Alien and native tree species having extrafloral nectaries as favorite hunting area for arboreal endemic Philippine tiger beetles (Coleoptera: Cicindelidae) in human‐disturbed habitat in Lanao del Sur Province, Mindanao, Philippines

Abstract To document a relation between abundance of arboreal, predatory tiger beetles, their ant prey, and extrafloral nectaries attracting the ants, we gathered data from more than 10 species of native and introduced trees and large, tree‐like perennial plants in Lanao del Sur Province, Mindanao, Philippines. All specimens of tiger beetles (two Tricondyla and two Neocollyris species, all endemic to the country) were noted on five tree species characterized by presence of extrafloral nectaries, including three alien/invasive and two native ones. Invasive Spathodea campanulata and native Hibiscus tiliaceus were the most inhabited ones (respectively, 56% and 19% of beetles). Presence of tiger beetles on these trees most probably depends on high abundance of ants, which are typical prey for arboreal Cicindelidae, while occurrence of ants can result from presence of extrafloral nectaries on different parts of the plants. This suggests a new mutualistic insect–plant interaction between native and invasive species.

It is estimated that extrafloral nectaries occur in 1%-2% of terrestrial plant species. These highly diversified structures, located at almost every above-ground organ of plants, produce aqueous solution of carbohydrates and other compounds, similar to "classic" nectar produced in flowers (Weber & Keeler, 2013). As different species of ants (Formicidae) are the most common insects utilizing this source of food, it is generally accepted that occurrence of extrafloral nectaries is evidence of coevolution between plant and ants.
There are three main models of this coevolution: protection, flower distraction, and ant distraction. Probably, first of them is the most common; in this model, the ants protect plant against herbivores, obtaining in return "reward" from the plant. Two other models assume that plant with help of extrafloral nectaries distracts ants from flowers, where they may affect pollination, or from their herbivorous trophobionts (Del-Claro et al., 2016).

| S TUDY ARE A , MATERIAL , AND ME THODS
Adult tiger beetles were collected through opportunistic sampling from tree trunks, branches, and leaves using entomological hand net from 9 April to 4 August 2019 in the municipality of Masiu, Saguiran, and Marawi City, Lanao del Sur Province, Mindanao, Philippines. A total of 10 sampling sites were established in these areas (Table 1). Some sampling sites were visited several times, while others surveyed once due to time constraints and availability. During every visit in every sampling site, all specimens of every tree species present in particular areas were checked (Table 1). Sampling sites were located in human-disturbed habitat close to human settlements or even in the villages. In total, over 10 tree and large, tree-like perennial species have been checked, including three taxa alien and/or invasive for the Philippines ( and Bambusa spp., have been checked as well. According to literature (Elias & Prance, 1978;Feinsinger & Swarm, 1978;Peng, 2015;Savage et al., 2009;Seibert, 1948;Sugiura et al., 2006;Zhang et al., 2012;Zimmermann, 1932), all invasive species and first two of the mentioned native taxa can be characterized by presence of extrafloral nectaries.
Insects were collected usually between 7 a.m. and noon and from 1 p.m. to 5 p.m. as it is the time of highest activity of adult tiger beetles (Pearson & Vogler, 2001). Material was preserved in 96% ethanol and is currently deposited in collections of the first and last author.
Numbers of tiger beetles noted on alien/invasive and native tree species were tested for statistical significant differences applying Mann-Whitney U tests, while in case of tree preferences of Tricondyla species, chi-square tests were conducted using TIBCO Statistica v. 13.3 software.

| RE SULTS
Tiger beetles were found only on plants belonging to five species In total, 78% recorded tiger beetles were collected from alien/ invasive tree species while only 22% from native ones. All beetles were noted on tree species characterized by presence of extrafloral nectaries on their different parts (91% if Mangifera indica is excluded as in case of this species presence of such nectaries is sometimes questioned). Invasive Spathodea campanulata was the most preferred tree with a total of 56% of Cicindelidae (128 individuals) noted in the study, including, respectively, 125 specimens of Tricondyla elongata (65% individuals of the species), two individuals of T. cavifrons (7% of the species), and a single Neocollyris cf. albitarsis (25% of the species). Native Hibiscus tiliaceus and Erythrina fusca were recorded as the second and the third preferred tree by tiger beetles (respectively, 19% and 13% of all collected specimens) followed by alien taxa Mangifera indica L. (9%), and Acacia mangium (4%). The result of Mann-Whitney U tests shown nonsignificant differences between number of tiger beetles recorded on alien/invasive and native trees. Figure 2 shows that Tricondyla elongata was observed at five tree species; however, Spathodea campanulata and Hibiscus tiliaceus seemed to be the favorable for this species (respectively, 65%

