Osmoderma, and strategies for their conservation (Coleoptera, Scarabaeidae, Cetoniinae). Fragmenta. Entomol.39, 273–290. doi: 10.13133/2284-4880/124

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  BakerT. R.Honorio CoronadoE. N.PhillipsO. L.MartinJ.van der HeijdenG. M.GarciaM.et al. (2007). Low stocks of coarse woody debris in a southwest Amazonian forest. Oecologia152, 495–504. doi: 10.1007/s00442-007-0667-5

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BaoT.WangB.LiJ.DilcherD. (2019). Pollination of cretaceous flowers. Proc. Natl. Acad. Sci.116, 24707–24711. doi: 10.1073/pnas.1916186116

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  BarclayM. V.BouchardP. (2023a). “ Archostemata, archaic beetles,” in Beetles of the World: A Natural History ( Princeton University Press, Oxford, United Kingdom), 84–85.

  • Google Scholar
  • View reference in article

• 8

  BarclayM. V.BouchardP. (2023b). “ Tenebrionidae: stenochiinae, metallic darkling beetles,” in Beetles of the World: A Natural History ( Princeton University Press, Oxford, United Kingdom), 194–195.

  • Google Scholar
  • View reference in article

• 9

  Bastrup-BirkA.NevilleP.ChiriciG.HoustonT. (2007). The Biosoil Forest Biodiversity Field Manual (Hamburg: ICP Forests).

  • Google Scholar
  • View reference in article

• 10

  BatesD.MächlerM.BolkerB.WalkerS. (2015). Fitting linear mixed-effects models using lme4. J. Stat. Softw.67, 1–48. doi: 10.18637/jss.v067.i01

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  BengK. C.CorlettR. T. (2020). Applications of environmental DNA (eDNA) in ecology and conservation: opportunities, challenges and prospects. Biodivers. Conserv.29, 2089–2121. doi: 10.1007/s10531-020-01980-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  BossoL.SmeraldoS.RapuzziP.SamaG.GaronnaA. P.RussoD. (2018). Nature protection areas of Europe are insufficient to preserve the threatened beetle Rosalia alpina (Coleoptera: Cerambycidae): evidence from species distribution models and conservation gap analysis. Ecol. Entomol.43, 192–203. doi: 10.1111/een.12485

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  BougetC.BrustelH.BrinA.ValladaresL. (2009). Evaluation of window flight traps for effectiveness at monitoring dead wood‐associated beetles: the effect of ethanol lure under contrasting environmental conditions. Agric. For. Entomol.11 10, 143–152. doi: 10.1111/j.1461-9563.2008.00400.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  BougetC.BrustelH.BrinA.NoblecourtT. (2008). Sampling saproxylic beetles with window flight traps: methodological insights. Rev. D'écol.10, 21–32. doi: 10.3406/revec.2008.1457

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  BourguignonT.DahlsjöC. A.SalimK. A.EvansT. A. (2018). Termite diversity and species composition in heath forests, mixed dipterocarp forests, and pristine and selectively logged tropical peat swamp forests in Brunei. Insectes. Soc65, 439–444. doi: 10.1007/s00040-018-0630-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  BoyleM. J.BonebrakeT. C.Dias da SilvaK.DongmoM. A.MaChado FrançaF.GregoryN.et al. (2025). Causes and consequences of insect decline in tropical forests. Nat. Rev. Biodiversity.1, 315–331. doi: 10.1038/s44358-025-00038-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  BoyleM. J.SharpA. C.BarclayM. V.ChungA. Y.EwersR. M.de RougemontG.et al. (2024). Tropical beetles more sensitive to impacts are less likely to be known to science. Curr. Biol.34, R770–R771. doi: 10.1016/j.cub.2024.06.059

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  BrillE.FollettP. A.KawabataA. M. (2021). Feeding habits, movement, and reproduction of the predatory flat bark beetles Cathartus quadricollis (Coleoptera: Silvanidae) and Leptophloeus sp. (Coleoptera: Laemophloeidae) in Hawaii coffee and macadamia nut. Int. J. Trop. Insect Sci.41, 285–294. doi: 10.1007/s42690-020-00205-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  BrookB. W.SodhiN. S.NgP. K. (2003). Catastrophic extinctions follow deforestation in Singapore. Nature424, 420–423. doi: 10.1038/nature01795

