THE QUARTERLY REVIEW of Biology

Volume 91, No. 4 December 2016 THE QUARTERLY REVIEW of Biology

PATTERNS AND PROCESSES IN NOCTURNAL AND CREPUSCULAR POLLINATION SERVICES

Renee M. Borges* Centre for Ecological Sciences, Indian Institute of Science Bangalore 560 012 India e-mail: [email protected]

Hema Somanathan Indian Institute of Science Education and Research Thiruvananthapuram Kerala 695016 India e-mail: [email protected]

Almut Kelber Vision Group, Department of Biology, Lund University 22362 Lund, Sweden e-mail: [email protected]

*Corresponding author.

keywords aridity, crepuscular pollination, dawn, dusk, nocturnal pollination network, water stress

abstract Night, dawn, and dusk have abiotic features that differ from the day. Illumination, wind speeds, turbulence, and temperatures are lower while humidity may be higher at night. Nocturnal pollination oc-

curred in 30% of angiosperm families across 68% of orders, 97% of families with C3, two-thirds of fam-

389 390 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

ilies with crassulacean acid metabolism (CAM), and 71% dicot families with C4 photosynthesis. Despite its widespread occurence, nocturnal pollination occurs in more families with xerophytic adaptations than helophytes or mesophytes, suggesting that nocturnal flowering is primarily an adaptation to water stress since flowering is a water-intensive process. We propose the arid or water stress hypothesis for nocturnal flowering suggesting that plants facing water stress in a habitat (e.g., deserts) or a habitat stratum (e.g., upper canopy for epiphytes) gain a selective advantage by nocturnal flowering by reducing water loss through evapotranspiration, leading to larger flowers that provide more nectar or other resources, to support pollinators with higher rewards. Contrary to the wide taxonomic occurrence of nocturnal flowering, few animal taxa serve as nocturnal pollinators. We discuss the sensory and physiological abilities that enable pollinator movement, navigation, and detection of flowers within the nocturnal temporal niche and present a unified framework for investigation of nocturnal flowering and pollination.

Introduction Nocturnal Pollination OLLINATION is a process in which Patterns in the Angiosperms Pa stationary plant exchanges gametes Since there has not been a review of noc- with other plants or with itself through the turnal/crepuscular pollination systems af- action of abiotic agents such as the wind or ter Baker (1961), there is a lot of ground biotic agents that range in size from minute to cover. We followed the classification of gall midges to large primates. Most plants angiosperms prepared by the Angiosperm are metabolically active during the day when Phylogeny Group (APG III; Bremer et al. the important reactions of photosynthesis 2009) and the linear sequence of plant occur. The wind is strongest during the day, families in Haston et al. (2009). We did an and it appears that the majority of pollinat- exhaustive search of Google Scholar using ing animal taxa are also active during the a combination of keywords that included day. If this is the case, then most pollina- plant families, and such keywords as noctur- tion should occur during the day and noc- nal pollination, night, evening, crepuscular, turnal pollination should be the exception. matinal, as well as keywords for known noc- What, then, are the factors that facilitate turnal/crepuscular pollinators such as bats, the origin and maintenance of nocturnal beetles, moths, and nocturnal mammals. For pollination? plant traits that were not considered in the The purpose of this review is severalfold. papers we reviewed, we also consulted Watson First, it will document nocturnal pollination and Dallwitz (1992; http://delta-intkey.com). patterns in the angiosperms. Second, it will We have taken the caveats mentioned in this present features of the physical environ- online resource seriously and have made in- ment that characterize the night (includ- dependent checks in the primary literature ing dawn and dusk). Third, it will elaborate wherever possible. We also consulted Ste- on factors that make nocturnal pollination vens (2001; http://www.mobot.org/mobot an advantage to plants. Fourth, it will docu- /research/apweb) for plant traits and for ment the animal taxa involved in nocturnal determining the family affiliation of all gen- pollination and review the special features era with nocturnal/crepuscular pollination that enable the activity of nocturnal polli- especially since several genera have been re- nators given the physical conditions of the assigned to families based on recent phyloge- night. Finally, it will lay out new hypotheses netic information. to fuel future investigations into this fasci- We score pollination to be truly nocturnal nating field. Considering the vast scope of and/or crepuscular (active during dawn or this subject, this review is intended as the dusk) if there is evidence that pollinators beginning of a synthesis toward a better un- are attracted to the flowers during noctur- derstanding of how and why nocturnally nal and/or crepuscular hours, and if there flowering plants and their nocturnal polli- is reasonable evidence of pollen transfer. nators occupy such a special temporal niche. We do not include some systems, for exam- December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 391 ple, brood site pollination by thrips where only the Acorales and the Petrosaviales did it is clear that the thrips are attracted to not have any families with nocturnal pollina- the flowers in large numbers during the tion, although the other orders had represen- day and they continue to be present in the tation of this trait. In the other “clades,” there flowers during the night. Similarly, we in- is no evidence for nocturnal pollination in clude thermogenic pollination systems as the following orders: Ceratophyllales, Troch- being nocturnal only if there is evidence odendrales, Buxales, Gunnerales, Dilleniales, that the attraction to the heat-producing Vitales, Fagales, Picramniales, and Garryales. flowers is during the crepuscular/noctur- In the following orders, pollination systems nal hours; for example, while Aristolochi- are unknown in some or all families, so the aceae and Rafflesiaceae are thermogenic presence of nocturnal pollination cannot be families (Thien et al. 2009), there is no determined: Berberidopsidales, Huerteales, evidence that the flies are attracted to the Bruniales, and Paracryphiales. Across orders, flowers during the night itself. Similarly, Magnoliales, Pandanales, Arecales, Zingibe- Hydnoraceae includes thermogenic trap rales, Myrtales, Gentianales, and Escalloniales blossoms, but the odor that attracts beetles (represented by only one family) had the is produced during the day and no noctur- highest percentage (at least 60%) of families nal visitors have yet been observed (Bolin with nocturnal pollination (Table 1). It must et al. 2009). Also, in Nelumbonaceae, there be remembered that our scoring of nocturnal is no evidence of nocturnal attraction by pollination is at the family level; some orders beetles, which are the major pollinators, and clades have many more families than oth- although the flowers are thermogenic (Li ers, and this factor must be considered when and Huang 2009). This is in contrast to the evaluating the phylogenetic occurrence of thermogenic Araceae, e.g., Amorphophallus, nocturnal pollination. which begin to emit odor at dusk (Punekar Using Phylomatic (Webb and Donoghue and Kumaran 2010). 2005; http://www.phylodiversity.net/phylo Of the 413 families recognized in the matic), we constructed a phylogeny for the APG III classification, 113 families had spe- 413 APG III angiosperm families and lo- cies with nocturnal pollination, 37 families cated the nocturnal/crepuscular families have unknown pollination mechanisms, and on the obtained phylogenetic tree, which 263 families have no evidence of nocturnal was visualized using the program FigTree pollination (Table 1, see Appendixes 1 and (http://tree.bio.ed.ac.uk/software/figtree/; 2 available at http://www.journals.uchicago Figure 1). The value of good phylogenies is .edu/loi/qrb). Therefore 113/376, or ap- essential to any analysis on functional traits proximately one-third of angiosperm fam- (Hinchliff et al. 2015). As the figure shows, ilies, whose pollination systems are known, nocturnal pollination occurs throughout have species with nocturnal pollination. This the tree of angiosperms and does not appear is likely to be a conservative estimate since to be restricted to any particular groups, most studies on pollination are conducted whether early or more recent. With its wide on day-flowering species. At a higher taxo- distribution, nocturnal pollination does not nomic level, at least 42/61 = 69% of orders appear to be restricted to particular types (with designated families) have species with of flowers. Is it possible then to understand nocturnal pollination (Table 1; excludes all what types of processes may have given rise orders marked with a “?”). This indicates the to nocturnal pollination? We therefore ex- wide taxonomic occurrence of nocturnal amine the abiotic environment that is char- pollination, presumably in some cases, due acteristic of the night since pollinators that to convergent evolution. use this temporal niche must have abilities Examining these trends phylogenetically, to function within this part of the diel cy- Malpighiales had the maximum number of cle, and plants must also have an advantage families with nocturnal pollination (i.e., 10 to open their flowers during nocturnal or out of 35 families; Table 1). In the monocots, crepuscular periods. 392 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

TABLE 1 The frequency of nocturnal pollination across the angiosperm tree of life (the orders and families follow Bremer et al. 2009 and Haston et al. 2009) Families with nocturnal pollination (percent Families Families of families with with without Families with Order nocturnal pollination unknown nocturnal thermogenesis excluding those with pollination pollination* unknown pollination mechanisms mechanisms)

ANITA grade Amborellales 0 (0%) 1 0 0 Nymphaeales 1 (33.33%) 2 0 1 Austrobaileyales 1 (33.33%) 2 0 1 Undeﬁned Chloranthales 0 (0%) 1 0 0 Magnoliids Canellales 0 (0%) 2 0 0 Piperales 1 (20%) 4 0 2 Magnoliales 4 (66.67) 2 0 2 Laurales 3 (50%) 3 1 0 Monocots Acorales 0 (0%) 1 0 0 Alismatales 1 (7.7%) 12 0 1 Petrosaviales 0 (0%) 1 0 0 Dioscoreales 1 (33.33%) 2 0 0 Pandanales 3 (60%) 2 0 1 Liliales 2 (22.22%) 7 1 0 Asparagales 7 (50%) 7 0 0 Monocots Arecales 1 (100%) 0 0 1 (Commelinids) Commelinales 1 (20%) 4 0 0 Zingiberales 5 (62.5%) 3 0 0 Poales 1 (6.25) 15 0 0 Eudicots Ceratophyllales 0 (0%) 1 0 0 Ranunculales 1 (12.5%) 7 0 0 Proteales 1 (33.33%) 2 0 1 Trochodendrales 0 (0%) 1 0 0 Buxales 0 (0%) 2 0 0 Core Eudicots Gunnerales 0 (0%) 2 0 0 Dilleniales? 0 (0%) 1 0 0 (includes Dilleniaceae) Saxifragales 1 (11.11%) 8 5 0 ?(includes 0 (0%) 1 0 0 Cynomoriaceae) Rosids Vitales 0 (0%) 1 0 0 Rosids (Fabids) Zygophyllales 1 (50%) 1 0 0 Fabales 1 (25%) 3 0 0 Rosales 3 (33.33%) 6 0 0 Fagales 0 (0%) 7 ?(includes 0 (0%) 1 0 0 Apodanthaceae) Cucurbitales 1 (16.67%) 5 1 0 Celastrales 1 (50%) 1 0 0 Oxalidales 1 (20%) 4 2 0 Malpighiales 10 (31.25%) 22 3 1

continued December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 393

TABLE 1 Continued Families with nocturnal pollination (percent Families Families of families with with without Families with Order nocturnal pollination unknown nocturnal thermogenesis excluding those with pollination pollination* unknown pollination mechanisms mechanisms)

Rosids (Malvids) Geraniales 1 (33.33%) 2 0 0 Myrtales 6 (75%) 2 1 0 Crossosomatales 2 (40%) 3 2 0 Picramniales 0 (0%) 1 0 0 Sapindales 4 (50%) 4 1 0 Huerteales 0 (0%) 2 1 0 Malvales 4 (40%) 6 0 0 Brassicales 5 (41.67%) 7 5 0 Core Eudicots Berberidopsidales 0 (0%) 1 1 0 Santales 3 (42.85%) 4 0 0 Caryophyllales 5 (16.67%) 25 4 0 Asterids Cornales 1 (16.67%) 5 0 0 Ericales 7 (31.81%) 15 0 0 Asterids Unassigned 0 (0%) 1 2 0 (Lamiids) (includes Oncothecaceae, Metteniusaceae, Icacinaceae) Garryales 0 (0%) 2 0 0 Gentianales 4 (80%) 1 0 0 Unassigned 1 (50%) 1 0 0 (includes Vahliaceae, Boraginaceae) Solanales 2 (50%) 2 1 0 Lamiales 9 (42.86%) 12 2 0 Asterids Aquifoliales 0 (0%) 4 1 0 (Campanulids) Asterales 2 (18.18%) 9 0 0 Escalloniales 1 (100%) 0 0 0 Bruniales 0 (0%) 1 1 0 Paracryphiales 0 (0%) 0 1 0 Dipsacales 1 (50%) 1 0 0 Apiales 2 (33.33%) 4 1 0

113 263 37 11

*This excludes Dasypoganales whose position in the Commelinids appears to be unsure; this family has no nocturnal pollination reported. ?, Undeﬁned, or Unassigned indicates that the taxonomic position is unclear; families are arranged in the linear sequence of Haston et al. (2009).

