Start PagePlacentationBreeding Behaviour


Although there are frequent insect visitors to Philodendron, especially small Hemiptera in the genus Neelia which neither appear to never feed nor mate on the inflorescences (H. Young, pers. comm.) only the larger beetles are known to be pollinators. The system of pollination is nearly identical to that shown for Dieffenbachia (Croat, 1983a; Young, 1986, 1990). Pollinators are members of subfamily Dynastinae in the family Scarabaeidae (Fig. 32). All beetles determined to date from either Central American or South American Philodendron are members of the genera Cyclcephala or Erioscelis. Some species of beetles are not particularly host-specific, visiting members of both P. subg. Philodendron and P. subg. Pteromischum and sometimes even other genera such as Dieffenbachia, Homalomena, Syngonium, and Xanthosoma and sometimes even other families including some palms (Arecaceae), Cyclanthus bipartitus (Cyclanthaceae), as well as Annona and Cymbopetalum in the Annonaceae (M. Grayum, pers. comm.; Schatz, 1990). Cyclocephala negerrima Bates, for example, has been found visiting both P. brenesii and P. tysonii but also P. standleyi a member of P. subg. Pteromischum. Though beetles are not very species-specific pollinators, individual beetles of some species tend to be somewhat strata-specific, visiting only those species which are in a certain range above the ground (Schatz, loc. cit.; Helen Young, pers. comm.). The beetles are attracted to the Philodendron inflorescence, usually late in the day or at dusk. Attractants are apparently a combination of a scent (at least in many cases), a source of food (oil-bearing sterile staminate flowers), and shelter.

Scents produced by Philodendron species are not always obvious, at least in P. subgen. Philodendron. George Schatz & Helen Young, pers. comm.) (unpublished data) has documented floral odors for species of Philodendron and has identified the principal constituents of these aromas. Some species have noticeably sweet aromas in the early evening hours while other species show no noticeable aromas, at least during the early evening hours on the first day of anthesis. Philodendron megalophyllum, a South American species, had a faint spicy aroma detectable directly at the spathe during the evening but even this faint aroma was absent the following morning on day two of the flowering event. At the same time the stigmas were soft, juicy and sticky but without an obvious sweet taste both in the evening and the following morning. Schatz (1990) believes that the pattern of visitation to Philodendron as is exhibited by beetles at La Selva in Costa Rica are to a great degree explained by odor. He points, for example, to the high degree of specificity which is exhibited by Philodendron radiatum and by and undescribed species of the beetle Cyclocephala. The aroma given off by P. radiatum is particularly unique in that it is made up of compounds that are unique to P. radiatum.

At anthesis the open spathe of Philodendron provides ample space at the base in the area surrounding the pistillate portion of the spadix. The open spathe provides a certain amount of protection, often with the spathe blade being projected slightly forward and effective in providing protection against rain falling directly into the spathe tube. Beetles typically spend the night and most of the following day inside the spathe where they remove the sticky exudate from the pistils, eat pollen (Gottsberger & Silberbauer-Gottsberger, 1991) feed on the sterile staminate flowers and use this opportunity for mating. Studying P. bipinnatifidum Gottsberger & Silberbauer-Gottsberger (1991) found that the beetles were active mostly during the first ten to twenty minutes after arrival and during the strongest production of scent. Copulation was strongest immediately after arriving and eating also subsided by the time the spadix had cooled off. The staminodia which serve as a food source for the beetles are typically high in lipids (Armstrong & Irvine, 1990) making them a high source of food value for the beetles. Old inflorescences which have been pollinated often exhibit the sterile staminate portion of the spadix entirely eaten away.

Typically the number of beetles found in a single inflorescence is modest, frequently no more than five and sometimes up to a dozen beetles. However, sometimes the numbers are simply astounding, with Gottsberger & Amaral (1984) reporting as many as 200 beetles trying to enter a single inflorescence of P. bipinnatifidum. While visiting beetles are nearly always of the same species, sometimes more than one species of beetle may be found in the inflorescence. Though beetles typically do not leave the spathe until about dusk, they will leave when the inflorescence is disturbed, for example by removing it. Beetles crawl slowly up the spathe or spadix and appear at the rim of the open spathe at which point they generally fall promptly to the ground where they disappear in the leaf litter or soil with remarkable speed.