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  BrooksM. E.KristensenK.Van BenthemK. J.MagnussonA.BergC. W.NielsenA.et al. (2017). glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R. J.9, 378–400. doi: 10.32614/RJ-2017-066

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  BullE. L.BeckwithR. C.HolthausenR. S. (1992). Arthropod diet of pileated woodpeckers in northeastern Oregon. Northwest. Nat.73 (2), 42–45. doi: 10.2307/3536688

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  ChaveJ.CoomesD.JansenS.LewisS. L.SwensonN. G.ZanneA. E. (2009). Towards a worldwide wood economics spectrum. Ecol. Lett.12, 351–366. doi: 10.1111/j.1461-0248.2009.01285.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  ChénierJ. V. R.PhilogèneB. J. R. (1989). Evaluation of three trap designs for the capture of conifer-feeding beetles and other forest Coleoptera. Can. Entomol.121, 159–167. doi: 10.4039/Ent121159-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  CheongL. F. (2019). Estimating saproxylic beetle (Coleoptera: Cerambycidae and Buprestidae) diversity in Bukit Timah Nature Reserve, Singapore, with a methodological and biological review. Gardens’. Bull. Singapore.71, 339–368. doi: 10.26492/gbs71(suppl.1).2019-14

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  ChungA. Y. C.EggletonP.SpeightM. R.HammondP. M.CheyV. K. (2000). The diversity of beetle assemblages in different habitat types in Sabah, Malaysia. Bull. Entomol. Res.90, 475–496. doi: 10.1017/S0007485300000602

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  CiH.GuoC.SaiB. L.TuoB.ZhaoW. W.QinH. T.et al. (2025). The wood economics spectrum modulates the positive effects of termite foraging intensity on deadwood invertebrate diversity. Funct. Ecol.39, 1498–1509. doi: 10.1111/1365-2435.70021

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  DentD. H.WrightS. J. (2009). The future of tropical species in secondary forests: a quantitative review. Biol. Conserv.142, 2833–2843. doi: 10.1016/j.biocon.2009.05.035

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  Díaz-CalafatJ.Jaume-RamisS.SoachaK.ÁlvarezA.PieraJ. (2024). Revealing biases in insect observations: A comparative analysis between academic and citizen science data. PloS One19, e0305757. doi: 10.1371/journal.pone.0305757

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  Díaz-NavarroM.BolívarP.AndrésM. F.Gómez-MuñozM. T.Martínez-DíazR. A.ValcárcelF.et al. (2021). Antiparasitic effects of potentially toxic beetles (Tenebrionidae and Meloidae) from steppe zones. Toxins13, 489. doi: 10.3390/toxins13070489

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  Di LibertoE. A.BattagliaG.PelleritoR.CurcurutoG.DintchevaN. T. (2024). Biodegradation of polystyrene by plastic-eating Tenebrionidae larvae. Polymers16, 1404. doi: 10.3390/polym16101404

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  DunnR. R. (2005). Modern insect extinctions, the neglected majority. Conserv. Biol.19, 1030–1036. doi: 10.1111/j.1523-1739.2005.00078.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  EggletonP. (2000). “ Global patterns of termite diversity,” in Termites: Evolution, Sociality, Symbioses, Ecology. Eds. AbeT.BignellD. B.HigashiM. ( Springer, Dordrecht, Netherlands), 25–51.

  • Google Scholar
  • View reference in article

• 33

  Falconí-LópezA.GrellaN.DonosoD. A.FeldhaarH.TremlettC. J.MüllerJ. (2024). Patterns of deadwood amount and deadwood diversity along a natural forest recovery gradient from agriculture to old-growth lowland tropical forests. Eur. J. For. Res.143, 1321–1332. doi: 10.1007/s10342-024-01671-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  GiamX. (2017). Global biodiversity loss from tropical deforestation. Proc. Natl. Acad. Sci.114, 5775–5777. doi: 10.1073/pnas.1706264114

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  GrafM.SeiboldS.GossnerM. M.HaggeJ.WeißI.BässlerC.et al. (2022). Coverage based diversity estimates of facultative saproxylic species highlight the importance of deadwood for biodiversity. For. Ecol. Manage.517, 120275. doi: 10.1016/j.foreco.2022.120275

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  GrimmR. (2011). New and little known species of Tenebrionidae (Coleoptera) from Borneo (2). Stuttgarter. Beiträge. zur. Naturkunde. A. Neue. Serie.4, 249–257. doi: 10.18476/sbna.v10.a4