The Night (Including Dawn The most direct effects of reduced electro- and Dusk) as a Niche Axis magnetic radiation from the sun are lowered During sunrise and sunset, the physical en- light intensity and changes of temperature. vironment changes in many respects that are However, these changes themselves strongly important for plants and their pollinators. inﬂuence several other parameters, including 394 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

Figure 1. Nocturnal/Crepuscular Families (in Bold and Marked with Black Circles Near Their Names) Located on the Apg III Angiosperm Tree of Life See the online edition for a full-size version of this figure. humidity, speed, and turbulence in air mo- ent ways that will be discussed in later section—winds—and this again influences the tions but we will, in this section, concentrate speed, efficiency, and distance over which on the changes themselves. pollen or floral volatiles can be dispersed by abiotic factors. The changes in these parameters are not the same in all habitats; they dif- light and infrared radiation fer between tropical and temperate zones, As the sun sets, the intensity of electromag- between mountains and lowland, between netic radiation falls continuously until the open habitats and forests, and between the sun is 18° below the horizon. This is the same understory and the canopy of a forest. In ad- for visible light (for animal vision, this means dition, moon phase affects light levels. These light between 300 and 800 nm) and infrared changes influence the relationship between or heat radiation (wavelengths up to 300 μm). plants and their pollinators in many differ- Heat radiation influences temperature, but December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 395 temperature changes vary to a large degree that dark, and twilight can last for six hours, with landscape, vegetation, and weather con- while at 70° latitude, the sun does not set ditions, and differ vastly between seasons and for more than two months. Figure 2C shows latitude. These changes are complicated be- the duration of the midsummer night and cause they depend on changes in air move- the different phases of twilight as a func- ments, and will therefore be discussed with tion of latitude. The respective light inten- the other changes of the atmospheric bound- sities are in Figure 2B. The full moon rises ary layer that occur at night. around sunset and sets around sunrise, Between a sunny day and full moon night, and during full moon nights, light levels the light intensity differs by a factor of about never drop below those at nautical twilight one million, albeit with a similar spectrum (when the sun is 12° below the horizon). (Figure 2A,B). On a moonless night, light The half moon rises or sets around midnight intensity provided by stars is dimmer than and reaches only lower light levels, and moonlight by a factor of approximately 100 during the darkest moon phase, the moon is (Figure 2B), and the spectrum is consider- never up during the dark part of the night. ably shifted toward long wavelengths (Fig- The duration of the night therefore dif- ure 2A; Johnsen et al. 2006; Johnsen 2012). fers largely across latitudes while the lunar Twilight is blue-shifted in comparison to di- phases vary similarly at each latitude, factors rect sunlight or moonlight. The changes in that should be relevant in a global under- light color are less obvious than the intensity standing of nocturnal pollination. change (Figure 2B) but may nevertheless be important for the detection of flower colors and pollinator vision under these conditions. the atmospheric boundary layer Finally, both the moon and the sun create at night: wind and temperature a pattern of polarized light scattered from The atmospheric boundary layer is the the sky, and can be used by insects to navi- lowest part of the atmosphere where wind gate (Warrant and Dacke 2011), although no is influenced by surface friction from veg- such pattern is present in a moonless sky. etation and topography, and wind speed In different habitats, and depending on increases with height over ground. The weather conditions, light intensity can change thickness of the boundary layer depends for reasons other than described above. Un- on wind speeds and temperature, and the der the more open canopy of a temperate or structure differs strongly between day and deciduous forest, intensity is reduced only by night (e.g., Grant 1997; Nadeau et al. 2011). a factor of approximately 10, while in some Although the boundary layer is thick and tropical rainforests, the canopy may transmit characterized by convective turbulences as little as 0.1% of the light available above, due to solar heat radiation during the day, it and the wavelength spectrum is largely dom- is more stably stratified, thinner, and solely inated by the green and infrared light that is driven by wind shear at night when the not absorbed by chlorophyll (Endler 1993). Earth’s surface is colder than the air above. This means that pollinators or flowers in The temperature inversion starts close to the the understory at dawn, dusk, or at night ground shortly before sunset, because the would be under more stringent illumination ground cools down, while warm air rises constraints. (Figure 3). As stratification depends on Close to the equator, the dark period temperature, it is more stable in temperate of the day lasts approximately 12 hours all zones where it can even last during day- year, including morning and evening twi- time, although in the tropics, stratification light periods of little more than one hour may be less stable even at night. With a each. The further away from the equator, clear sky, the ground cools down more rap- the influence of the seasons on day–night idly, and inversion is more likely to occur cycles is more pronounced. For example, than with a cloudy sky. Thus, in humid cli- at 50° latitude, the sun does not sink 18° mates, night temperature does not sink as below the horizon for 40 summer days, at much as in dry climates, and temperature 60° latitude, one-third of the year never gets inversion is less common. 396 THE QUARTERLY REVIEW OF BIOLOGY Volume 91 December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 397

Generally, wind speeds at night tend to deserts (Dai 2006). The diurnal humidity cy- be lower than during the day (Figure 3), cle varies considerably depending on terrain and there are two different scenarios: with (e.g., Gebhart et al. 2001; Duane et al. 2008), strongly stable stratification, wind speeds and in heterogeneous structures such as val- above the inversion capping the boundary leys, the characteristics of the boundary layer layer are larger than surface wind speeds, differs from textbook conditions described while with less stable stratification, wind for homogeneous and flat terrain. For in- speeds are distributed in a similar way as stance, in a coastal valley, temperature de- during the day (Monahan et al. 2011). In creases over the late afternoon and transport forested mountainous areas, at night, winds of cold moist air is important for nocturnal above and below the canopy are most often dew deposition (Khodayar et al. 2008). decoupled. The lower canopy flow is most Therefore, temperature, humidity, and often downhill (Sedlák et al. 2010) and wind latitude interact to influence ambient con- speeds are usually lower at night under tropi- ditions in different ways that can affect cal rainforest canopies (McCay 2003). flow ering plants and pollinators. Given the Temperature inversion has several effects. par ticular abiotic features of dawn, dusk, Most importantly, with the stable stratifica- and the night, we now examine why and how tion under an inversion layer, pollutants but plants and pollinators may use the noctur- possibly also odorants released from plants nal niche. are trapped and accumulate more than usual. Although this effect is well known, the changes in the atmospheric boundary layer Physiological Constraints in Plants during the late afternoon and evening, and Leading to Nocturnal Pollination again during the morning are not well un- Despite the fact that nocturnal pollina- derstood (Nadeau et al. 2011), and few stud- tion appears to be widespread, are there ies on the shape of odor plumes at night ecophysiological traits and processes that are available (e.g., Willis 2008; Girling et al. can help to predict its occurrence? We used 2013). BayesTraits V2 (http://www.evolution.rdg.ac .uk/BayesTraits.html) to determine whether plant traits that we examined evolved in- humidity dependently on the phylogenetic tree. We In addition to temperature and winds, hu- used the Discrete module for examining midity changes systematically between day correlated evolution between pairs of these and night, with generally higher relative hu- discrete binary traits such as nocturnal pol- midity at night (Figure 3). As a result of large lination and any other traits. We used Mar- diurnal temperature variations, mean rela- kov Chain Monte Carlo (MCMC) methods tive humidity at night is up to 15% higher for calculating Bayes Factors (BF). A step- than daytime relative humidity over most ping stone sampling method was used to land areas. This effect is weaker in humid estimate marginal likelihoods of the mod- climates and strongest in dry areas such as els following the procedures recommended

Figure 2. Light Conditions During Dusk, Night, and Dawn A. The number of photons of different wavelengths before sunset (sun at 11.4° elevation), at sunset, and during nautical twilight (sun at 10.4° below the horizon), and with a full moon and starlight. Note that dim twilight is dimmer and strongly blue-shifted compared to moonlight, and that starlight is strongly red-shifted. B. The illuminance as a function of solar (upper two curves) and lunar (lower three curves) elevation. Starlight illuminance (gray shaded zone) differs with time of year and region. Vertical lines give limits: 0°: sunset/ sunrise; -6°: civil twilight; -12°: nautical twilight; -18° of astronomical twilight, which is when the sun or moon no longer contributes to illuminance. The luminance values are approximations, and indicate the luminance of a white surface positioned horizontally, in an open landscape given the indicated illuminance. C. Duration of the night in different regions, given for 21 June in the Northern Hemisphere, or 21 December in the Southern Hemisphere as a function of latitude. 398 THE QUARTERLY REVIEW OF BIOLOGY Volume 91 December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 399

in Cooper et al. (2015) who adopted this in families with C3 photosynthesis. Moreover, method since it has been shown to be su- of the 31 angiosperm families in which CAM perior at estimating marginal likelihoods photosynthesis has been reported (Silvera compared to other methods. In the step- et al. 2010), nocturnal pollination is present ping stone method, we sampled 50 step- in 21 families (Table 2). In two CAM families ping stones from a beta distribution with where nocturnal pollination is absent, the α = 0.4 and β = 1; each stone was sampled plants are aquatic and exhibit aquatic CAM. for 20,000 iterations with the ﬁrst 5000 iter- Therefore, 21 of 31 families (67.8%) with ations (burn-in) being discarded (Cooper CAM are nocturnally pollinated. We found et al. 2015). MCMC runs equalled 1,000,000 that over the phylogeny, CAM and nocturnal iterations with a prior drawn from an expo- pollination are correlated traits (marginal nential distribution of mean 10. In the Dis- likelihood values for the dependent model = crete module, we examined models where -342.7, and for the independent model = χ 2 the traits evolve in a correlated fashion -353.1; BF = 20.7, p < 0.001 under a dis- such that the rate of change in one trait tribution). Of the 17 angiosperm families in depends upon the background state of the which C4 photosynthesis has been reported other (dependent model), as well as mod- (Sage 2001), three are monocots (Poaceae, els where the traits evolve independently Cyperaceae, Hydrocharitaceae) that are of each other (independent model). In mostly pollinated by wind and have no these cases, BF = 2 (likelihood of depen- nocturnal pollination; the remaining 14 dent model – likelihood of independent families (Amaranthaceae, Euphorbiaceae, model). Correlated evolution between pairs Asteraceae, Polygonaceae, Acanthaceae, of binary states is considered to have oc- Portulacaceae, Caryophyllaceae, Zygophyl- curred when the BF value exceeds the crit- laceae, Boraginaceae, Aizoaceae, Nyctag- ical value of a χ 2 distribution with df = 4 inaceae, Scrophulariaceae, Molluginaceae, (considering the number of possible tran- and Capparaceae) are dicots of which 10 sitions between the states; Pagel 1999; Giv- families have examples of nocturnal pol- nish et al. 2014). lination; there appears to be a prepon- Since a major difference between day and derance of nocturnal pollination in dicot night pollination is the light level at which families with C4 photosynthesis (71.4%). pollination occurs, and since plants need Since CAM and C4 photosynthesis are be- light for photosynthesis, can the type of lieved to be adaptations to warm and dry photosynthesis help to explain the observed conditions (Sage 2001; Lüttge 2004; Sage patterns of nocturnal pollination? Of the et al. 2014), this therefore led us to ask 113 plant families with evidence of noctur- whether there is a preponderance of noc- nal pollination, 109 (96.5%) engage in C3 turnal pollination in families whose repre- photosynthesis (with or without C4 or CAM sentatives are reported within xerophytic photosynthesis), nine families employ the conditions or have speciﬁc adaptations to

C4 photosynthesis system, and 21 fam ilies hold water, e.g., succulence. We catego- employ CAM photosynthesis (with three rized the 113 families in which nocturnal families that had exclusively CAM photo- pollination has been reported into: helo- synthesis; Table 2). It is clear therefore that phytic families if most plants in this fam- nocturnal pollination is mostly represented ily appeared to be living near streams and

Figure 3. Schematic of the Changes in (from Top to Bottom) Wind Speed, Relative Humidity, Air and Surface Temperature, and Temperature Gradient A normal gradient means that temperature decreases with elevation above the surface while inversion indicates that a warmer air layer is found at higher elevation. Note that sunrise and sunset hours (approximately indicated by gray shaded areas) as well as the amplitude of all of these changes differ considerably between latitudes and time of the year. Gradients also differ with absolute humidity and elevation over sea level. For instance, in the wet tropics, the temperature changes are much smaller than in a dry climate. Wind speeds also strongly depend on surface topography. 400 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

TABLE 2 Traits in nocturnally pollinated families Nocturnally Traits possibly pollinated family* Habit* Pollinator reward# Photosynthesis pathway related to water+

Nymphaeaceae h h, p, m C3 l

Schisandraceae m h, p C3 o, m Piperaceae m p, b CAM o, m

Myristicaceae m p C3 o, m, l

Magnoliaceae m h, p C3 o, m

Eupomatiaceae m b, p (staminode material) C3 o

Annonaceae m h, p C3 m, r, o

Siparunaceae m b C3 o

Monimiaceae m f C3 o, m

Lauraceae m p, n C3 o, m

Araceae m h, p C3, CAM l

Burmanniaceae m n C3

Velloziaceae x n C3

Cyclanthaceae m h, m, b, f C3 l, m

Pandanaceae m n, p C3 m

Colchicaceae m n, p C3

Liliaceae x n C3

Orchidaceae x n C3, CAM m

Asteliaceae x n C3 m

Iridaceae x n C3 m, o

Xeronemataceae x n C3

Xanthorrhoeaceae x n, p C3, CAM m

Amaryllidaceae m n C3 m, l

Asparagaceae x n C3, CAM m

Arecaceae/Palmae x h, m, b, n, f, p C3 m

Haemodoraceae x n C3 m?

