Settling moths as potential pollinators of Uncaria rhynchophylla ( Rubiaceae )

Nocturnal pollinators such as moths have received less attention than diurnal insects. To elucidate whether nocturnal moths are important pollinators, we observed both the diurnal and nocturnal visitors to the fl owers of Uncaria rhynchophylla (Rubiaceae) in a warm-temperate forest in central Japan. The diurnal visitors included various taxonomic groups (e.g., bees, hoverfl ies and butterfl ies). The nocturnal visitors were exclusively moths (Geometridae, Erebidae, Noctuidae and Crambidae). Pollen grains of U. rhynchophylla were attached to both diurnal and nocturnal visitors. Although diurnal fl ower visitors carried pollen grains of other plant species, nocturnal moths did not carry heterospecifi c pollen grains. These results suggest that nocturnal moths, as well as diurnal insects, are important pollinators of U. rhynchophylla.

Uncaria rhynchophylla (Miq.)Miq.(Rubiaceae) is a vine found in Japan and China (Fig. 1A; Ridsdale, 1978).This vine has spherical infl orescences (20-25 mm in diameter; Ridsdale, 1978), each of which is composed of many yellow fl owers (Fig. 1B).Previous research reports that U. rhynchophylla fl owers in a warm-temperate forest on the island of Yakushima in southwestern Japan are visited by diurnal insects such as small bees, vespid wasps and hoverfl ies (Yumoto, 1987).However, the pollen attached to these fl ower visitors was not investigated and the observations made only during daytime (Yumoto, 1987).
In Borneo, fl owers of another Uncaria species, U. longifl ora (Poir.)Merr., are reportedly visited by many kinds of diurnal insects, such as Hymenoptera and Diptera (Momose et al., 1998).Flower-visitors belonging to seven insect groups (butterfl ies, bees, hoverfl ies, fl ies, wasps, beetles and moths) were recorded.Uncaria rhynchophylla blooming along the survey route were the survey route were 343 on 21 June, 465 on 23 June, and 286 on 26 June.Because a previous study indicates that very few insects visit the fl owers of U. rhynchophylla (Yumoto, 1987), it was necessary to observe many infl orescences in order to record most of the potential pollinators of U. rhynchophylla.
During the observations, we randomly sampled fl ower visitors using insect nets (pole length: 1.2-8.0m), to examine the pollen grains attached to them and measure the lengths of their proboscis.In the laboratory, we measured the proboscis length of fl ower visitors to the nearest 0.1 mm using slide callipers or with an ocular micrometer under a stereomicroscope.Following the categorisation of Okamoto et al. (2008), we quantifi ed the amount of U. rhynchophylla pollen grains on each fl ower visitor under a stereomicroscope; i.e., the numbers of pollen grains were categorised as 0, 1-10, 11-100 and > 100.We also recorded the presence or absence of heterospecifi c pollen grains on each visitor.Uncaria rhynchophylla pollen grains could easily be distinguished from those of other species of plants (L.japonicum and M. japonicus) based on their morphological characters (e.g., size) under a stereomicroscope.
To determine the fl oral morphology of U. rhynchophylla we randomly sampled nine fl owers collected from nine infl orescenc-es.The lengths of the following two fl oral parts were measured to the nearest 0.1 mm using slide callipers: the distances from the fl oral entrance to the stylar head (x) and fl oral base (y) (Fig. 1C).
To assess the amount of nectar available during the day and at night, the amount of nectar in corolla tubes was quantifi ed using 1-μl microcapillary tubes (Minicaps; Hirschmann, Heidelberg, Germany).The volume of the nectar sampled was determined by measuring the length of the column of nectar in the capillary tubes and calculating it in terms of the proportion of the entire length of the capillary tube (Morrant et al., 2009).We randomly sampled nine fl owers collected from nine infl orescences.Nectar sampling was done during daytime (12:00) and at night (20:30) on 26 June 2016.Because it was diffi cult to insert microcapillary tubes into the narrow corolla tubes they were cut open to collect the nectar.Nectar concentration was measured using a refractometer (Eclipse 45-81: 0-50%; Bellingham and Stanley, Kent, UK).As the amount of nectar was too small to directly measure nectar concentration in several nectar samples, one microlitre of water was added to these samples.
Specimen photographs were taken using a Canon EOS 70D camera with a Canon MP-E 65 mm 1-5× macro lens (Canon Inc, Tokyo, Japan).The software Zerene Stacker v.1.04was used to stack the images.