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  GroveS. J. (2002a). Saproxylic insect ecology and the sustainable management of forests. Annu. Rev. Ecol. System.33, 1–23. doi: 10.1146/annurev.ecolsys.33.010802.150507

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  GroveS. J. (2002b). The influence of forest management history on the integrity of the saproxylic beetle fauna in an Australian lowland tropical rainforest. Biol. Conserv.104, 149–171. doi: 10.1016/S0006-3207(01)00140-9

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  GroveS. J.StorkN. E. (1999). The conservation of saproxylic insects in tropical forests: a research agenda. J. Insect Conserv.3, 67–74. doi: 10.1023/a:1009616112275

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  GutowskiJ. M. (1990). Pollination of the orchid Dactylorhiza fuchsii by longhorn beetles in primeval forests of Northeastern Poland. Biol. Conserv.51, 287–297. doi: 10.1016/0006-3207(90)90114-5

  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  HövemeyerK.SchauermannJ. (2003). Succession of Diptera on dead beech wood: a 10-year study. Pedobiologia47, 61–75. doi: 10.1078/0031-4056-00170

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  HsiehT. C.MaK. H.ChaoA. (2016). iNEXT: An R package for interpolation and extrapolation of species diversity (Hill numbers). Methods Ecol. Evol.7, 1451–1456. doi: 10.1111/2041-210X.12613

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  JohnsonS.Domínguez-GarcíaV.DonettiL.MunozM. A. (2014). Trophic coherence determines food-web stability. Proc. Natl. Acad. Sci.111, 17923–17928. doi: 10.1073/pnas.1409077111

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  JonsellM.SchroederM.LarssonT. (2003). The saproxylic beetle Bolitophagus reticulatus: its frequency in managed forests, attraction to volatiles and flight period. Ecography26, 421–428. doi: 10.1034/j.1600-0587.2003.03449.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45

  KassJ. M.GuénardB.DudleyK. L.JenkinsC. N.AzumaF.FisherB. L.et al. (2022). The global distribution of known and undiscovered ant biodiversity. Sci. Adv.8, eabp9908. doi: 10.1126/sciadv.abp9908

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46

  KirkendallL. (2007). Ongoing invasions of old-growth tropical forests: establishment of three incestuous beetle species in southern Central America (Curculionidae: Scolytinae). Zootaxa1588, 53–62. doi: 10.11646/zootaxa.1588.1.3

  • CrossRef
  • Google Scholar

• 47

  KirkendallL. R.BiedermannP. H.JordalB. H. (2015). “ Evolution and diversity of bark and ambrosia beetles,” in Bark Beetles ( Academic Press, USA), 85–156. doi: 10.1016/B978-0-12-417156-5.00003-4

  • CrossRef
  • Google Scholar

• 48

  KnuffA. K.WinigerN.KleinA. M.SegelbacherG.StaabM. (2019). Optimizing sampling of flying insects using a modified window trap. Methods Ecol. Evol.10, 1820–1828. doi: 10.1111/2041-210X.13258

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  KulhavyD.MillerM. C. (1989). “ Potential for biological control of Dendroctonus and Ips bark beetles,” in Austin State University Faculty Publications. Texas, USA Ed. StephenF.

  • Google Scholar

• 50

  KundungalH.SynshiangK.DevipriyaS. P. (2021). Biodegradation of polystyrene wastes by a newly reported honey bee pest Uloma sp. larvae: An insight to the ability of polystyrene-fed larvae to complete its life cycle. Environ. Challenges.4, 100083. doi: 10.1016/j.envc.2021.100083

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51

  LachatT.MüllerJ. (2018). “ Importance of primary forests for the conservation of saproxylic insects,” in Saproxylic Insects: Diversity, Ecology and Conservation. Ed. UlyshenM. D. ( Springer International Publishing, Cham, Switzerland), 581–605.

  • Google Scholar

• 52

  LarssonJ.GustafssonP. (2018). “ A case study in fitting area-proportional euler diagrams with ellipses using eulerr”, in: Proceedings of International Workshop on Set Visualization and Reasoning, 2116, 84–91.