Strelitziaceae m n C3 m

Heliconiaceae m n C3 m

Musaceae m n C3 l, m

Cannaceae m n C3 m

Zingiberaceae m n C3 Bromeliaceae x n CAM m

Ranunculaceae m n C3

Proteaceae x n, p C3 s Crassulaceae x n CAM m

Zygophyllaceae x n, p C3, C4 m

Fabaceae/Leguminosae x n, p C3 m, r, o

Rosaceae x n C3 m

Rhamnaceae x n C3 m

Moraceae m b, n C3 m, l

Cucurbitaceae x n, p C3, CAM l

Celastraceae x n C3 m, l

Elaeocarpaceae x n, p C3 m

Rhizophoraceae h n C3 m

Euphorbiaceae x n C3, C4, CAM m, l

Phyllanthaceae m n C3 m

Chrysobalanaceae m n C3 m

Passiﬂoraceae x n, p C3, CAM m

Salicaceae x n C3 m, r, h, o

Violaceae m n, p C3 m, o

Achariaceae m n, p C3 m

Caryocaraceae x n C3 m, r? continued December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 401

TABLE 2 Continued Nocturnally Traits possibly pollinated family* Habit* Pollinator reward# Photosynthesis pathway related to water+

Clusiaceae/Guttiferae x n, p? C3, CAM l, r, o

Geraniaceae x n C3, CAM m, o

Combretaceae x n C3 h, m

Lythraceae x n C3 m

Onagraceae x n, p C3 m, o

Vochysiaceae x n C3 m, s

Myrtaceae x n C3 m, l, o

Melastomataceae x n, p C3 m, o Strasburgeriaceae m n s, m

Stachyuraceae m n C3

Anacardiaceae x n C3 l, r, m

Sapindaceae x n C3 m, l

Rutaceae x n C3 m, o

Meliaceae m n C3 m, l, r Cytinaceae m n

Malvaceae x n, p C3 m

Thymelaeaceae x n C3 m

Dipterocarpaceae m n, p C3 r, m

Caricaceae x n C3 l

Salvadoraceae x n C3

Capparaceae x n, p C3, C4 m

Cleomaceae x n C3, C4 Brassicaceae/

Cruciferae x n C3 m Balanophoraceae m n, p, b

Santalaceae x n C3

Loranthaceae x n C3 m

Nepenthaceae h n C3

Caryophyllaceae x n C3, C4 m

Aizoaceae x n C4, CAM m

Nyctaginaceae x n C3, C4

Cactaceae x n C3, CAM l, m

Loasaceae x n, p? C3

Balsaminaceae m n C3 h, m

Marcgraviaceae m n C3 m

Polemoniaceae x n C3 m

Lecythidaceae m n C3 m

Sapotaceae m n?, f C3 l, m, r

Primulaceae x n C3 h, s

Ericaceae x n C3 m, o

Rubiaceae x n C3, CAM m

Gentianaceae m n C3 m, o

Loganiaceae x n C3 m

Apocynaceae x n, p C3, CAM l, m

Boraginaceae x n, p C3, C4 m

Convolvulaceae x n, p C3 l

Solanaceae x n, p C3 m, o

Gesneriaceae m n C3, CAM r

Plantaginaceae x n C3, CAM h

Scrophulariaceae x n C3, C4 o, h

Lamiaceae/Labiatae x n C3, CAM h, o continued 402 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

TABLE 2 Continued Nocturnally Traits possibly pollinated family* Habit* Pollinator reward# Photosynthesis pathway related to water+

Orobanchaceae m n C3 h

Lentibulariaceae h n C3

Acanthaceae x n C3, C4 m unique phloem for

Bignoniaceae m n, p C3 holding water

Verbenaceae x n C3 m

Campanulaceae x n C3 l

Asteraceae/Compositae x n C3, C4, CAM l

Escalloniaceae m n C3

Adoxaceae m n C3

Araliaceae x n C3 s, r

Apiaceae x n? f? C3, CAM s, r, m *h = helophytic; m = mesophytic; x = xerophytic #h = heat; p = pollen; m = mating site; b = brood site; f = floral tissue; n = nectar +l = laticifer; o = oil; m = mucilage; r = resin canal; h = hydathode; s = secretory canal. Blanks indicate no information. *Linear sequence of families follows Haston et al. (2009). have an affinity for water; mesophytic fami- (81.4%) have at least one of these features, lies if plants in this family occupied habitats while in only 21 of these families was there of intermediate water regimes, neither too either no information on their occurrence wet nor too dry; and xerophytic families or none of these features occurred. We sim- if there was evidence that at least some ilarly found that nocturnal pollination and plants in this family are xerophytes and oc- the occurrence of any water-holding trait cupy areas where water is scarce (Table 2). (Table 2) in angiosperm families are highly We found that 67 families with nocturnal correlated traits (marginal likelihood val- pollination have xerophytic representa- ues for the dependent model = -467.4, and tives, four families are mostly helophytes, for the independent model = -520.9; BF = χ2 and 42 are mesophytic. When we scored 107.0, p <<< 0.001 under a distribution). these traits over the entire phylogeny (and pooled helophytic and mesophytic traits into one category), we found that noctur- The Arid Hypothesis for Nocturnal nal pollination and xerophytism are highly Flowering and Pollination correlated traits (marginal likelihood val- We therefore propose the arid hypoth- ues for the dependent model = -427.1, and esis for nocturnal flowering and thereby for the independent model = -472.8; BF = nocturnal pollination, according to which χ 2 91.5, p << 0.001 under a distribution). plants that are water-stressed should pref- Therefore, xerophytism is more likely to erentially flower at night since they can occur in families with nocturnal/crepuscu- reduce water loss by doing so. Flowering lar pollination than those with helophytic is a water-demanding process requiring wa- or mesophytic representatives. We there- ter at all stages from flower bud maturation fore scored the occurrence of the follow- to flower opening, and also for mainte- ing features that may aid plants to cope nance of turgor in floral organs as well with water stress or function as an adap- as nectar production (Mohan Ram and tation to water regimes: mucilage, resin, Rao 1984; Galen et al. 1999; Galen 2000; hydathodes, secretory canals, laticifers, or De la Barrera et al. 2009). Such water de- any other special feature for water storage mands can only be met from the vegeta- (Table 2). We found that plants in 92 out tive parts of the plant after water loss by of the 113 nocturnally pollinated families evapotranspiration. December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 403

In the Polemoniaceae, a family with C3 then, by having flowers open at night when photosynthesis, the influence of water stress transpiration losses are likely to be much has, for example, led to flower sizes of di- lower than in the day, the constraint on urnally opening flowers in Leptosiphon bi- flower size may be lifted. This may conse- color being smaller in sites with lower soil quently also allow for the production of moisture than in those with higher soil larger flowers or inflorescences with abun- moisture (Lambrecht 2013). Large flow- dant nectar, which may support larger noc- ers could only be produced in dry sites if turnal pollinators such as bats and large leaves closed their stomata during the day moths. Plants with C4 photosynthesis have to prevent water loss (Lambrecht 2013). greater water use efficiency (defined as the Similarly, Galen et al. (1999) showed that ratio of the rate of carbon assimilation by flower size in the diurnal Polemonium vis- photosynthesis to the rate of water loss by cosum (Polemoniaceae) is positively related transpiration); sustaining high rates of pho- to water uptake during bud expansion and tosynthesis even when stomatal conduc- anthesis; plants with larger flowers take up tance is low results in lowered transpiration more water. In another set of five species of and protection of the hydraulic system un- unknown photosynthesis type, flower size der water stress conditions, allowing C4 was larger in sites with greater moisture plants to colonize dry environments (Os- content (Lambrecht and Dawson 2007). borne and Sack 2012). It is also possible Clearly, flowering imposes a heavy water that those plants in which stomata are cost on plants. In avocado trees, 13% of open at night (as happens during CAM transpirational water loss was due to loss photosynthesis) would be predisposed to from floral organs (Whiley et al. 1988). nocturnal flowering. Considering that noc- This may be a general trend across plants. turnal pollination is significantly more In Agave deserti (Asparagaceae), a noc- frequent in those families that occupy water- turnally pollinated CAM plant, a mature stressed environments, it is reasonable to plant requires about 18 kg of water during speculate that opening flowers at night flowering; this is mainly supplied from leaf helps to relieve water stress. It appears nec- water stores; 4 kg of this water is taken up essary to learn much more about water as by the leaves of which 80% is transpired an important cost of flowering to under- (Nobel 1977). Abundant watering of A. de- stand the evolution of nocturnal flowering serti can cause these plants to shift from the and nocturnal pollination in plants. CAM mode of nocturnal stomatal opening Despite the fact that water is so important to diurnal stomatal opening (Hartsock and in the process of flowering, it is phloem and Nobel 1976), affirming that CAM photo- not xylem that supplies water to nectar (De synthesis is a response to water stress (Her- la Barrera and Nobel 2004). However, floral rera 2009). The large flowers of the diurnal nectar is not merely excreted phloem sap, CAM plant Opuntia ficus-indica (Cactaceae) and considerable synthesis of nectar com- lose 15% of their mass by transpiration ponents probably occurs within the nectary during anthesis (De la Barrera and Nobel itself (Escalante-Pérez and Heil 2012; Cha- 2004). It can be clearly seen that water is a nam et al. 2015). Floral nectar, especially limiting factor on flower size and this may that produced by succulents, is copious; be more so for C3 plants in which stomata e.g., Aloe marlothii in southern Africa pro- are necessarily open during the day. In duces 50–100 liters of nectar/ha resulting such plants, opening flowers at night and in 100,000–200,000 kj/ha in terms of en- engaging in nocturnal pollination would ergy production (Wolf and Hatch 2011). be a valuable water conservation strategy. Therefore, plants that can save water by It is not surprising, therefore, that noctur- opening during the night could support nal pollination is mostly represented in populations of nocturnal pollinators by pro-

C3 families as we have shown earlier. Fur- ducing sufficient quantities of nectar. This thermore, if nocturnal blooming is an ad- may have resulted in the radiation of large aptation primarily to deal with water stress nocturnal nectarivores such as bats and large 404 THE QUARTERLY REVIEW OF BIOLOGY Volume 91 moths. Bat-pollinated plants, at least in the cestral bird pollination (Tschapka and von Neotropics, also tend to be more repre- Helversen 1999) or to prevent discovery by sented in arid habitats (Fleming et al. 2009). sphingid nectar robbers (von Helversen 1993). Since nocturnal pollination also occurs Clearly, the visual and overall sensory sys- in a wide range of families, it is possible that tems of pollinators need to be taken into a variety of plant species that experience consideration to understand these flower varied amounts of water stress, depending color tran sitions. on which habitat or habitat stratum they occupy, may profit from flowering at night and hence may make the switch from di- Nocturnal Pollinators urnal to nocturnal flowering. For example, low availability of water is an important For nocturnal pollination, plants have stressor in epiphytes (Zotz and Hietz 2001) to recruit nocturnal animals as pollinators even when they occur in rainforests (Lüttge by offering rewards. Considering the di- 2010); consequently, many epiphytes adopt versity of animals that are active at night, CAM photosynthesis (Lüttge 2004) espe- the taxonomic representation of known cially when they are in dry forest canopies nocturnal pollinators is limited (Table 3).