F I G U R E 2 Tree preferences in
Tricondyla tiger beetle species in studied habitats in Lanao sel Sur Province, Mindanao, Philippines as for now there is a lack of such data for Philippine arboreal tiger beetles, similar results are obtained in case of epigeic species studied in this country (Jaskuła et al., in prep.). Moreover, there are numerous published examples of negative impact of alien/invasive plant species on animal/insect diversity in areas where they were introduced (e.g., Dostál et al., 2013;Harvey & Fortuna, 2012), even if gradually some of insect species start to use alien plants as food or habitat (e.g., Brändle et al., 2008;Karolewski et al., 2014;Meijer et al., 2012;Novotny et al., 2003). Taking all these facts into consideration, one can note that high abundance of endemic predatory tiger beetles (Cabras et al., 2016) occurring not only on native but also on invasive/alien tree species in Lanao del Sur is unusual and surprising, until we will pay attention on biology and ecology of both arboreal Cicindelidae and tree species on which these beetles were recorded in current study. Both Tricondyla and Neocollyris tiger beetles are known as day active predators hunting on trees and bushes on different small insects, with preference of ants (Clausen, 1940;Naviaux, 2002;Trautner & Schawaller, 1996; R. Jaskuła-personal observation). On the other hand, most (all?) tree species which were recorded by us as attractive for arboreal tiger beetles can be characterized by presence of extrafloral nectaries on different parts of the plants (Weber et al., 2020). For example, Spathodea campanulata has nectaries on fruits, sides of midrib, surface of the leaves, flower pedicels, and external surface of the petals (Elias & Prance, 1978;Seibert, 1948;Zimmermann, 1932), and in Hibiscus tiliaceus, they are present on leaf base, main nerve, and sepals (Savage et al., 2009;Sugiura et al., 2006;Zimmermann, 1932), while Acacia mangium has them on basal parts of leaf stalks (Zhang et al., 2012). In Erythrina fusca, their presence was unspecified (Feinsinger & Swarm, 1978;Savage et al., 2009) but other species classified in this genus have external nectaries on stipules and on the sepals (Zimmermann, 1932). In case of Mangifera indica, some sources state that nectaries are present on young leaves, petioles, and developing fruits (Peng, 2015) while other suggest that they are absent in the species (Weber et al., 2020).
Presence of extrafloral nectaries on mentioned above plants was not studied by us in details, but high number of insects, especially different ant species, was noted on all these trees checked in all study sites. Although we are not able to confirm whether these insects Although no details about Formicidae species were provided by the authors, we believe that most probably such example comes also from the paper by Abeywardhana et al. (2019)  presence of alien/invasive tree species in human-changed habitats because of high concentration of their potential prey. As Hibiscus tiliaceus and Erythrina fusca, which are native to the Philippines, were the second and the third tree species according to number of tiger beetles noted in this study, we suppose that presence of these insects on trees having extrafloral nectaries is not fully a novel hunting adaptation (even if never studied before), but can suggest that at least some arboreal tiger beetles can easily adapt to environmental changes like is was observed earlier in other representatives of this family (e.g., Cabrera et al., 2019;Gilbert, 1997;Rewicz & Jaskuła, 2018;Riggins & Hoback, 2005). Moreover, our results clearly suggest that presence of alien/invasive tree species, even if generally it has negative impact on biodiversity (e.g., Dostál et al., 2013;Harvey & Fortuna, 2012), occasionally can support local populations of some native or even endemic taxa, especially in highly human-disturbed habitats.

ACK N OWLED G M ENTS
We would like to thank Andrey Matalin (Russia) and two anonymous reviewers for their comments to the first version of manuscript as well as to Axel Schwerk (Poland) for some comments according to statistics used in the paper. The study was possible due to the Wildlife

CO N FLI C T O F I NTE R E S T
All authors declare no conflict of interest including any financial, personal, or other relationships with other people or organizations within 3 years of beginning the submitted work that could inappropriately influence, or be perceived to influence, their work.