  • Google Scholar

• 53

  LaurieA.SeidenstickerJ. (1977). Behavioural ecology of the sloth bear (Melursus ursinus). J. Zool.182, 187–204. doi: 10.4324/9781003234340-10

  • CrossRef
  • Google Scholar

• 54

  LeeR. H.CheungK.FellowesJ. R.GuénardB. (2017). Insights into the Chinese pangolin’s (Manis pentadactyla) diet in a peri-urban habitat: a case study from Hong Kong. Trop. Conserv. Sci.10, 1940082917709648. doi: 10.1177/1940082917709648

  • CrossRef
  • Google Scholar

• 55

  LegalovA. A. (2018). Annotated key to weevils of the world. Part 1. Families Nemonychidae, Anthribidae, Belidae, Ithyceridae, Rhynchitidae, Brachyceridae and Brentidae. Ukrainian. J. Ecol.8, 780–831. doi: 10.15421/2018_280

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  LeungT. K. C.BonebrakeT. C. (2021). Abundance, distribution and substrate association of Hong Kong stag beetles (Coleoptera: Lucanidae) in secondary forests. Insect Conserv. Diversity14, 609–619. doi: 10.1111/icad.12492

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 57

  LindénA.MäntyniemiS. (2011). Using the negative binomial distribution to model overdispersion in ecological count data. Ecology92, 1414–1421. doi: 10.1890/10-1831.1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58

  LoreauM.De MazancourtC. (2013). Biodiversity and ecosystem stability: a synthesis of underlying mechanisms. Ecol. Lett.16, 106–115. doi: 10.1111/ele.12073

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59

  MarshallS. A. (2018). “ Identifying beetles,” in Beetles: The Natural History and Diversity of Coleoptera ( Firefly Book, Ontario, Canada), 721–764.

  • Google Scholar

• 60

  MaruyamaM.NomuraS. (2011). Taxonomy of the genus Stereomera Arrow 1905 (Coleoptea, Scarabaeidae, Aphodiinae), with description of a new species from Peninsular Malaysia. Special. Publ. Japanese. Soc. Scarabaeoidol.1, 171–175. doi: 10.37520/aemnp.2022.010

  • CrossRef
  • Google Scholar

• 61

  MeliP.HollK. D.Rey BenayasJ. M.JonesH. P.JonesP. C.MontoyaD.et al. (2017). A global review of past land use, climate, and active vs. passive restoration effects on forest recovery. PloS One12, e0171368. doi: 10.1371/journal.pone.0171368

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62

  Meteorological Service Singapore (2025). Climate of Singapore. Available online at: https://www.weather.gov.sg/climate-climate-of-Singapore/ (Accessed June 1, 2025).

  • Google Scholar

• 63

  MillerD. R. (2025). Ethanol: dose-dependent flight responses of bark and woodboring beetles, and associated species of Coleoptera. Environ. Entomol.54, 467–479. doi: 10.1093/ee/nvaf032

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  MüllerJ.BußlerH.KneibT. (2008). Saproxylic beetle assemblages related to silvicultural management intensity and stand structures in a beech forest in Southern Germany. J. Insect Conserv.12, 107–124. doi: 10.1007/s10841-006-9065-2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 65

  National Parks Board (NParks) (2025a). About Our Parks, Nature Reserves and Park Connectors. Available online at: https://www.nparks.gov.sg/visit/when-visiting-parks/about-parks-nature-reserves-pcns/nature-park-networks-areas (Accessed September 3, 2025).

  • Google Scholar

• 66

  National Parks Board (NParks) (2025b). Onemilliontrees Movement: 60 Years of Greening Singapore. Available online at: https://www.nparks.gov.sg/treessg/one-million-trees-movement (Accessed September 3, 2025).

  • Google Scholar

• 67

  OksanenJ.SimpsonG.BlanchetF.KindtR.LegendreP.MinchinP.et al. (2025). vegan: community ecology package. doi: 10.32614/CRAN.package.vegan

  • CrossRef
  • Google Scholar

• 68

  OngX. R.TanB.ChangC. H.PuniamoorthyN.SladeE. M. (2025). Identifying the knowledge and capacity gaps in Southeast Asian insect conservation. Ecol. Lett.28, e70038. doi: 10.1111/ele.70038

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 69

  ParkerJ. (2016). Myrmecophily in beetles (Coleoptera): evolutionary patterns and biological mechanisms. Myrmecol. News22, 65–108. doi: 10.4324/9780429325687-21