Thermogenesis, the production of heat in the spadix by the rapid oxidation of stored starch or by lipids (Walker et al., 1983; Gottsberger, 1990), plays an important role in the pollination of Philodendron (Herk, 1937, 1937a, 1837b; Van der Pijl, 1937; Knutsen, 1974; Seymour et al., 1984; Gottsberger, 1984; Gottsberger, 1986), studying P. bipinnatifidum, have shown that, though oxidation of carbohydrates took place during preheating of the spadix, lipids were oxidized thereafter during maximum heating and that the lipids involved were consumed directly, not after conversion to carbohydrates. This makes the biochemistry of that species similar to heat production in some animals. The thermogenic reaction, occurring principally in the staminodial region of the staminate spadix, is a cyanide resistent pathway (James & Beevers, 1950; Henry & Nyns, 1975) involving the inner surface of mitochondrial membranes (Urdentlich, et al., 1991) which is triggered by an accumulation of acetosalicitic acid (Meeuse & Buggeln, 1969; Raskin et al., 1987, 1989; Meeuse, 1975, 1978; Raskin, 1992). The end result of this high rate of respiration in these plants is the production of heat rather than ATP as in animals (Meeuse, 1966). To accomplish this high increase in metabolism the plants mitochondria in the inflorescence switch to an electron transport pathway commonly refered to as the "cyanide resistant pathway".

The thermogenetic heat rise in Philodendron is sometimes dramatic with temperatures rising well above ambient temperature but the maximum temperatures are not affected by ambient temperature (Nagy, 1972; Knutsen, 1974). Gottsberger & Silberbauer-Gottsberger (1991), working with P. bipinnatifidum, reported temperatures of spadices occasionally to 46ðE C. with the highest and most efficient temperatures for the emission of scents being maintained for 20-40 minutes. Thermogensis does not create even or constant temperatures, but rather produces fluctuations depending on the time of day with definite peaks (Leick, 1910, 1916; Engler, 1920; Foster, 1949; Nagy et al., 1972; Sheridan, 1960; Gottsberger & Amaral, 1984) with the greatest temperature peaks occurring when beetle visitation is greatest and odor is most intense (Gottsberger & Amaral, 1984). Temperature peaks may occur on two or more successive days. While the increased temperature is presumably responsible for the generation of scent compounds (Nagy et al., 1972) and while the production of temperature and scent appears to be closely correlated with the peaks in temperature, there is still controversy over the exact function of the heat production at least as it pertains to genera which produce foul aromas. Moodie (1976) suggests that heat production and the higher levels of carbon dioxide production are components of a carrion, dung and mammal mimicry syndrome and that heat production alone aids in providing sufficient warmth in colder climates for the activity of pollinatin organisms. The subject of thermogenesis and its role in pollination has been reviewed in great detail by Mayo (1986), by Grayum (1990) and Bay (1995[1996]).