(Nadkarni et al. 2001), while C3 photosyn- Among the invertebrates, moths and bee- thesis prevails in epiphytes in the canopies tles are the major nocturnal pollinators. of cloud forests (Nadkarni et al. 2001). It Moths are well studied while beetle pollina- would be interesting to speculate on the tion is generally understudied with respect specific occurrence of nocturnal pollina- to diurnal versus nocturnal activity. Of the tion in different habitats and even different hymenopterans, several species of smaller habitat strata with varying levels of water and larger bees such as Sphecodogastra or stress. Information about abiotic factors is Megalopta are crepuscular pollinators (Kel- clearly required to predict the occurence ber et al. 2006), but the large carpenter of nocturnal pollination. bee Xylocopa tranquebarica is currently the Given the huge diversity of families that only known nocturnal bee to be active even show examples of nocturnal pollination, during moonless nights (Somanathan et al. it is difficult to make general statements 2008a). Yet, even the generally diurnal rock about floral traits of nocturnally pollinated honey bee Apis dorsata and the African race plants. However, nocturnally pollinated flow- of honey bee Apis mellifera adansonii can ers tend to be light/pale/white in color forage and act as pollinators during half- (Baker 1961). Since flower pigments such moon nights (Dyer 1985). In the Diptera, as flavonoids provide floral organs with pro- gall midges and fungus gnats are impor- tection against heat stress resulting in higher tant nocturnal pollinators (e.g., Yuan et al. pollen performance and seed production 2008; Luo et al. 2010; Duque-Buitrago et al. (Coberly and Rausher 2003), it is tempting 2013). A few plants are nocturnally polli- to suggest that by flowering at night, plants nated by mosquitoes and calliphorids, cock- are removed from the constraints of produc- roaches and orthopterans (a cricket and ing pigmented flowers and could therefore a weta), and ants (Table 3, Appendixes 1 produce white or pale-colored flowers, which and 2). may also have the added advantage of being Of the vertebrates, bats are the most im- highly conspicuous during the night. Since portant nocturnal pollinators (Table 3). In flower pigment genes also have pleiotropic the order Chiroptera, nectar foraging has effects (Coberly and Rausher 2008), the tar- evolved independently more than once in gets of selection on white-colored flowers the Old World family Pteropodidae, and may be varied; however, all else being equal, in the New World family Phyllostomidae it is worth considering the lower tempera- (Fleming et al. 2009; Fleming and Kress ture at night as a positive selection pressure 2013). Several nocturnal nonflying mam- for white corollas. On the other hand, dark mals such as marsupials, lemurs, shrews, red flower color may have been retained in and rodents also participate in pollination some bat-pollinated flowers as a relic of an- (Carthew and Goldingay 1997; Goldingay December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 405

2000). In a few cases, geckos engage in Long-proboscid nectar-feeding insects may nocturnal pollination (e.g., Eifler 1995; have pollinated angiosperm flowers since Godínez-Álvarez 2004). Rodent pollination is the Cretaceous over 100 million years ago prevalent in geophytic plants (e.g., Turner (Labandeira 2010, 2013) and it is possible et al. 2011), as is the case for ant pollina- that some of these (for instance, nematoc- tion (e.g., de Vega et al. 2009). erans) were nocturnal. In contrast, the first Many examples of unusual pollinators nectar-feeding phyllostomid bats evolved such as geckos occur in habitats such as from initially insectivorous bats less than islands (Whitaker 1987), where there may 30 million years ago in the early Miocene be a dearth of regular nocturnal pollina- (Datzmann et al. 2010), and other noc- tors. Other such examples include rodents turnal vertebrates were recruited as polli- substituting for bats in a cloud forest (Lu- nators possibly even later. How the more mer 1980), moths substituing for bats on ancient origin of insect versus vertebrate remote Japanese islands where bats are ab- nectar feeding has affected the divergence sent (Norio 2004), or weta pollination in of nocturnally pollinated plant lineages is stressful habitats such as high-altitude for- completely unknown, and may be signifi- ests in New Zealand (Lord et al. 2013). cant given that nectar is an important re- Such unusual pollinators may be referred ward in most nocturnally pollinated plant to as opportunistic, i.e., being in the right families. place at the right time. In some cases, these unusual pollinators are nectar specialists as in the New Zealand Hoplodactylus gecko Sensory and Physiological that feeds on honeydew when nectar is Adaptations and Constraints scarce (Gardner-Gee and Beggs 2010). of Nocturnal Pollinators The requisites for pollen transfer, which Nocturnal pollinators, obviously, have are mobility as well as pollen adherence, may to combine traits that enable them to: use restrict the types of taxa that can be effective nectar and/or pollen or other rewards of- during nocturnal pollination. This is why, fered by nocturnal flowers; move pollen for example, although thrips are found in efficiently between flowers and plants; and flowers during the night, they cannot move be active and do all of this at night. For all long distances between plants since they are of the major groups of pollinators, there dependent on the wind for transportation, is copious literature including several re- making them rather unreliable for cross- cent books on the first two topics, i.e., the pollination (Sakai et al. 1999; Webber et al. general adaptations to a nectar or pollen 2008) since wind speeds are lower at night. diet and to moving pollen (e.g., Waser and The vast majority of nocturnal pollinators Ollerton 2006; Patiny 2012; Fleming and consume nectar as a reward, although Hy- Kress 2013). Therefore, we will focus on menoptera, some rodents, gall midges, and the third topic, the sensory and physiolog- beetles collect pollen and, in some families, ical adaptations of pollinators to a noctur- floral tissue or staminodes are also rewards nal lifestyle. (Table 2). Some pollinators utilize floral re- Adaptations to features of the night in sources as mating sites and/or brood sites nocturnal pollinators include adaptations where eggs are laid and offspring develop; to locomotion in low temperature and low heat produced by floral tissue is a reward in humidity, adaptations to sensing flowers, several systems (Table 2). and to navigation between them in dim The basal or ANITA grade angiosperm light and with low wind speeds. Although families are mostly pollinated by Diptera nocturnal birds and bats prey upon noctur- and Coleoptera (Thien et al. 2009) with nal pollinators, the night may be a temporal gall midges and beetles engaging in noc- niche with lower predation pressure (e.g., turnal pollination in several families (Ap- Wcislo et al. 2004). Nocturnal pollinators pendix 1). Insect pollination may have include a diverse range of animal groups occurred as early as the Permian (250– (Table 3), among which moths and bats are 300 million years ago; Labandeira 2013). most species-rich and quantitatively most 406 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

TABLE 3 Taxonomic groups with nocturnal pollinators Order Family References (see also Appendixes 1 and 2)

Class Insecta Orthoptera Gryllacrididae Micheneau et al. 2010 Blattodea Blattellidae Vlasáková et al. 2008 Hymenoptera Formicidae de Vega et al. 2009 Vespidae Nakase and Kato 2012 Andrenidae Kelber et al. 2006 Halictidae Kelber et al. 2006 Apidae Somanathan and Borges 2001; Kelber et al. 2006; Somanathan et al. 2008a,b Colletidae Linsley and Cazier 1970 Coleoptera Scarabaeidae Hirthe and Porembski 2003; Gottsberger et al. 2011 Anthicidae Armstrong and Drummond 1986 Curculionidae Bergstrom et al. 1991; Franz 2007 Lepidoptera Noctuidae Faegri and van der Pijl 1979 Geometridae Sphingidae Prodoxidae yucca moths Pellmyr 2003 Gracillariidae Kawakita and Kato 2004 Pyralidae Kawakita and Kato 2002 Diptera Culicidae Branjtes and Leemans 1976; Kato 1996 Cecidomyiidae Feil 1992; Yuan et al. 2008 (gall midges) Mycetophilidae Vogel and Martens 2000; Duque-Buitrago et al. 2013 (fungus gnats) Sciaridae (fungus gnats) Vogel and Martens 2000; Duque-Buitrago et al. 2013 Tipulidae (craneﬂies) Primack 1983 Calliphoridae Kato 1993 Class Reptilia Squamata Gekkonidae Newstrom and Robertson 2005 Class Mammalia Infraclass Marsupialia Didelphimorphia Didelphidae Carthew and Goldingay 1997 Diprotodontia Phalangeridae Carthew and Goldingay 1997 Petauridae Carthew and Goldingay 1997 Burramyidae Acrobatidae Tarsipedidae Dasyuromorphia Dasyuridae Infraclass Eutheria Chiroptera Phyllostomidae Fleming et al. 2009 Mystacinidae Pteropodidae Macroscelidea Macroscelididae (elephant Carthew and Goldingay 1997; shrew) Johnson et al. 2011 Rodentia Muridae (mice and rats) Gliridae (dormice) Carthew and Goldingay 1997

Cricetidae (voles) Carthew and Goldingay 1997

Petauristidae Ganesh and Devy 2000 continued December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 407

TABLE 3 Continued Order Family References (see also Appendixes 1 and 2)

Carnivora Viverridae Carthew and Goldingay 1997 Procyonidae Carthew and Goldingay 1997 Primates Cheirogaleidae (Dwarf lemurs) Carthew and Goldingay 1997; Heymann 2011 Lemuridae Carthew and Goldingay 1997; Galagidae Heymann 2011 Lorisidae (potto) Aotidae (night monkey) important, and we will mainly concentrate Many nocturnal insect pollinators have on adaptations in these groups but mention good temperature control and preheat flight other cases, such as the repeated evolution muscles prior to taking flight. For instance, of nocturnal activity in bees (e.g., Kelber hawkmoths fly at 40°C muscle temperature et al. 2006). Animal groups that could not allowing for fast long-distance flight, and adapt to all of these demands do not con- they can easily overheat at high environ- tribute to nocturnal pollination. As pointed mental temperatures (Heinrich 1971). A out earlier, we need to keep in mind that direct example of insects using the crepus- the features of the night differ between lat- cular period to escape overheating is seen itudes and climate zones. Perhaps the most in the hawkmoth Macroglossum stellatarum, obvious difference in the duration of the which is generally purely diurnal but has night is reflected in the activity patterns of an activity peak after sunset under Medi- crepuscular pollinators, which are active terranean hot summer conditions (Her- during dawn and dusk periods. Although rera 1992). In colder climates, nocturnal crepuscular moths have a long activity pe- or crepuscular insects often have an insu- riod in northern Sweden, their activity win- lating layer of cuticular hairs as in moths dows can be as short as half an hour close (Heinrich 1993). Also, the truly nocturnal to the equator ( Johnson and Nilsson 1999; Asian carpenter bee X. tranquebarica, which Kelber et al. 2006; Martins and Johnson can fly at low temperatures, is more densely 2007). pubescent than its diurnal congeners (per- sonal observations). Thick pelage, thoracic shivering, and counter-current exchangers locomotion at low that slow down heat flow to the head and ab- temperatures and humidity domen enable moths to fly at low tempera- For efficient pollen transfer, pollinators are tures (Heinrich 1987). Geometrid moths required to move between plants; thus most may fly at low temperatures by having un- pollinators—with the exception of some in- usually low wing loading that enables low en- sects (such as ants), geckos, and nonflying ergy expenditure for flight even if thoracic mammals—use flight for locomotion. Lo- muscle temperature is at ambient freezing comotion at night functions under very dif- temperatures (Heinrich and Mommsen ferent constraints in different climate zones. 1985). Midges and mosquitoes that fly at In tropical rainforests, temperature does not low temperatures have high metabolic rates, change much between day and night. How- and scarab beetles can also raise thoracic ever, in temperate and dry habitats, night temperatures and fly at low temperatures temperatures are considerably lower than (Morgan 1987; Heinrich 1993). day temperatures (see earlier sections), and Freeze tolerance has independently here nocturnal or crepuscular flight may be evolved at least six times among the insects an adaptation to avoid high temperatures (in the Blattaria, Orthoptera, Coleoptera, and low humidity during the day (Willmer Hymenoptera, Diptera, and Lep idoptera; and Stone 1997). Sinclair et al. 2003). The New Zealand weta 408 THE QUARTERLY REVIEW OF BIOLOGY Volume 91 that pollinates megaherbs at night on the (Heath and Derraik 2005) while nocturnal subantarctic Campbell Island is active at moths fly at all strata (Ashton et al. 2016). temperatures of 3 – 4°C (Lord et al. 2013). In tropical and humid habitats, temperatures do not change much between day and sensory adaptations for night. It is therefore interesting that only finding flowers at night freeze-tolerant groups of insects are in- Pollinators sense flowers mostly by vision volved in nocturnal pollination, which may and olfaction. Visual cues become less reli- imply that these groups have wider thermal able when light intensities are low, thus noc- tolerance than other insect groups, facili- turnal pollinators—just as other nocturnal tating their locomotion and functionality animals—have a number of adaptations for even at extremely low temperatures. seeing well in dim light: large eyes relative Among mammals, nectar feeders have to body size, large pupils relative to focal body temperatures that, although higher length, highly sensitive photoreceptors, than those of carnivores, are lower than those and neural mechanisms for pooling signals of other herbivorous species (Clarke and in space and time (e.g., Kelber and Roth O’Connor 2014). The high energy demands 2006; Warrant 2008). This allows them to of both large hawkmoths and bats indicate reliably detect the highly reflective noctur- that nocturnal flowers have to produce rel- nal flowers against the dark backgrounds of atively large amounts of nectar, especially the night sky and vegetation. In addition, since nectar-feeding bats, unlike some other larger nocturnal pollinating insects, such as bats, consume their body fat on a daily basis hawkmoths and large carpenter bees, can (Voigt and Speakman 2007) and do not have discriminate colors even in dim light when the ability to go into a state of torpor when humans and most other animals are color inactive to save energy when resting (e.g., blind (Kelber et al. 2002; Kelber and Roth Bartholemew et al. 1970). Copious nectar 2006; Somanathan et al. 2008b). production has long been known for bat-pol- Although vision becomes less reliable in linated flowers (e.g., Faegri and van der Pijl the dim light of the night, flower odor may 1979; Winter and von Helversen 2001), but become a more reliable and more long- may be more variable among groups polli- ranging cue. Wind speeds are lower at nated by moths (Haber and Frankie 1989). night, but there is less turbulence (Nadeau Since large body size is also correlated with et al. 2011); thus odor plumes are carried low ambient temperatures in ectotherms longer distances with less disturbances (see (Atkinson 1994), some nocturnal pollinators above) allowing pollinators to find flowers may also be larger than their diurnal coun- from further away. It is therefore not sur- terparts. For example, larger body size and prising that nocturnal moths—in contrast thick pelage enable cold tolerance in bees to diurnal species—rely more on olfaction (Bishop and Armbruster 1999). than on vision, both innately and after Stratum of flight at night also needs to learning a food source (Balkenius et al. be understood in terms of pollinator phys- 2006). Despite this preference, the majority iological capabilities and thereby stratum of nocturnal pollinators use both olfactory of pollinated flowers; e.g., nocturnal Meg- and visual stimuli (e.g., Kelber et al. 2002; alopta bee species were found in high can- Raguso and Willis 2003; Riffell and Alarcón opy traps at 30 m as well as in low traps 2013). Additionally, nocturnal hawkmoths use