  • CrossRef
  • Google Scholar

• 70

  PaulsenM. J. (2018). Generic changes in the stag beetle tribe Aesalini (Coleoptera: Lucanidae: Aesalinae) with the description of two new species. Insecta. Mundi.0666, 1–10. doi: 10.5281/zenodo.3710074

  • CrossRef
  • Google Scholar

• 71

  PerottiM. A.YoungD. K.BraigH. R. (2016). The ghost sex-life of the paedogenetic beetle Micromalthus debilis. Sci. Rep.6, 27364. doi: 10.1038/srep27364

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 72

  PhilipsT. K. (2001). A record of Micromalthus debilis (Coleoptera: Micromalthidae) from Central America and a discussion of its distribution. Florida Entomol.84, 159–160. doi: 10.2307/3496680

  • CrossRef
  • Google Scholar

• 73

  PollockD. A.NormarkB. B. (2002). The life cycle of Micromalthus debilis LeConte, (1878) (Coleoptera: Archostemata: Micromalthidae): historical review and evolutionary perspective. J. Zool. Syst. Evol. Res.40, 105–112. doi: 10.1046/j.1439-0469.2002.00183.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 74

  Posit team (2025). Rstudio: Integrated Development Environment for R (Boston, MA: Posit Software, PBC). Available online at: http://www.posit.co/ (Accessed May 4, 2025).

  • Google Scholar

• 75

  QuintoJ.Marcos-GarcíaM.D.L.Á.BrustelH.GalanteE.MicóE. (2013). Effectiveness of three sampling methods to survey saproxylic beetle assemblages in Mediterranean woodland. J. Insect Conserv.17, 765–776. doi: 10.1007/s10841-013-9559-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 76

  Ramírez-HernándezA.Martínez-FalcónA. P.MicóE.AlmendarezS.Reyes-CastilloP.EscobarF. (2019). Diversity and deadwood-based interaction networks of saproxylic beetles in remnants of riparian cloud forest. PloS One14, e0214920. doi: 10.1371/journal.pone.0214920

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 77

  Ramos-RoblesM.Vargas-CardosoO. R.Corona-LópezA. M.Flores-PalaciosA.Toledo-HernándezV. H. (2020). Spatio-temporal variation of Cerambycidae-host tree interaction networks. PloS One15, e0228880. doi: 10.1371/journal.pone.0228880

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 78

  R Core Team (2025). R: A Language and Environment for Statistical Computing (Vienna, Austria: R Foundation for Statistical Computing). Available online at: https://www.R-project.org/ (Accessed May 4, 2025).

  • Google Scholar

• 79

  RooneyN.McCannK. S. (2012). Integrating food web diversity, structure and stability. Trends Ecol. Evol.27, 40–46. doi: 10.1016/j.tree.2011.09.001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 80

  Saint-GermainM.BuddleC. M.DrapeauP. (2006). Sampling saproxylic coleoptera: scale issues and the importance of behavior. Environ. Entomol.35, 478–487. doi: 10.1603/0046-225X-35.2.478

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 81

  SeiboldS.RammerW.HothornT.SeidlR.UlyshenM. D.LorzJ.et al. (2021). The contribution of insects to global forest deadwood decomposition. Nature597, 77–81. doi: 10.1038/s41586-021-03740-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 82

  ShonoK.DaviesS. J.KhengC. Y. (2006). Regeneration of native plant species in restored forests on degraded lands in Singapore. For. Ecol. Manage.237, 574–582. doi: 10.1016/j.foreco.2006.10.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 83

  ShuvoM. J.TabassumT.SegelbacherG. (2025). Optimizing eDNA metabarcoding techniques for assessing arthropod communities in tree‐related microhabitats. Environ. DNA7, e70064. doi: 10.1002/edn3.70064

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 84

  SkubalaP.MaslakM. (2010). “ Succession of oribatid fauna (Acari, Oribatida) in fallen spruce trees: Deadwood promotes species and functional diversity,” in Trends in Acarology. Eds. SabelisM.BruinJ. ( Springer, Dordrecht, Netherlands), 123–128. doi: 10.1007/978-90-481-9837-5_19

  • CrossRef
  • Google Scholar

• 85

  SladeE. M.OngX. R. (2023). The future of tropical insect diversity: strategies to fill data and knowledge gaps. Curr. Opin. Insect Sci.58, 101063. doi: 10.1016/j.cois.2023.101063

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 86

  SlipinskiS. A. (1990). A monograph of the world Cerylonidae (Coleoptera; Cucujoidea). Part 1: Introduction and higher classification. Annali. del. Musei. Civico. di. Storia. Naturale. “Giacomo. Doria”.88, 1–273.