Those Central American species observed showed that the spathe closes slightly the morning after the first night of anthesis. The evolutionary significance of this slight closure is uncertain but the minor constriction, even in the most extreme cases, closes off only the pistillate portion. The spathe continues to close on the evening of the second day after being open for about 24 hours (see discussion below). The beetles, which may enter the spathe on the first night of opening, spend about 24 hours in the spathe tube. This partial closing process usually corresponds with staminal anthesis. The now crowded condition of the spathe tube and the probable desire on the part of the beetles to seek a new food source with the onset of dusk encourages the beetles to leave the inflorescence. Since the spathe is somewhat constricted at this stage, the departing beetles must now crawl through the constriction caused by the partially closed spathe. The beetles emerge from the spathe tube by either climbing up the side of the spadix or up the inside wall of the spathe. By the time they emerge from the spathe it is quite constricted and they must squeeze through the constriction which fits rather tightly around the spadix just above the sterile staminate spadix. In order to depart the beetles must literally crawl through the copious strands of pollen that emerge from the apical pores of the stamens. The constriction of the spathe and its corresponding constricted area on the spadix help to insure that most of the pollen which falls into the spathe tube accumulates and is carried forward and out of the spathe by the departing beetles. Araceae pollen is not very tacky and probably does not adhere well to the smooth, hard surfaces of the beetles. Alternatively the beetles themselves are usually quite sticky from the sticky sugary secretions of the styles and especially from the resin which arises usually from the inner surface of the spathe or sometimes from the spadix itself (Fig. 128). Once the beetles have emerged they fly off in search of another place to spend the night, usually another open inflorescence. The beetles apparently have a keen detection of infra-red radiation or of scent because they are often seen in a "honing-in" pattern which is quite direct to the next available inflorescence (John Rawlins, pers. comm.). Gottsberger & Silberbauer-Gottsberger (1991) reported that beetles flew in a typical zig-zag pattern toward the center of fragrance concentration, indicating that they were very sensitive to the aroma being produced. They also reported that once the beetle was in within sight of the inflorescence they shifted to a straight line of flight until they hit the inner surface of the spathe blade, whence they fell into the lower portion of the spathe. Gottsberger & Siberbauer-Gottsberger (loc. cit.) have proven experimentally in the case of P. pinnatifidum (a member of P. subg. Meconostigma) that the beetles use only visual references for location as they near the inflorescence. Shelter, warmth, food and copulation are the driving forces behind this pollination strategy and although selectivity is not perfect in such beetle-pollinated systems (Young, 1986; 1988) fruit set in undisturbed populations is high. The precision and high degree of synchrony of thermogenesis gives evidence of a highly evolved system of pollination despite the presence of the rather unsophisticated beetle pollinators.

Armbruster (1984), studying the role of resin in angiosperm pollination, has questioned the efficacy of floral resin in the transport of pollen, citing its possible toxicity and the difficulty of transporting pollen embedded in resin. While he stresses the role of resin for other purposes, mainly in nest building by bees, it must be pointed out that bees which use resin for nest building play no role whatever in Philodendron pollination. In contrast, the near universal availability of resin, its close association with pollen delivery, the non-tacky nature of Philodendron pollen and its availability only at anthesis of flowers all point to a strong role in Philodendron pollination. In those species with resiniferous spadices (Fig.128) the pollen is shed with and incorporated in the resin from the moment of thecae dehiscence. Alternatively those species which lack staminal resin and instead have resin only on the inner spathe surface have pollen presented as slender filaments.

The exact role that thermogenesis plays in the pollination of Philodendron is still poorly known and rather few plants have been studied on an experimental basis. Despite its probable occurence in all Philodendron species, most detailed studies of thermogenesis in Philodendron thus far have been made on either P. selloum or P. bipinnatifidum, now assumed by Mayo (1991) to be synonymous.

Even though all detailed observations thus far have been made with P. bipinnatifidum, there is considerable confusion regarding the results. Although he measured no heat peaks, horticulturist Ron Weeks (Homestead, Florida) reports (pers. comm.) that three members of P. subg. Philodendron, P. bipinnatifidum, P. speciosum, and P. williamsii, show no variation in the schedule of spathe opening, of plants being pollinated on the first evening of opening or in the shedding of pollen on the evening of the second day. On the other hand, he reported that P. eichleri shows great variation in opening periods, temperature changes, fragrance and pollen shed, perhaps owing to weather conditions. Scientific studies carried out on other plants in P. subg. Meconostigma showed considerable variance. Four separate and conflicting reports were made on material determined as belonging to P. pinnatifidum in Brazil. Warming (1867, 1883) reported a two day pollination event with two heat peaks (early evening and late morning respectively) with the spathe closing then reopening during the first night. Gottsberger & Amaral (1984) reported on two plants, one as P. selloum (now considered to be a synonym of P. pinnatifidum) with a 3-day pollination event with two unequal early evening heat peaks and one as P. bipinnatifidum with a 4-day pollination event with three unequal early-evening heat peaks. The spathe was not reported to close during the event. Confirming the complexity of the thermogenesis riddle is the fact that Seymour et al. (1983), studying a cultivated but similar plant, believed to be the same species (Mayo, 1986), found both types of pollination events that had previously been described by Gottsberger and Amaral but this time in a single plant! Clearly more investigation must take place, at least in P. subg. Meconostigma to determine the pollination behavior.