(Roubik 1993). Mycetophilids and sciarids gradients of CO2 that indicate floral respira- (both fungus gnats), on the other hand, tion (Thom et al. 2004; Goyret et al. 2008) fly close to the ground (Peng et al. 1992; and gradients of humidity arising from nec- Eberhard and Flores 2002) and being weak tar evaporation (von Arx et al. 2012) to lo- fliers (Lewis 1967) are suitable for pollina- cate flowers at closer range. Moreover, once tion of the Araceae and terrestrial orchids having found flowers by using long-distance (Appendix 2). Craneflies or tipulids are cues, nocturnal moths appear to also use also weak fliers and keep to ground levels tactile cues from flowers to a greater extent December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 409 than diurnal moths (Balkenius et al. 2006; tle research, but both hawkmoths and bats Goyret 2010; Goyret and Kelber 2011) sug- are known to have good spatial memories gesting that successful nocturnal flower vis- (hawkmoths: Balkenius et al. 2004; bats: itation may demand the recruitment of a Winter and Stich 2005). Nocturnal moths greater number of sensory modalities. (see Warrant and Dacke 2011), bees (Baird Among bats, the Pteropodidae also use a et al. 2011), and bats (Fleming et al. 2009) combination of visual and olfactory cues for use visual cues for flight control and naviga- flower detection (Fleming et al. 2009), but tion emphasizing the necessity for large, in dim light they use their rod photorecep- highly sensitive eyes in nocturnal pollinators. tors restricting them to the use of brightness instead of color cues (Winter et al. 2003). Interestingly, although many bat-pollinated Gaps and the Future flowers are white or brightly colored, others The night is a temporal niche in which have dull colors, and thus are more likely to light, wind, and temperature levels are gen- rely on their scent to attract bats. In addition erally lower while humidity levels are higher to visual and olfactory cues, the Phyllostomi- than during the day. Although it is a niche dae can also use echoes from flowers (von that frees plants from water constraints, a Helversen and von Helversen 1999; Simon night-flowering plant is unlikely, for instance, et al. 2011), yet this is necessarily a shorter- to be wind-pollinated. If plants are to be biot- distance cue compared to vision or olfaction. ically pollinated at night, they need the ser- Gall midges are weak fliers and are likely vices of those pollinators that can overcome attracted by scent (Luo et al. 2010). They the physiological constraints imposed by the have very sensitive antennae (Hall et al. 2012) entire night or certain parts of it (see Table 4 and also a unique type of fused sensillum for known and suggested traits of noctur- called the circumfila (Boddum 2013). Gall nally pollinated plants and pollinators). For midges also have infrared (IR) receptors example, some plants may open flowers (Zahradnik et al. 2012) and may use these in the early evening or morning (dusk and to find the large thermogenic flowers of the dawn) and use the services of those pollina- Schisandraceae (Takács et al. 2009). At night, tors that are not fully visually adapted to the such flowers present a higher IR contrast night but have just high enough visual sensi- against the background and this may be used tivity to navigate in twilight. On the pollina- as an effective flower location signal. Al- tor side, very small species may often lack the though the eyes of diurnal and nocturnal capacity to see well enough at night. What mosquitoes differ (Land et al. 1999), whether constrains the evolution of nocturnal activity they use vision to find flowers is unknown. across pollinator taxa is not well understood. Nocturnal mosquitoes are, however, attracted We are not aware of any attempt to explain by floral scents (Brantjes and Leemans why common diurnal pollinator groups such 1976; Jhumur et al. 2007). Very little work has as birds have not expanded their activity into been done on color vision in flower-visiting the night, or why there is scarce evidence of beetles (e.g., Dafni et al. 1990; Keasar et al. taxa such as wasps being engaged in noctur- 2010; Martínez-Harms et al. 2012), while nal pollination. considerably more is known about olfactory Some plants may inhabit areas where cli- attraction in beetles, including those visiting matic conditions are harsh both during the flowers at night (Maia et al. 2012; Pereira day and night; such plants, as in the high et al. 2014). The spread of the olfactory sig- Andean Espeletia (a giant caulescent rossette nal may also be helped by thermogenesis in plant) may utilize the services of diurnal and some beetle-pollinated flowers. nocturnal pollinators since pollinator num- In addition to finding flowers, nocturnal bers are usually low in such extreme con- pollinators need to be able to move between ditions (Fagua and Gonzalez 2007). Many flowers at night, demanding good orienta- plants that inhabit similarly challenging en- tion, navigation, and a spatial memory. These vironments in South Africa such as the Ka- topics have generally attracted relatively lit- roo employ nocturnal rodents as pollinators 410 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

TABLE 4 Plant and pollinator traits that could facilitate nocturnal pollination Plant traits Pollinator traits

Low light a) High conspicuousness of flowers by increased a) Simple eyes: large eyes, large pupils, short focal levels achromatic contrast length, long photoreceptor outer segments, spatial and/or temporal pooling b) If flower pigments are selected for protection against heat and ultraviolet damage, then at b) Compound eyes: for moths: superposition night, these constraints are removed; therefore, optics; for bees: large ommatidia, large flowers can afford to be pigmentless (i.e., white) rhabdoms, spatial and/or temporal pooling; for at night mosquitoes: large interommatidial angles, wide fused rhabdoms, strong adaptation changes c) Large floral display per plant (many small flowers or few large flowers) for high c) Large ocelli conspicuousness d) Well-developed olfactory system d) Flowers easily accessible and therefore readily e) Well-developed ability to sense temperature visible differentials such as infrared detection e) Visual signalling likely to be coupled with chemical signaling in the form of scents produced by floral and associated tissues

Reduced wind a) No reliance on wind pollination or on biotic a) Ability to ﬂy without wind assistance: suitable agents that are wind-assisted in ﬂight wing loading, wing musculature, fuel reserves

b) Pollen donation is not proﬂigate

c) Pollen likely to be sticky

d) Pollen receipt structures likely to be commensurate with more specialized donation

e) Adequate rewards to fuel nonwind-assisted ﬂight of pollinators

Lower a) Flowers should not be open during conditions a) Larger body size for adequate heat retention, temperature of frost or when temperatures are very low heat generation under low temperatures

b) If ﬂowers are thermogenic, then heat can be a b) Thick chitinized cuticle for reduced reward for pollinators under lower temperatures temperature loss

c) Thicker pubescence (bees), scales (moths), or fur (mammals) for heat retention under low temperatures

d) Greater cold tolerance

e) Ability to sense temperature differentials using infrared (IR) detectors

f ) Moths and bees: ability to preheat ﬂight muscles

Higher a) Flowers should not be open during rain a) Ability to cope with dew and/or rain humidity b) Flowers should be downfacing to prevent damage by water droplets or nectar dilution

(Kleizen et al. 2008). It would be interesting opening times, flower sizes, and rewards that to investigate whether plants have shifted corresponds to a gradient in the energetic flowering to particular parts of the night or needs and sensory capabilities of the polli- day in response to the availability of physio- nators that would also partly scale with their logically capable pollinators, or vice versa. It is body sizes. Even within a single habitat, we also possible that there is a gradient in flower may expect gradients in drought and expo- December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 411 sure that favor nocturnal pollination, e.g., in difference in cold tolerance in insects with bromeliads (Graham and Andrade 2004). It many more insects adapted to the cold in would be interesting to examine the prob- the Southern Hemisphere as compared ability of transitions between diurnal and with the north (Sinclair and Chown 2005); nocturnal pollination and correlated evo- it is possible that there may be more plants lution between functional traits such as pollinated at night by insects in the south water-holding capacity or type of photosyn- compared with the north. For example, thesis at lower taxonomic levels. This is New Zealand wetas and cockroaches show be cause, even within a family, transitions great cold tol erance (Wharton 2011) and from nocturnal to diurnal pollination can pollinate plants at night. occur (Rentsch and Leebens-Mack 2014). Since the night may also feature greater We need much more data on the match- humidity, flowers may also be downfacing ing of traits between plants and pollinators to prevent nectar dilution by water drop- and their evolutionary trajectories. Addi- lets (Aizen 2003) or have structures that tionally, there appears to be a hemispheric prevent pollen damage by rain (Mao and

Box 1 Gaps in our knowledge and a sample of questions to be answered t8IBUJTUIFSFBMEJWFSTJUZPGOPDUVSOBMMZCMPPNJOHQMBOUTQFDJF T t "TTVNJOHUIBUUIFBSJEIZQPUIFTJTPGOPDUVSOBMGMPXFSJOHJTDPSSFDU BOEJG$".

and C4 photosynthesis already confers plants with protection against drought, why are there many examples of plants with nocturnal pollination in families with

CAM and C4 photosynthesis? t $BOBDPNQBSJTPOPGDMPTFMZSFMBUFEEJVSOBMBOEOPDUVSOBMQMBOUTSFWFBMUIFNPS- phological and physiological constraints or adaptations necessary for diurnal or nocturnal blooming? t %PQMBOUTBMPOHBOBMUJUVEJOBMPSMBUJUVEJOBMHSBEJFOUIBWFGMPXFSPQFOJOHBOE closing times that match the activity periods of their pollinators based on differences in the length of dawn and dusk periods and other abiotic factors character- istic of the night at these different geographical locations? t 8IBUBSFUIFGFBUVSFTPGOPDUVSOBMQPMMJOBUJPOOFUXPSLTJOBTJOHMFIBCJUBUBOE how do they compare with diurnal networks in the same habitat? t 8IBUJTUIFTFBTPOBMBWBJMBCJMJUZPGEJVSOBMMZBOEOPDUVSOBMMZGMPXFSJOHQMBOUTJO a single habitat; how does this availability change with seasonal variation in abun- dance and diversity of pollinators? t 8IBUBSFUIFTFOTPSZBCJMJUJFTPGOPDUVSOBMMZQPMMJOBUJOHUBYBTVDIBTNJEHFT CFF- tles, orthopterans, rodents, geckos, and primates? t 8IBUBSFUIFOBWJHBUJPOBOENPWFNFOUDBQBCJMJUJFTPGUIFOPDUVSOBMMZQPMMJOBUJOH taxa mentioned above? t %PUIFGMPSBMBUUSBDUJPOTJHOBMTQSPEVDFECZQMBOUTNBUDIUIFTFOTPSZBCJMJUJFTPG these understudied groups of nocturnal pollinators? t )PXEPUIFGMPSBMTJHOBMTPGDMPTFMZSFMBUFEBOETZNQBUSJDEJVSOBMBOEOPDUVSOBM plants compare? t "SFOPDUVSOBMQPMMJOBUPSTNPSFTQFDJBMJTUTPSHFOFSBMJTUT *TJUQPTTJCMFUPQSFEJDU the degree of specialization based on the taxon of pollinator and on the macro- habitat (geographical location) and microhabitat (understory versus canopy, for example)? t "SFOPDUVSOBMGMPXFSTBMTPTFSWJDFECZEJVSOBMQPMMJOBUPST *GTPUPXIBUFYUFOU Do floral traits of such plant species differ from flowers that are strictly open only at night or in the day? 412 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