  • Google Scholar

• 87

  SpeightM. C. D. (1989). Saproxylic Invertebrates and Their Conservation (Strasbourg, France: Council of Europe), 79.

  • Google Scholar

• 88

  SugiuraN.MatsumuraS. I.YokotaM. (2021). Beetle pollination of Luisia teres (Orchidaceae) and implications of a geographic divergence in the pollination system. Plant Species. Biol.36, 52–59. doi: 10.1111/1442-1984.12294

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 89

  ThomaesA.KervynT.MaesD. (2008). Applying species distribution modelling for the conservation of the threatened saproxylic stag beetle (Lucanus cervus). Biol. Conserv.141, 1400–1410. doi: 10.1016/j.biocon.2008.03.018

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 90

  TroudetJ.GrandcolasP.BlinA.Vignes-LebbeR.LegendreF. (2017). Taxonomic bias in biodiversity data and societal preferences. Sci. Rep.7, 9132. doi: 10.1038/s41598-017-09084-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 91

  UlyshenM. D. (2016). Wood decomposition as influenced by invertebrates. Biol. Rev.91, 70–85. doi: 10.1111/brv.12158

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 92

  UlyshenM. D.ŠobotníkJ. (2018). “ An introduction to the diversity, ecology, and conservation of saproxylic insects,” in Saproxylic Insects: Diversity, Ecology and Conservation. Ed. UlyshenM. D. ( Springer International Publishing, Cham, Switzerland), 1–51.

  • Google Scholar

• 93

  United Nations (2025). United Nations Decade on Ecosystem Restoration 2021 – 2030. Available online at: https://www.decadeonrestoration.org/ (Accessed June 1, 2025).

  • Google Scholar

• 94

  Vander ZandenM. J.OldenJ. D.GrattonC. (2006). “ Food-web approaches in restoration ecology,” in Foundations of Restoration Ecology. Eds. FalkD. A.PalmerM. A.ZedlerJ. B. ( Island Press, Washington D.C. USA), 165–189.

  • Google Scholar

• 95

  WeissM.DidhamR. K.ProcházkaJ.SchlaghamerskýJ.BassetY.OdegaardF.et al. (2019). Saproxylic beetles in tropical and temperate forests–A standardized comparison of vertical stratification patterns. For. Ecol. Manage.444, 50–58. doi: 10.1016/j.foreco.2019.04.021

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 96

  WeslienJ.RegnanderJ. (1992). The influence of natural enemies on brood production in Ips typographus (Col. Scolytidae) with special reference to egg-laying and predation by Thanasimus formicarius (Col.: Cleridae). Entomophaga37, 333–342. doi: 10.1007/BF02372435

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 97

  WinigerN.SeiboldS.LucekK.MüllerJ.SegelbacherG. (2022). Lost in dead wood? Environmental DNA sequencing from dead wood shows little signs of saproxylic beetles. Environ. DNA4, 654–660. doi: 10.1002/edn3.284

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 98

  YangY.YangJ.WuW. M.ZhaoJ.SongY.GaoL.et al. (2015). Biodegradation and mineralization of polystyrene by plastic-eating mealworms: part 2. Role of gut microorganisms. Environ. Sci. Technol.49, 12087–12093. doi: 10.1021/acs.est.5b02663

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 99

  YeeA. T. K.ChongK. Y.SeahW. W.LuaH. K.YangS. (2019). “ Vegetation of Singapore. Flora of Singapore,” in Flora of Singapore Vol. 1. Eds. MiddletonD. J.Leong-ŠkorničkováJ.LindsayS. ( National Parks Board, Singapore), 47–70.

  • Google Scholar

• 100

  YeeA. T. K.CorlettR. T.LiewS. C.TanH. T. (2011). The vegetation of Singapore—an updated map. Gardens’. Bull. Singapore.63, 205–212.

  • Google Scholar

• 101

  ZhangH.LyK. C. S.LiuX.ChenZ.YanM.WuZ.et al. (2020). Biologically inspired flexible photonic films for efficient passive radiative cooling. Proc. Natl. Acad. Sci.117, 14657–14666. doi: 10.1073/pnas.2001802117

  • Pubmed Abstract
  • CrossRef
  • Google Scholar