RESULTS AND DISCUSSION
Both diurnal and nocturnal insects visited U. rhynchophylla infl orescences (Table 1).During the daytime, butterfl ies, bees, hoverfl ies, fl ies, wasps and beetles were observed visiting the infl orescences (Table 1).Among the diurnal visitors, bees (mainly male Bombus ardens ardens Smith) were the most abundant visitors (Tables 1, 2).Settling moths of four families (Geometridae, Erebidae, Noctuidae and Crambidae) visited the fl owers at night (Fig. 1E; Tables 1, 2).Pseudocollix hyperythra catalalia (Geometridae) was the most abundant species (Table 2).Although a greater diversity of groups were recorded as fl ower visitors dur- ing daytime than at night, nocturnal insects (i.e., moths) more frequently visited U. rhynchophylla fl owers than diurnal insects (Table 1).Previous studies have shown that the frequency with which the fl owers of some species of plants are visited by moths is highest a few hours just after sunset (Sugiura & Yamazaki, 2005;Okamoto et al., 2008).Because our night-time observations were limited to a 2-h period just after sunset, the high visitation frequency of moths recorded in this study might be overestimated.Both diurnal and nocturnal visitors carried pollen grains of U. rhynchophylla (Table 2).More than 100 pollen grains were found on each individual of the butterfl ies, hoverfl ies and bees sampled (Table 2).Pollen grains were attached to various parts of the bodies (e.g., legs and proboscises) of these diurnal insects.However, no pollen grains were found on the beetles Gametis jucunda (Faldermann) and Cetonia roelofsi roelofsi Harold (Table 2).Pollen grains were also found on 96.2% of the moths.A total of 54.0% of the moths carried > 100 pollen grains (Fig. 1E; Table 2).The body parts of the moths to which the pollen was attached were antennae, proboscis, compound eyes and forelegs (Fig. 1E).Thus, all fl ower visitors (except beetles) are potential pollinators of U. rhynchophylla.Although hoverfl ies and bees carried pollen grains of other plant species (e.g., L. japonicum and M. japonicas), moths and butterfl ies carried no heterospecifi c pollen grains (Table 2).Because interspecifi c pollen transfer has been known to decrease plant fi tness by clogging the recipient stigma with heterospecifi c pollen (Brown & Mitchell, 2001;Morales & Traveset, 2008), moths could be more effective pollinators of U. rhynchophylla than diurnal visitors that carry heterospecifi c pollen grains.
The mean (± SE) length of corolla tubes was 7.5 ± 0.1 mm (y; Fig. 1C) and style exsertion was 4.4 ± 0.0 mm (x; Fig. 1C).Nectar was detected in 44.4% (n = 9) and 22.2% (n = 9) of the fl owers during the day and at night, respectively.Among the fl owers containing nectar, the mean (± SE) volume of nectar per fl ower was 0.13 ± 0.03 μl (n = 4) during the day and 0.07 ± 0.01 μl (n = 2) at night.The mean (± SE) sugar concentration of fl oral nectar was 22.6 ± 7.2% (n = 3) during the day and 7.7 ± 0.8% (n = 2) at night.Thus, U. rhynchophylla fl owers contained more abundant and concentrated nectar during the day than at night.However, we did not experimentally measure nectar production during the day or night.Further experiments are needed to clarify the factors causing such differences in nectar quantity and quality.
Bumblebees, butterfl ies and some species of moths had proboscises longer (8.7-16.3mm) than the corolla tubes (Table 2).Because U. rhynchophylla fl owers contained nectar during the day and at night, these insects might have used their proboscis to feed on nectar.The nectar of U. rhynchophylla might be preferred by Lepidoptera because bumblebees and lepidopterans are known to prefer sugar concentrations of 30-55% and 15-30%, respectively (Willmer, 2011).The proboscises of hoverfl ies, halictids and a few species of moths were shorter (2.2-7.3 mm) than the corolla tubes (Table 2), indicating that they could not reach the nectar.In some species of plants, however, accumulations of nectar within fl oral tubes is known to increase nectar accessibility for shorttongued visitors (Wasserthal, 1997).This situation might occur in the narrow fl oral tubes of U. rhynchophylla and increase the nectar accessibility for short-tongued insects including P. hyperythra catalalia, the most abundant species.Pollen grains are important food resources for bees and hoverfl ies (Proctor et al., 1996).Therefore, short-tongued bees and hoverfl ies might have visited U. rhynchophylla fl owers to collect pollen grains.
Nocturnal moths and diurnal insects frequently visited fl owers of U. rhynchophylla.Pollen grains of U. rhynchophylla were found attached to both diurnal insects and nocturnal moths.Although diurnal visitors carried pollen grains of other species of plants, nocturnal moths did not.This is the fi rst study to document the importance of moths as visitors of the fl owers of the genus Uncaria.Further experiments are needed to clarify the relative importance of diurnal fl ower visitors and nocturnal Lepidoptera.

Fig. 1 .
Fig. 1.Flowers of Uncaria rhynchophylla and its insect visitors.(A) Shoots and infl orescences.(B) An infl orescence.The white arrow indicates an immature stylar head with pollen.(C) An individual fl ower.The distances from the fl oral entrance to the stylar head (x) and fl oral base (y) measured (see text for details).(D) A geometrid moth, Gandaritis fi xseni, visiting infl orescences.(E) Powder-like pollen grains attached to the antennae, compound eye and proboscis of a geometrid moth, Pseudocollix hyperythra catalalia.

Table 1 .
Frequency of fl ower visitors.