Leick (1916) reviewing work done by Kraus (Kraus, 1894; 1896) with P. pinnatifidum, a member of P. subg. Philodendron from Venezuela, reported a 2-day pollination event with temperature peaks in the evening of two consecutive days. While it is not certain that most Philodendron of Central America have elevated temperature on two consecutive days, the pattern in P. pinnatifidum would appear to match the events of Central American species observed in the field and under greenhouse conditions at the Missouri Botanical Garden. Further detailed studies of this phenomenon including a much broader survey of Central American species will be carried out by my student, Jane Whitehill, during graduate studies at the Missouri Botanical Garden.

My own measurements with a recording thermometer made on P. glanduliferum and P. advena in the field in Chiapas, Mexico, indicated a definite heating during the early evening hours, usually peaking between 18:00 and 19:00 hours. Phenological observations were made on many Central American species under greenhouse conditions at the Missouri Botanical Garden. Though no temperature readings were made in the greenhouse, the plants behaved in the same manner as they did in the field. The spathe began loosening during the early afternoon or evening of day one and was usually open, though not always fully open, by 18:00 hrs or sometimes as late at 19:00 hrs. Since the opening is believed to be in part light related it would probably begin earlier in tropical areas where day length is shorter. For example, P. megalophyllum, flowering in my office on May 1st in St. Louis was not fully open even at 19:00 hrs, perhaps due to the fact that at that time of the year on daylight savings time, it was still not very dark, whereas it would be fully dark by the same time of the day in Colombia where the plant originated.

Grayum's observations (Grayum, 1996) with Philodendron subg. Pteromischum showed a similar pattern with most species having the spathe beginning to loosen by early afternoon with the spathe fully open by mid afternoon. He reported that for the species of P. subg. Pteromischum that he observed the pollinators appeared at the opened inflorescence during a relatively brief time, usually between 19:00 and 19:15 hrs. An important feature in the pollination story reported by Grayum (1996) for the first time is that resin secretion from the inner spathe surface does not begin until 21:00 to 22:00 hours on the first day of anthesis and that it then continues until the end of anthesis.

Once opened, the spathes of P. subg. Philodendron apparently remained open during the night and were always open the next morning at the beginning of day two, remaining open during the course of most of the day. During the later part of the same day, usually in late afternoon of day two, the spathe begins to close and pollen begins to shed in long filaments. The spathe does not fully close at this time but remains open near the apex. It remains in this condition into the beginning of the evening of day two. By the beginning of day three the spathe is generally fully closed and the only evidence that it had ever opened is that there is often some loose pollen still remaining on the closed edges of the spathe. In addition the closed spathe is somewhat less turgid than before anthesis, sometimes allowing that it could be forced open without breaking the margins of the spathe. Doing the same with an unopened spathe is impossible without breaking the stiff and brittle spathe margins.

The entire pollination episode usually requires little more than 24 hours if you count just the time that the beetles are present. If you count the time that the spadix is to any extent open, the time could be as much as 8 hours longer, since it may open late in the afternoon and remain open for some hours after the pollen has been shed. Grayum (1996) reported that for the species of P. subg. Pteromischum studied in Costa Rica the average pollination event required about 30 hours (i.e., from spathe opening to spathe closing).

That the intensity of light must play an important role in flowering behavior was indicated by the fact that on cloudy days spathes in cultivated collections open earlier than usual, sometimes as early as noon on day one of the flowering sequence. This may reflect agreement with studies by Buggeln et al. (1971) who argue that darkness induces opening of the spathe and an elevated respiration rate in Sauromatum guttatum Schott.

