Topic 5.5: Plant Reproduction

Trees selected for phenological study should represent a range of sizes so as to allow estimation of the size of first reproduction.  In some forests in which harvesting is controlled simply by setting a minimum diameter for harvesting, the size at which trees commence fruiting and the minimum harvest size are often quite similar.  This is the case, for example, in Indonesia where the minimum felling diameter is often set at 60 cm; in nearby Peninsular Malaysia, the minimum diameter is generally 45 cm.  Commercially valuable tree species in both Malaysia and Indonesia generally start reproducing when somewhat smaller than these critical diameters, but species and site-specific data are rare and regeneration failures are common after harvesting.

Determining how crop species are pollinated is important because if the pollen vectors suffer under a particular management regime, future crops may be jeopardized.  For example, frequent burning may reduce populations of the bees that pollinate Brazil nut trees in the Amazon.  The actual field studies involved can simply entail observation of flower visitors, but the observer will quickly realize that many of the visitors are not pollinators.  Controlled experiments in which the possible pollen vectors are afforded selective access to virgin flowers can be difficult when dealing with large trees, but are nonetheless worthwhile.  Such studies in lowland forest dominated by trees in the Dipterocarpaceae in Malaysia, for example, revealed that several species share the same tiny insect pollinators (thrips) but avoid inter-specific pollen transfer by flowering at different times.  The researchers even suggested that successful reproduction of many important timber species may depend on the presence of species that flower earlier in the season; pollinator populations increase on earlier flowering species that tend to be less dependent on insect vectors for their pollination.

Most canopy trees in tropical forests are pollinated by small insects (most commonly bees and flies) that are unlikely to diminish greatly in number as a result of forest management operations.  There are enough exceptions to this rule, however, that it would be irresponsible to rely on assumptions when the necessary research is so straightforward and inexpensive.

Studies of the reproductive biology in many different parts of the tropics have revealed that as many as 50% of the tree species can be dioecious, i.e., have separate male and female trees.  Whether or not a tree species is dioecious can be of critical importance when reserving seed trees (half of which are likely to be male) and when trying to determine the minimum population size necessary to assure future harvests.

Forests dominated by members of the Dipterocarpaceae provide the best known but by no means unique example of the phenomenon of “masting”, super-annual bouts of reproduction.  Flowering and fruiting vary substantially between years in most species, but masting species characteristically produce few if any fruits in non-mast years.  In many Dipterocarpaceae, reproduction generally occurs at intervals of three to as many as seven years; flowering apparently occurs in response to unique climatological conditions, particularly long periods of clear, rainless weather, and may perhaps be triggered by low temperature events.  Masting in Dipterocarpaceae dominated forests is not by any means restricted to trees in this family; trees in other families, vines, shrubs, and even many epiphytes flower more intensively during mast years.  Masting in forests elsewhere in the world may be less spectacular than in Southeast Asia, but nonetheless represents a challenge for forest managers.  Lack of assurance of year-to-year availability of seeds wreaks havoc on replanting programs and often complicates silvicultural prescriptions.

Although we generally consider fruiting to inexorably follow flowering, this is unfortunately often not the case due to intervention by pests, pathogens, and inclement weather (e.g., storms or severe droughts).  In the few tree species that have been studied, often <5% of the flowers develop into fruits in some years and for some species, flowering fails to yield any fruits at all!  Insects that consume developing seeds and fruits are often responsible for fruiting failures; larvae that hatch from eggs laid on young fruits or flower ovaries bore into the seeds and consume the endosperm and other materials allocated to the developing embryo.  In many cases, the fruits and seeds continue to enlarge and appear normal on the outside, but are not viable.  Filled seeds, as opposed to those that contain an insect, can sometimes be detected by water emersion; insect-infested seeds tend to float.   Fumigating batches of seeds with insecticide may reduce the severity of attack, especially when seeds are intended for storage and with insect species capable of emerging from one seed and ovipositing on others in the same batch.  In forests regenerated naturally (i.e., not by planting), there is no obvious way to control pre­dispersal seed predators, but awareness of their presence could nevertheless be important.

Seeds are dispersed by a variety of biotic and abiotic vectors (e.g., birds, mammals, wind, and water) but few seeds are carried far from the mother tree.  The few “seed shadows” that have been quantitatively described generally show an exponentially diminishing number of seeds with increasing distance from the seed producer, with few seeds reaching more than 25 m from the crown edge.  Large seeds are particularly poorly dispersed, except those that are moved by water.  Information about seed dispersal distances can provide a preliminary basis for calculating the maximum spacing between seed trees to be retained during a logging operation.  For example, if at distances exceeding 20 m from the trunk of fruiting trees the seed “rain” averages less than 1 seed per square meter (the minimum density for full stocking), then at least eight evenly-spaced seed-producing trees need to be retained per hectare to assure full coverage with seeds.  For dioecious trees for which the sex is unknown, at least twice the number of potential seed trees needs to be retained. Seed predation continues after dispersal.  Many species of insects, birds, and mammals are obligate or facultative granivores.  The loss of entire seed crops to predators is not unusual, especially with large seeded species.  The preference of so many animal species for seeds is understandable given that the food value of seeds (e.g., caloric and protein contents) are generally much higher than vegetative tissue.  Furthermore, seeds can often be more easily transported and stored than leaves.

Seeds that escape pre- and post-dispersal predators may or may not germinate, depending on whether they are subjected to the appropriate environmental conditions.  Seeds of most species will germinate as long as they are warm and wet.  Scarification, either due to mechanical abrasion or the action of digestive enzymes in animal digestive tracks, stimulates seed germination in some species.  Dependence on gut passage is rare, but seeds that pass through animals or that are treated with acid to simulate this effect, tend to germinate rapidly and simultaneously, both of which may be advantages for nursery managers.  Several prominent tree species in mangrove forests produce seeds that actually germinate while still attached to the mother tree, i.e., before dispersal (e.g., Rhizophora spp., Nypa fruticans).  Other species, most notably some extremely light-demanding “pioneer” trees that dominate severely disturbed areas, have seeds that germinate only when exposed to direct sun.  Seeds of these “photoblastic” species contain phytochrome, a light-sensitive hormone that stimulates germination only when exposed to light that has a high ratio of red to far red wavelengths.  Since photosynthetic tissues absorb more red light than far red, light passing through leaves is depleted of red wavelengths.  In other words, the seeds detect whether there is sufficient light for establishment and growth not by measuring light intensity directly, but by measuring a characteristic of light that is strongly correlated with intensity.