Huang 2009). Whether these features also species within well-investigated families, occur in flowers that open at night and ex- there is a complete paucity of information perience greater humidity in the form of on the physiological capacities of many taxa dew (Berkelhammer et al. 2013) is not yet of potential nocturnal pollinators. These la- known. Flower orientation is another trait cunae become even more important given that needs greater investigation (Fenster the rapid erosion of biological diversity et al. 2009). Many flowering species also since we would be unable to predict which have flowers that remain open for several plant–pollinator nodes are more vulnerable days, and are open both during the day in a nocturnal plant–pollinator network. and the night; here pollination can also We can only begin to understand noc- occur during the day and night (Valdivia turnal pollination patterns and processes if and Niemeyer 2006), such that different there is collaboration between pollination pollinators may vary in pollination services. biologists, sensory biologists, plant ecophys- Pollinator services can differ even between iologists, and animal physiologists. This will nocturnal pollinators. For example, in the enable us to make sense of the correlation parasitic Balanophora, ants and cockroaches pleiades within and between plant and pol- contribute to geitonogamy or movement of linator traits (sensu Berg 1960; Armbruster pollen within a single plant while pyralid et al. 2014). There are good signs that this moths contribute to outcrossing (Kawakita is happening, but there is a long way to go. and Kato 2002). In the Nyctaginaceae, Aclei- We hope that this review will stimulate re- santhes longiflora is cleistogamous (closed search in this very important area. flowered) in summer when pollinating nocturnal hawkmoths are rare; therefore, in acknowledgments this case, cleistogamy may be an adaptation to a seasonal lack of nocturnal pollinators This review emerged out of thinking about nocturnal pollination following our joint work along with Eric (Douglas and Manos 2007). Warrant on a nocturnal carpenter bee that has color In this paper, we have sketched broad pat- vision under starlight. We thank Rob Raguso for crit- terns, and there are many gaps (Box 1). Just ical comments on this manuscript, as well as Yuvaraj as there is no information on pollination for Ranganathan and Vignesh Venkateswaran for assis- many plant families, and also for many plant tance with Figure 1 and the Bayesian analyses.

REFERENCES

Aizen M. A. 2003. Down-facing flowers, humming- Baird E., Kreiss E., Wcislo W., Warrant E., Dacke M. birds and rain. Taxon 52:675–680. 2011. Nocturnal insects use optic flow for flight Armbruster W. S., Pélabon C., Bolstad G. H., Hansen control. Biology Letters 7:499–501. T. F. 2014. Integrated phenotypes: understanding Baker H. G. 1961. The adaptation of flowering plants trait covariation in plants and animals. Philosophi- to nocturnal and crepuscular pollinators. Quarterly cal Transactions of the Royal Society B: Biological Sci- Review of Biology 36:64–73. ences 369:20130245. Balkenius A., Kelber A., Balkenius C. 2004. A model of Armstrong J. E., Drummond B. A. III. 1986. Floral bi- selection between stimulus and place strategy in a ology of Myristica fragrans Houtt. (Myristicaceae), hawkmoth. Adaptive Behavior 12:21–35. the nutmeg of commerce. Biotropica 18:32–38. Balkenius A., Rosén W., Kelber A. 2006. The relative Ashton L. A., Nakamura A., Basset Y., Burwell C. J., importance of olfaction and vision in a diurnal Cao M., Eastwood R., Odell E., de Oliveira E. G., and a nocturnal hawkmoth. Journal of Comarative Hurley K., Katabuchi M., Maunsell S., McBroom J., Physiology A: Neuroethology, Sensory, Neural, and Be- Schmidl J., Sun Z., Tang Y., Whitaker T., Laidlaw M. J., havioral Physiology 192:431–437. McDonald W. J. F., Kitching R. L. 2016. Vertical Bartholomew G. A., Dawson W. R., Lasiewski R. C. 1970. stratification of moths across elevation and lati- Thermoregulation and heterothermy in some of the tude. Journal of Biogeography 43:59–69. smaller flying foxes (Megachiroptera) of New Guinea. Atkinson D. 1994. Temperature and organism size—a Zeitschrift für Vergleichende Physiologie 70:196–209. biological law for ectotherms? Advances in Ecologi- Berg R. L. 1960. The ecological significance of cor- cal Research 25:1–58. relation pleiades. Evolution 14:171–180. December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 413

Bergström G., Groth I., Pellmyr O., Endress P. K., flowers: convergence for beetle pollination in the Med- Thien L. B., Hübener A., Francke W. 1991. Chem- iterranean region. Israel Journal of Botany 39:81–92. ical basis of a highly specific mutualism: chiral es- Dai A. 2006. Recent climatology, variability, and trends ters attract pollinating beetles in Eupomatiaceae. in global surface humidity. Journal of Climate 19: Phytochemistry 30:3221–3225. 3589–3606. Berkelhammer M., Hu J., Bailey A., Noone D. C., Still Datzmann T., von Helversen O., Mayer F. 2010. Evolu- C. J., Barnard H., Gochis D., Hsiao G. S., Rahn T., tion of nectarivory in phyllostomid bats (Phyllosto- Turnipseed A. 2013. The nocturnal water cycle in midae Gray, 1825, Chiroptera: Mammalia). BMC an open-canopy forest. Journal of Geophysical Re- Evolutionary Biology 10:165. search: Atmospheres 118:10,225–10,242. De la Barrera E., Nobel P. S. 2004. Nectar: properties, Bishop J. A., Armbruster W. S. 1999. Thermoreg- floral aspects, and speculations on origin. Trends in ulatory abilities of Alaskan bees: effects of size, Plant Science 9:65–69. phylogeny and ecology. Functional Ecology 13: De la Barrera E., Pimienta-Barrios E., Schondube J. E. 711–724. 2009. Reproductive ecophysiology. Pages 301–335 Boddum T. 2013. Gall midge olfaction and its role in in Perspectives in Biophysical Plant Ecophysiology: A speciation. PhD diss., Swedish University of Agri- Tribute to Park S. Nobel, edited by E. De la Barrera cultural Sciences. and W. K. Smith. Mexico City (Mexico): Universi- Bolin J. F., Maass E., Musselman L. J. 2009. Pollina- dad Nacional Autónoma de México. tion biology of Hydnora africana Thunb. (Hydnora- de Vega C., Arista M., Ortiz P. L., Herrera C. M., Tala- ceae) in Namibia: brood-site mimicry with insect vera S. 2009. The ant-pollination system of Cytinus imprisonment. International Journal of Plant Sciences hypocistis (Cytinaceae), a Mediterranean root holo- 170:157–163. parasite. Annals of Botany 103:1065–1075. Brantjes N. B. M., Leemans J. A. A. M. 1976. Silene otites Douglas N. A., Manos P. S. 2007. Molecular phylog- (Caryophyllaceae) pollinated by nocturnal Lepi- eny of Nyctaginaceae: taxonomy, biogeography, doptera and mosquitoes. Acta Botanica Neerlandica and characters associated with a radiation of xero- 25:281–295. phytic genera in North America. American Journal Bremer B., Bremer K., Chase M. W., Fay M. F., Reveal of Botany 94:856–872. J. L., Soltis D. E., Soltis P. S., Stevens P. F. 2009. Duane W. J., Pepin N. C., Losleben M. L., Hardy D. R. An update of the Angiosperm Phylogeny Group 2008. General characteristics of temperature and classification for the orders and families of flower- humidity variability on Kilimanjaro, Tanzania. Arc- ing plants: APG III. Botanical Journal of the Linnean tic, Antartcic, and Alpine Research 40:323–334. Society 161:105–121. Duque-Buitrago C. A., Alzate-Quintero N. F., Otero J. T. Carthew S. M., Goldingay R. L. 1997. Non-flying pri- 2013. Nocturnal pollination by fungus gnats of the mates as pollinators. Trends in Ecology and Evolution Colombian endemic species, Pleurothallis marthae 12:104 –108. (Orchidaceae: Pleurothallidinae). Lankesteriana 13: Chanam J., Kasinathan S., Pramanik G. K., Jagdeesh A., 407–417. Joshi K. A., Borges R. M. 2015. Foliar extrafloral Dyer F. C. 1985. Nocturnal orientation by the Asian nectar of Humboldtia brunonis (Fabaceae), a paleo- honey bee, Apis dorsata. Animal Behaviour 33:769– tropic ant-plant, is richer than phloem sap and 774. more attractive than honeydew. Biotropica 47:1–5. Eberhard W. G., Flores C. 2002. The behavior and Clarke A., O’Connor M. I. 2014. Diet and body tem- natural history of Hybosciara gigantea (Diptera: perature in mammals and birds. Global Ecology and Sciaridae). Journal of the Kansas Entomological Society Biogeography 23:1000–1008. 75:8–15. Coberly L. C., Rausher M. D. 2003. Analysis of a chal- Eifler D. A. 1995. Patterns of plant visitation by nectar- cone synthase mutant in Ipomoea purpurea reveals feeding lizards. Oecologia 101:228–233. a novel function for flavonoids: amelioration of Endler J. A. 1993. The color of light in forests and its heat stress. Molecular Ecology 12:1113–1124. implications. Ecological Monographs 63:1–27. Coberly L. C., Rausher M. D. 2008. Pleiotropic effects Escalante-Pérez M., Heil M. 2012. Nectar secretion: its of an allele producing white flowers in Ipomoea pur- ecological context and physiological regulation. purea. Evolution 62:1076–1085. Pages 187–218 in Secretions and Exudates in Biolog- Cooper N., Thomas G. H., Venditti C., Meade A., ical Systems, edited by J. M. Vivanco and F. Baluška. Freckleton R. P. 2015. A cautionary note on the Berlin (Germany): Springer-Verlag. use of Ornstein Uhlenbeck models in macroevo- Faegri K., van der Pijl L. 1979. Principles of Pollination lutionary studies. Biological Journal of the Linnean Ecology. Third Edition. Oxford (United Kingdom): Society 118:64–77. Pergamon Press. Dafni A., Bernhardt P., Shmida A., Ivri Y., Greenbaum Fagua J. C., Gonzalez V. H. 2007. Growth rates, re- S., O’Toole Ch., Losito L. 1990. Red bowl-shaped productive phenology, and pollination ecology of 414 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