Another possible difference between P. subg. Meconostigma, so adequately reported by Mayo (1986), and P. subg. Philodendron is the source of the heat, i.e., where the thermogenesis takes place on the spadix. His investigations with P. subg. Meconostigma indicated that the heat was centered on the sterile staminate section of the staminate spadix. In P. subg. Meconostigma that section is as large as or larger than the remaining fertile part of the spadix, a feature not common in P. subg. Philodendron where the staminodial segment is always a small percentage of the total spadix.

Contrary to the findings of Mayo, Leick (1916) reported that heating took place in the "middle and upper part" of the spadix, presumably indicating at least a part of the fertile staminate spadix. Concurring with this veiw, Ron Weeks (pers. comm.) reported that temperature rise occurs in both the sterile and fertile staminate portions of those species of P. subg. Meconostigma that he studied. Mayo (1986) theorized that because of the major morphological differences in the relative lengths of the sterile and fertile staminate portions of the spadix in the two subgenera that the two subgenera would likely have different thermogenetic patterns.

Breeding studies (see section on Breeding Behavior) have shown that Philodendron species have few if any barriers to cross-pollination, owing perhaps to the fact that there are other physical and temperal barriers to cross-pollination. Two species of Philodendron are seldom in flower simultaneously in a given area. Even when they are, there are other parameters which affect separation. Beetles tend to fly at certain elevations above the ground (Schatz, 1990), helping to prevent cross-pollination of species that occur at radically different strata. In addition, specific attractants, i.e., species-specific pheromones, exist in many species because most flowering species attract principally a single beetle. The sloppiness in the system, when it occurs, is owing to opportunistic beetle visitors (G. Schatz & H. Young, pers. comm.), pers. comm.) and this might produce some hybridization. Suspected hybrids, though rare, are seen in the field. It is not known if these suspected hybrids are themselves fertile and capable of reproducing.

Perhaps because of the substantial barriers already present, Philodendron appears to have developed the ability to cross between sections. In Anthurium, relatively little cross-breeding was possible between different sections in the genus (Croat, 1991). In contrast, quite unrelated species of Philodendron, even species in different sections, readily cross-pollinate and produce intermediate offspring (Keith Henderson, Cairns, Queensland, Australia, pers. comm.). For this reason phenological separation may be critical to maintaining distinct species lines. The pollinators of Philodendron, ruteline beetles, are for the most part not very species-specific and frequently switch from one species to any other in flower at the same time. Not only will some species of beetles switch from one species to the next, as is known for certain in Dieffenbachia (Young, 1986), but some individuals will also switch to another genus. For example, beetles which regularly visit D. longispatha Engl. & K. Krause at La Selva may visit P. grandipes, another species that is terrestrial and about the same height above the ground as that species (G. Schatz & H. Young, pers. comm.). Beetles are also reported to move from Dieffenbachia longispatha to Xanthosoma undipes. Some individuals of beetle species that regularly visit D. longispatha will even switch to Cyclanthus bipartitus Poit. (Cyclanthaceae). Schatz believes that this is owing to the fact the a small component of the pheromone emitted by Cyclanthus is the principal component of the scent given off by Dieffenbachia longispatha. He believes that during the end of the flowering season of Dieffenbachia and the beginning of the flowering season of Cyclanthus some confusion occurs in the pollinators behavior.

The unpublished observations of G. Schatz & H. Young (pers. comm.) and the published results reported by Helen Young (1986, 1988, 1988a) made principally with Dieffenbachia, probably are comparable to what is happening in Philodendron. Beetles which visit any particular species of Dieffenbachia are often predominantly of one species but they are often accompanied by other opportunistic species of beetles which visit occasionally as well. Schatz believes that these opportunistic species are not likely to be effective pollinators since they are so catholic in their tastes that they are not likely to make their next visit to a receptive Dieffenbachia. Alternatively Helen Young (Young, 1988) indicated that the most common species of beetles are not the most effective pollinators. However, they may be responsible for the occasional hybridization that is evident from observations on populations of Dieffenbachia, such as those at La Selva where the studies of both George Schatz and Helen Young were carried out. The pollination system described for Dieffenbachia by Young and Schatz is apparently true of Philodendron also. Schatz reported (pers. comm.) that while one undescribed beetle species (determined as Cyclocephala ampliata by H. Young), was found to visit only Philodendron radiatum it was accompanied occasionally by another more opportunistic species of beetle. Despite the presence of opportunistic beetle species, most beetle pollinators of most Philodendron species are probably much more species-specific. For example Grayum (1996), reporting on unpublished data collected by George Schatz, reported that two unrelated species of Philodendron subg. Pteromischum were pollinated by the same species of beetles and that both of the Philodendron species have floral odors featuring the same two principal components, 1,2-dimethyl-4(2-propenyl)-benzene. This leads one to the conclusion that there is a degree of specificity among pollinators for certain species based on their floral odors. In addition, in the list of pollinators known for P. subg. Philodendron (See List of Pollinators below) only three Philodendron species were observed to have more than a single species of beetles present at any one time. In each case two species of beetles were present. As can be seen from P. grandipes the beetle species need not always be the same species even if only a single species of beetle is found to be the pollinator. Though more studies must be made on pollination biology of Philodendron and even though the beetle pollination system is somewhat sloppy and imprecise, a combination of a moderately strong beetle-plant specificity, coupled with severe phenological constraints and narrow windows of pollination opportunities (perhaps as little as a few hours per year) work to reduce interspecific hybridization. Although hybrids can be readily produced under greenhouse conditions, evidence for hybridization is not apparent among wild populations

Plant species/ Beetle species Voucher Number  
P. anisotomum *Cyclocephala amblyopsis
*Erioscelis columbica Endrodi
P. brenesii Croat 35519 Cyclocephala nigerrima Bates
P. brevispathum Erioscelis proba Sharp 
P. correae Croat 66653 Cyclocephala conspicuaSharp
P. grandipes Croat 76594
Jimenez 6;

Jimenez 6;
*Cyclocephala gravis
Cyclocephala sexpunctata Castelnau

Erioscelis columbica Endrodi
P. grayumii Croat 74840
Croat 74840
Cyclocephala rubescens Bates
Cyclocephala sexpunctata Castelnau
P. jodavisianum Croat 35950 Cyclocephala ligyrina Bates
Cyclocephala mafaffa Burmeister

*Erioscelis columbica Endrodi
P. pterotum

Croat 10903

*Cyclocephala ampliata
*Cyclocephala amblyopsis

Cyclocephala ligyrina Bates

*Cyclocephala sexpunctata or C. tutilina
P. radiatum *Cyclocephala ampliata
*Erioscelis columbica Endrodi

*Cyclocephala amblyopsis

*Cyclocephalia ligyrina Bates

*Cyclocephala kaszabi
P. rothschuhianum (Young, 1987)  *Cyclocephala amblyopsis
Cyclocephala kaszabi
*Erioscelis columbica Endrodi
P. sagittifolium Thompson 4636 Cyclocephala sexpunctata Castelnau
P. schottianum Croat 36110 Cyclocephala melane Bates
P. triphyllum *Cyclocephala amblyopsis
*Cyclocephala kaszabi

*Erioscelis columbica Endrodi
P. tysonii Croat 66711 Cyclocephala nigerrima Bates
P. standleyi Thompson 5001 Cyclocephala nigerrima Bates
P. ptarianum Ramirez 1163 Cyclocephala rustica (Olivier)
P. callosum No voucher Cyclocephala rustica (Olivier)
* Note: Those entries marked with an asterisk were provided by Helen Young, Barnard College, New York and were based on unpublished observations made during 1984-1985 at La Selva (O.T.S. field station) in Costa Rica (with the assistance of George Schatz). The balance of the records were based on determinations made by John Rawlins at Carnegie Museum in Pittsburgh. Those collections by Sue Thompson were contributed independently to Rawlins.