Espeletia grandiflora (Asteraceae), a giant Andean Gottsberger G., Meinke S., Porembski S. 2011. First rec- caulescent rosette. Plant Biology 9:127–135. ords of flower biology and pollination in African Feil J. P. 1992. Reproductive ecology of dioecious Si- Annonaceae: Isolona, Piptostigma, Uvariodendron, paruna (Monimiaceae) in Ecuador—a case of gall Monodora and Uvariopsis. Flora—Morphology, Distri- midge pollination. Botanical Journal of the Linnean bution, Functional Ecology of Plants 206:498–510. Society 110:171–203. Goyret J. 2010. Look and touch: multimodal sensory Fenster C. B., Armbruster W. S., Dudash M. R. 2009. control of flower inspection movements in the Specialization of flowers: is floral orientation an nocturnal hawkmoth Manduca sexta. Journal of Ex- overlooked first step? New Phytologist 183:502–506. perimental Biology 213:3676–3682. Fleming T. H., Kress W. J. 2013. The Ornaments of Life: Goyret J., Kelber A. 2011. How does a diurnal hawkmoth Coevolution and Conservation in the Tropics. Chicago find nectar? Differences in sensory control with a (Illinois): University of Chicago Press. nocturnal relative. Behavioral Ecology 22:976–984. Fleming T. H., Geiselman C., Kress W. J. 2009. The Goyret J., Markwell P. M., Raguso R. A. 2008. Context-

evolution of bat pollination: a phylogenetic per- and scale-dependent effects of floral CO2 on nec- spective. Annals of Botany 104:1017–1043. tar foraging by Manduca sexta. Proceedings of the Franz N. M. 2007. Reproductive trade-offs in a special- National Academy of Sciences of the United States of ized plant/pollinator system involving Asplundia America 105:4565–4570. uncinata Harling (Cyclanthaceae) and a derelo- Graham E. A., Andrade J. L. 2004. Drought tolerance mine flower weevil (Coleoptera: Curculionidae). associated with vertical stratification of two co- Plant Systematics and Evolution 269:183–201. occurring epiphytic bromeliads in a tropical dry Galen C. 2000. High and dry: drought stress, sex- forest. American Journal of Botany 91:699–706. allocation trade-offs, and selection on flower size Grant A. L. M. 1997. An observational study of the eve- in the alpine wildflower Polemonium viscosum (Pol- ning transition boundary-layer. Quarterly Journal of emoniaceae). American Naturalist 156:72–83. the Royal Meteorological Society 123:657–677. Galen C., Sherry R. A., Carroll A. B. 1999. Are flowers Haber W. A., Frankie G. W. 1989. A tropical hawk- physiological sinks or faucets? Costs and correlates moth community: Costa Rican dry forest Sphingi- of water use by flowers of Polemonium viscosum. dae. Biotropica 21:155–172. Oecologia 118:461–470. Hall D. R., Amarawardana L., Cross J. V., Francke W., Ganesh T., Devy M. S. 2000. Flower use by arboreal Boddum T., Hillbur Y. 2012. The chemical ecology mammals and pollination of a rain forest tree in of cecidomyiid midges (Diptera: Cecidomyiidae). south Western Ghats, India. Selbyana 21:60–65. Journal of Chemical Ecology 38:2–22. Gardner-Gee R., Beggs J. R. 2010. Challenges in food- Hartsock T. L., Nobel P. S. 1976. Watering converts

web restoration: an assessment of the restoration a CAM plant to daytime CO2 uptake. Nature 262: requirements of a honeydew–gecko trophic inter- 574 –576. action in the Auckland region, New Zealand. Res- Haston E., Richardson J. E., Stevens P. F., Chase M. W., toration Ecology 18:295–303. Harris D. J. 2009. The Linear Angiosperm Phylog- Gebhart K. A., Copeland S., Malm W. C. 2001. Diurnal eny Group (LAPG) III: a linear sequence of the and seasonal patterns in light scattering, extinc- families in APG III. Botanical Journal of the Linnean tion, and relative humidity. Atmospheric Environ- Society 161:128–131. ment 35:5177–5191. Heath A. C. G., Derraik J. G. B. 2005. Adult Diptera Girling R. D., Higbee B. S., Cardé R. T. 2013. The trapped at two heights in two native forests and an plume also rises: trajectories of pheromone plumes urban environment in New Zealand. Weta 29:22–32. issuing from point sources in an orchard canopy at Heinrich B. 1971. Temperature regulation of the night. Journal of Chemical Ecology 39:1150–1160. sphinx moth, Manduca sexta. I. Flight energetics Givnish T. J., Barfuss M. H. J., Van Ee B., Riina R., and body temperature during free and tethered Schulte K., Horres R., Gonsiska P. A., Jabaily R. S., flight. Journal of Experimental Biology 54:141–152. Crayn D. M., Smith J. A. C., Winter K., Brown G. K., Heinrich B. 1987. Thermoregulation by winter-flying Evans T. M., Holst B. K., Luther H., Till W., Zizka G., endothermic moths. Journal of Experimental Biology Berry P. E., Sytsma K. J. 2014. Adaptive radiation, 127:313–332. correlated and contingent evolution, and net spe- Heinrich B. 1993. The Hot-Blooded Insects: Strategies and cies diversification in Bromeliaceae. Molecular Phy- Mechanisms of Thermoregulation. Cambridge (Massa- logenetics and Evolution 71:55–78. chusetts): Harvard University Press. Godínez-Álvarez H. 2004. Pollination and seed dis- Heinrich B., Mommsen T. P. 1985. Flight of winter persal by lizards: a review. Revista Chilena de Historia moths near 0°C. Science 228:177–179. Natural 77:569–577. Herrera A. 2009. Crassulacean acid metabolism and Goldingay R. L. 2000. Small dasyurid marsupials—are fitness under water deficit stress: if not for carbon they effective pollinators? Australian Journal of Zo- gain, what is facultative CAM good for? Annals of ology 48:597–606. Botany 103:645–653. December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 415

Herrera C. M. 1992. Activity pattern and thermal Kelber A., Balkenius A., Warrant E. J. 2002. Scotopic biology of a day-flying hawkmoth (Macroglossum colour vision in nocturnal hawkmoths. Nature 419: stellatarum) under Mediterranean summer condi- 922–925. tions. Ecological Entomology 17:52–56. Kelber A., Warrant E. J., Pfaff M., Wallén R., Theo- Heymann E. W. 2011. Florivory, nectarivory, and pol- bald J. C., Wcislo W. T., Raguso R. A. 2006. Light lination—a review of primate-flower interactions. intensity limits foraging activity in noctur nal and Ecotropica 17:41–52. crepuscular bees. Behavioral Ecology 17:63–72. Hinchliff C. E., Smith S. A., Allman J. F., et al. 2015. Syn- Khodayar S., Kalthoff N., Fiebig-Wittmaack M., Kohler thesis of phylogeny and taxonomy into a comprehen- M. 2008. Evolution of the atmospheric boundary- sive tree of life. Proceedings of the National Academy of layer structure of an arid Andes Valley. Meteorology Sciences of the United States of America 112:12764 –12769. and Atmospheric Physics 99:181–198. Hirthe G., Porembski S. 2003. Pollination of Nymphaea Kleizen C., Midgley J., Johnson S. D. 2008. Pollination lotus (Nymphaeaceae) by rhinoceros beetles and systems of Colchicum (Colchicaceae) in Southern bees in the northeastern Ivory Coast. Plant Biology Africa: evidence for rodent pollination. Annals of 5:670–676. Botany 102:747–755. Jhumur U. S., Dötterl S., Jürgens A. 2007. Electrophys- Labandeira C. C. 2010. The pollination of mid Meso- iological and behavioural responses of mosquitoes zoic seed plants and the history of long-proboscid to volatiles of Silene otites (Caryophyllaceae). Arthro- insects. Annals of the Missouri Botanical Garden 97: pod-Plant Interactions 1:245. 469–513. Johnsen S. 2012. The Optics of Life: A Biologist’s Guide to Labandeira C. C. 2013. A paleobiologic perspective Light in Nature. Princeton (New Jersey): Princeton on plant–insect interactions. Current Opinion in University Press. Plant Biology 16:414–421. Johnsen S., Kelber A., Warrant E. J., Sweeney A. M., Lambrecht S. C. 2013. Floral water costs and size vari- Widder E. A., Lee R. L. Jr., Hernández-Andrés J. ation in the highly selfing Leptosiphon bicolor (Polem- 2006. Crepuscular and nocturnal illumination and oniaceae). International Journal of Plant Sciences 174: its effects on color perception by the nocturnal 74 –84. hawkmoth Deilephila elpenor. Journal of Experimental Lambrecht S. C., Dawson T. E. 2007. Correlated varia- Biology 209:789–800. tion of floral and leaf traits along a moisture avail- Johnson S. D., Nilsson L. A. 1999. Pollen carryover, ability gradient. Oecologia 151:574 –583. geitonogamy, and the evolution of deceptive pol- Land M. F., Gibson G., Horwood J., Zeil J. 1999. Fun- lination systems in orchids. Ecology 80:2607–2619. damental differences in the optical structure of Johnson S. D., Burgoyne P. M., Harder L. D., Dötterl S. the eyes of nocturnal and diurnal mosquitoes. Jour- 2011. Mammal pollinators lured by the scent of a nal of Comparative Physiology A: Neuroethology, Sen- parasitic plant. Proceedings of the Royal Society B: Bio- sory, Neural, and Behavioral Physiology 185:91–103. logical Sciences 278:2303–2310. Lewis T. 1967. The horizontal and vertical distribution Kato M. 1993. Floral biology of Nepenthes gracilis (Ne- of flying insects near artificial windbreaks. Annals penthaceae) in Sumatra. American Journal of Botany of Applied Biology 60:23–31. 80:924 –927. Li J.-K., Huang S.-Q. 2009. Flower thermoregulation Kato M. 1996. Plant-pollinator interactions in the un- facilitates fertilization in Asian sacred lotus. Annals derstory of a lowland mixed dipterocarp forest in of Botany 103:1159–1163. Sarawak. American Journal of Botany 83:732–743. Linsley E. G., Cazier M. A. 1970. Some competitive Kawakita A., Kato M. 2002. Floral biology and unique relationships among matinal and late afternoon pollination system of root holoparasites, Bala- foraging activities of caupolicanine bees in south- nophora kuroiwai and B. tobiracola (Balanophora- eastern Arizona (Hymenoptera, Colletidae). Jour- ceae). American Journal of Botany 89:1164 –1170. nal of the Kansas Entomological Society 43:251–261. Kawakita A., Kato M. 2004. Obligate pollination mu- Lord J. M., Huggins L., Little L. M., Tomlinson V. R. tualism in Breynia (Phyllanthaceae): further doc- 2013. Floral biology and flower visitors on subant- umentation of pollination mutualism involving arctic Campbell Island. New Zealand Journal of Bot- Epicephala moths (Gracillariidae). American Journal any 51:168–180. of Botany 91:1319–1325. Lumer C. 1980. Rodent pollination of Blakea (Melas- Keasar T., Harari A. R., Sabatinelli G., Keith D., Dafni tomataceae) in a Costa Rican cloud forest. Brit- A., Shavit O., Zylbertal A., Shmida A. 2010. Red tonia 32:512–517. anemone guild flowers as focal places for mating Luo S.-X., Chaw S.-M., Zhang D., Renner S. S. 2010. and feeding by Levant glaphyrid beetles. Biological Flower heating following anthesis and the evolu- Journal of the Linnean Society 99:808–817. tion of gall midge pollination in Schisandraceae. Kelber A., Roth L. S. V. 2006. Nocturnal colour vi- American Journal of Botany 97:1220–1228. sion—not as rare as we might think. Journal of Ex- Lüttge U. 2004. Ecophysiology of crassulacean acid perimental Biology 209:781–788. metabolism (CAM). Annals of Botany 93:629–652. 416 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

Lüttge U. 2010. Ability of crassulacean acid metab- chid Coelogyne fimbriataens. Entomological Science 15: olism plants to overcome interacting stresses in 253–256. tropical environments. AoB PLANTS 2010:plq005. Newstrom L., Robertson A. 2005. Progress in under- Maia A. C. D., Dötterl S., Kaiser R., Silberbauer- standing pollination systems in New Zealand. New Gottsberger I., Teichert H., Gibernau M., do Zealand Journal of Botany 43:1–59. Amaral Ferraz Navarro D. M., Schlindwein C., Nobel P. S. 1977. Water relations of flowering of Agave Gottsberger G. 2012. The key role of 4-methyl-5- deserti. Botanical Gazette 138:1–6. vinylthiazole in the attraction of scarab beetle pol- Norio T. 2004. Pollination of Barringtonia racemosa (Lecy- linators: a unique olfactory floral signal shared by thidaceae) by moths on Iriomote Island, Japan. Annonaceae and Araceae. Journal of Chemical Ecol- Annals of the Tsukuba Botanical Garden 23:17–20.

ogy 38:1072–1080. Osborne C. P., Sack L. 2012. Evolution of C4 plants:

Mao Y.-Y., Huang S.-Q. 2009. Pollen resistance to water a new hypothesis for an interaction of CO2 and in 80 angiosperm species: flower structures pro- water relations mediated by plant hydraulics. Phil- tect rain-susceptible pollen. New Phytologist 183: osophical Transactions of the Royal Society B: Biological 892–899. Sciences 367:583–600. Martínez-Harms J., Vorobyev M., Schorn J., Shmida Pagel M. 1999. The maximum likelihood approach A., Keasar T., Homberg U., Schmeling F., Menzel to reconstructing ancestral character states of dis- R. 2012. Evidence of red sensitive photoreceptors crete characters on phylogenies. Systematic Biology in Pygopleurus israelitus (Glaphyridae: Coleoptera) 48:612–622. and its implications for beetle pollination in the Patiny S. 2012. Evolution of Plant-Pollinator Relation- southeast Mediterranean. Journal of Comparative ships. Cambridge (United Kingdom): Cambridge Physiology A: Neuroethology, Sensory, Neural, and Be- University Press. havioral Physiology 198:451–463. Pellmyr O. 2003. Yuccas, yucca moths, and coevolu- Martins D. J., Johnson S. D. 2007. Hawkmoth pollination: a review. Annals of the Missouri Botanical Gar- tion of aerangoid orchids in Kenya, with special den 90:35–55. reference to nectar sugar concentration gradients Peng R. K., Sutton S. L., Fletcher C. R. 1992. Spatial in the floral spurs. American Journal of Botany 94:650– and temporal distribution patterns of flying Dip- 659. tera. Journal of Zoology 228:329–340. McCay M. G. 2003. Winds under the rain forest can- Pereira J., Schlindwein C., Antonini Y., Maia A. C. D., opy: the aerodynamic environment of gliding tree Dötterl S., Martins C., do Amaral Ferraz Navarro frogs. Biotropica 35:94 –102. D. M., Oliveira R. 2014. Philodendron adamantinum Micheneau C., Fournel J., Warren B. H., Hugel S., (Araceae) lures its single cyclocephaline scarab pol- Gauvin-Bialecki A., Pailler T., Strasberg D., Chase linator with specific dominant floral scent volatiles. M. W. 2010. Orthoptera, a new order of pollinator. Biological Journal of the Linnean Society 111:679-691. Annals of Botany 105:355–364. Primack R. B. 1983. Insect pollination in the New Zea- Mohan Ram H. Y., Ramanuja Rao I. V. 1984. Physiol- land mountain flora. New Zealand Journal of Botany ogy of flower bud growth and opening. Proceedings 21:317–333. of the Indian Academy of Sciences (Plant Sciences) 93: Punekar S. A., Kumaran K. P. N. 2010. Pollen mor- 253–274. phology and pollination ecology of Amorphophallus Monahan A. H., He Y., McFarlane N., Dai A. 2011. species from North Western Ghats and Konkan re- The probability distribution of land surface wind gion of India. Flora—Morphology, Distribution, Func- speeds. Journal of Climate 24:3892–3909. tional Ecology of Plants 205:326–336. Morgan K. R. 1987. Temperature regulation, energy Raguso R. A., Willis M. A. 2003. Hawkmoth pollina- metabolism and mate-searching in rain beetles tion in Arizona’s Sonoran Desert: behavioral re- (Pleocoma spp.), winter-active, endothermic scarabs sponses to floral traits. Pages 43–65 in Butterflies: (Coleoptera). Journal of Experimental Biology 128: Evolution and Ecology Taking Flight, edited by C. L. 107–122. Boggs, W. B. Watt, and P. R. Ehrlich. Chicago (Illi- Nadeau D. F., Pardviak E. R., Higgins C. W., Fernando nois): University of Chicago Press. H. J. S., Parlange M. B. 2011. A simple model for Rentsch J. D., Leebens-Mack J. 2014. Yucca aloifolia the afternoon and early evening decay of con- (Asparagaceae) opts out of an obligate pollination vective turbulence over different land surfaces. mutualism. American Journal of Botany 101:2062– Boundary-Layer Meteorology 141:301. 2067. Nadkarni N. M., Merwin M. C., Nieder J. 2001. Forest Riffell J. A., Alarcón R. 2013. Multimodal floral sig- canopies, plant diversity. Pages 27–40 in Encyclope- nals and moth foraging decisions. PLOS ONE 8: dia of Biodiversity, Volume 3, edited by S. A. Levin. e72809. San Diego (California): Academic Press. Roubik D. W. 1993. Tropical pollinators in the canopy Nakase Y., Kato M. 2012. A nocturnal Provespa wasp and understory: field data and theory for stratum species as the probable pollinator of epiphytic or- “preferences.” Journal of Insect Behavior 6:659–673. December 2016 NOCTURNAL AND CREPUSCULAR POLLINATION 417

Sage R. F. 2001. Environmental and evolutionary pre- ability to moths. Journal of Chemical Ecology 30:

conditions for the origin and diversiﬁcation of the C4 1285–1288. photosynthetic syndrome. Plant Biology 3:202–213. Tschapka M., von Helversen O. 1999. Pollinators of Sage R. F., Khoshravesh R., Sage T. L. 2014. From syntopic Marcgravia species in Costa Rican lowland

proto-Kranz to C4 Kranz: building the bridge to rain forest: bats and opossums. Plant Biology 1:382–

C4 photosynthesis. Journal of Experimental Botany 388. 65:3341–3356. Turner R. C., Midgley J. J., Johnson S. D. 2011. Ev- Sakai S., Momose K., Yumoto T., Kato M., Inoue T. idence for rodent pollination in Erica hanekomii 1999. Beetle pollination of Shorea parvifolia (sec- (Ericaceae). Botanical Journal of the Linnean Society tion Mutica, Dipterocarpaceae) in a general flow- 166:163–170. ering period in Sarawak, Malaysia. American Journal Valdivia C. E., Niemeyer H. M. 2006. Do floral syn- of Botany 86:62–69. dromes predict specialisation in plant pollination Sedlák P., Aubinet M., Heinesch B., Janouš D., Pavelka systems? Assessment of diurnal and nocturnal pol- M., Potužníková K., Yernaux M. 2010. Night-time lination of Escallonia myrtoidea. New Zealand Journal airflow in a forest canopy near a mountain crest. of Botany 44:135–141. Agricultural and Forest Meteorology 150:736–744. Vlasáková B., Kalinová B., Gustafsson M. H. G., Silvera K., Neubig K. M., Whitten W. M., Williams Teichert H. 2008. Cockroaches as pollinators of N. H., Winter K., Cushman J. C. 2010. Evolution Clusia aff. sellowiana (Clusiaceae) on inselbergs in along the crassulacean acid metabolism contin- French Guiana. Annals of Botany 102:295–304. uum. Functional Plant Biology 37:995–1010. Vogel S., Martens J. 2000. A survey of the function of Simon R., Holderied M. W., Koch C. U., von Helversen the lethal kettle traps of Arisaema (Araceae), with O. 2011. Floral acoustics: conspicuous echoes of records of pollinating fungus gnats from Nepal. a dish-shaped leaf attract bat pollinators. Science Botanical Journal of the Linnean Society 133:61–100. 333:631–633. Voigt C. C., Speakman J. R. 2007. Nectar-feeding bats Sinclair B. J., Chown S. L. 2005. Climatic variability fuel their high metabolism directly with ex ogenous and hemispheric differences in insect cold toler- carbohydrates. Functional Ecology 21:913–921. ance: support from southern Africa. Functional Ecol- von Arx M., Goyret J., Davidowitz G., Raguso R. A. ogy 19:214 –221. 2012. Floral humidity as a reliable sensory cue for Sinclair B. J., Addo-Bediako A., Chown S. L. 2003. Cli- profitability assessment by nectar-foraging hawk- matic variability and the evolution of insect freeze moths. Proceedings of the National Academy of Sciences tolerance. Biological Reviews 78:181–195. of the United States of America 109:9471–9476. Somanathan H., Borges R. M. 2001. Nocturnal pol- von Helversen D., von Helversen O. 1999. Acoustic lination by the carpenter bee Xylocopa tenuiscapa guide in a bat-pollinated flower. Nature 398:759–760. (Apidae) and the effect of floral display on fruit von Helversen O. 1993. Adaptations of flowers to set in Heterophragma quadriloculare (Bignoniaceae) pollination by glossophagine bats. Pages 41–59 in in India. Biotropica 33:78–89. Animal-Plant Interactions in Tropical Environments, Somanathan H., Borges R. M., Warrant E. J., Kelber edited by W. Barthlott, C. M. Naumann, K. Schmidt- A. 2008a. Visual ecology of Indian carpenter bees Loske, and K.-L. Schuchmann. Bonn (Germany): I: light intensities and flight activity. Journal of Com- Zoologisches Forschungsinstitut und Museum Al- parative Physiology A: Neuroethology, Sensory, Neural, exander Koenig. and Behavioral Physiology 194:97–107. Warrant E. J. 2008. Nocturnal vision. Pages 53–86 in Somanathan H., Borges R. M., Warrant E. J., Kelber A. The Senses: A Comprehensive Reference, Volume 2, Vi- 2008b. Nocturnal carpenter bees learn landmark sion II, edited by A. I. Basbaum, A. Kaneko, G. M. colours in starlight. Current Biology 18:R996–R997. Shepherd, G. Westheimer, T. D. Albright, R. H. Stevens P. F. 2001. Angiosperm Phylogeny Website, Ver- Masland, P. Dallos, D. Oertel, S. Firestein, G. K. sion 9, June 2008. http://www.mobot.org/mobot Beauchamp, M. C. Bushnell, J. H. Kass, and E. /research/apweb/. Gardner. San Diego (California): Academic Press. Takács S., Bottomley H., Andreller I., Zahradnik T., Warrant E. J., Dacke M. 2011. Vision and visual navi- Schwarz J., Bennett R., Strong W., Gries G. 2009. gation in nocturnal insects. Annual Review of Ento- Infrared radiation from hot cones on cool conifers mology 56:239–254. attracts seed-feeding insects. Proceedings of the Royal Waser N. M., Ollerton J. 2006. Plant-Pollinator Interac- Society B: Biological Sciences 276:649–655. tions: From Specialization to Generalization. Chicago Thien L. B., Bernhardt P., Devall M. S., Chen Z.-D., (Illinois): University of Chicago Press. Luo Y.-B., Fan J.-H., Yuan L.-C., Williams J. H. 2009. Watson L., Dallwitz M. J. 1992. The families of flower- Pollination biology of basal angiosperms (ANITA ing plants: descriptions, illustrations, identification, grade). American Journal of Botany 96:166–182. and information retrieval. http://delta-intkey.com. Thom C., Guerenstein P. G., Mechaber W. L., Hilde- Wcislo W. T., Arneson L., Roesch K., Gonzalez V.,

brand J. G. 2004. Floral CO2 reveals ﬂower pro ﬁt- Smith A., Fernández H. 2004. The evolution of 418 THE QUARTERLY REVIEW OF BIOLOGY Volume 91

nocturnal behaviour in sweat bees, Megalopta gena- Winter Y., Stich K. P. 2005. Foraging in a complex nat- lis and M. ecuadoria (Hymenoptera: Halictidae): an uralistic environment: capacity of spatial working escape from competitors and enemies? Biological memory in flower bats. Journal of Experimental Biol- Journal of the Linnean Society 83:377–387. ogy 208:539–548. Webb C. O., Donoghue M. J. 2005. Phylomatic: tree Winter Y., von Helversen O. 2001. Bats as pollinators: assembly for applied phylogenetics. Molecular Ecol- foraging energetics and floral adaptations. Pages ogy Notes 5:181–183. 148–170 in Cognitive Ecology of Pollination, edited by Webber B. L., Curtis A. S. O., Cassis G., Woodrow I. E. L. Chittka and J. D. Thomson. Cambridge (United 2008. Flowering morphology, phenology and flower Kingdom): Cambridge University Press. visitors of the Australian rainforest tree Ryparosa Winter Y., López J., von Helversen O. 2003. Ultravio- kurrangii (Achariaceae). Australian Entomologist 35: let vision in a bat. Nature 425:612–614. 7–17. Wolf B. O., Hatch K. A. 2011. Aloe nectar, birds and Wharton D. A. 2011. Cold tolerance of New Zealand stable isotopes: opportunities for quantifying tro- alpine insects. Journal of Insect Physiology 57:1090– phic interactions. Ibis 153:1–3. 1095. Yuan L.-C., Luo Y.-B., Thien L. B., Fan J.-H., Xu H.-L., Whiley A. W., Chapman K. R., Saranah J. B. 1988. Yukawa J., Chen Z.-D. 2008. Pollination of Kadsura Water loss by floral structures of avocado (Persea longipedunculata (Schisandraceae), a monoecious americana cv. Fuerte) during flowering. Australian basal angiosperm, by female, pollen-eating Megom- Journal of Agricultural Research 39:457–467. mata sp. (Cecidomyiidae: Diptera) in China. Bio- Whitaker A. H. 1987. The roles of lizards in New Zea- logical Journal of the Linnean Society 93:523–536. land plant reproductive strategies. New Zealand Zahradnik T., Takács S., Strong W., Bennett R., Kuz- Journal of Botany 25:315–328. min A., Gries G. 2012. Douglas-fir cone gall midges Willis M. A. 2008. Odor plumes and animal orienta- respond to shape and infrared wavelength attri- tion. Pages 771–782 in The Senses: A Comprehensive butes of host tree branches. Canadian Entomologist Reference, Volume 4: Olfaction and Taste, edited by 144:658–666. S. Firestein and G. K. Beauchamp. San Diego (Cal- Zotz G., Hietz P. 2001. The physiological ecology of ifornia): Academic Press. vascular epiphytes: current knowledge, open ques- Willmer P., Stone G. 1997. Temperature and water tions. Journal of Experimental Botany 52:2067–2078. relations in desert bees. Journal of Thermal Biology 22:453–465. Handling Editor: James D. Thomson