Field-notes on the biology of Pterocelastrus tricuspidatus (Celastraceae), highlighting the puzzle of seed-dispersal

@dianastuder @tonyrebelo @jeremygilmore @ludwig_muller @botaneek @troos @benjamin_walton @vynbos @yvettevanwijk1941 @richardgill @dhoare @craigpeter @adriaan_grobler @koosretief @chris_whitehouse @robertarcher397 @sandraf @elmarvrooyen @geoffnichols @strandloper @alastairpotts

In and around May 2000, I visited West Coast National Park (https://en.wikipedia.org/wiki/West_Coast_National_Park), Gansbaai (https://en.wikipedia.org/wiki/Gansbaai), Nature's Valley (https://en.wikipedia.org/wiki/Nature%27s_Valley), and Addo National Park (https://en.wikipedia.org/wiki/Addo_Elephant_National_Park).

I made the following field-notes on Pterocelastrus tricuspidatus (https://www.inaturalist.org/taxa/526208-Pterocelastrus-tricuspidatus and http://pza.sanbi.org/pterocelastrus-tricuspidatus).

Growth-form:

This is a shrub rather than a tree. It regenerates vegetatively (https://www.inaturalist.org/observations/14285413) rather than germinatively.

In West Coast National Park and at Gansbaai, the plants are only about 2 m high (https://www.inaturalist.org/observations/40291312 and https://www.inaturalist.org/observations/30876554).

At Nature's Valley, the plants are generally >3 m high. Presumably old specimens, 6 m high, are still multi-stemmed. The species is common in the afromontane forests at Knysna and Tsitsikamma (https://en.wikipedia.org/wiki/Knysna%E2%80%93Amatole_montane_forests), as well as occurring in coastal scrub.

The reddish wood is hard and dense (https://knysnawoodworkers.co.za/articles/characteristics-of-our-indigenous-trees/candlewood-pterocelastrus-tricuspidatus/#:~:text=The%20wood%20is%20beautifully%20marked,wood%20produces%20striking%20cathedral%20figuring.&text=The%20wood%20saws%20easily%2C%20despite,a%20smooth%20finish%20is%20obtainable.), similar to that of 'dwarf ebonies'. The species is reputed to be slow-growing, which also implies great longevity.

Leaf-size is variable. It is relatively small in West Coast National Park. However, it is relatively large at Nature's Valley (e.g. width 4.5 cm, https://www.inaturalist.org/observations/10785564 and https://www.inaturalist.org/observations/10947188). This holds true even in an extremely windy (but somehow not halophytic) situation, where P. tricuspidatus grows as part of a wind-shorn mat/hedge, 0.5 m high, of intermingled spp. of plants, at the edge of a pebble beach, e.g. near Salt River (https://en.wikipedia.org/wiki/Salt_River_(Nature%27s_Valley)).

Leaf texture:

The leaves are notably thickened. However, they conform to neither sclerophylly nor fleshiness.

The leaf is brittle, and can be carved easily with the thumbnail, at least in the case of the large-leafed form of the fire-free littoral. This differs from coexisting Sideroxylon inerme, which is truly semi-sclerophyllous (leaf is dry and not similarly brittle), even in littoral situations.

I would place the leaves of P. tricuspidatus as belonging to a separate category, 'designed to attract neither herbivory nor fire', but also palatable to Loxodonta africana (von Gadow, 1973) and Sylvicapra grimmia grimmia (https://www.inaturalist.org/observations?taxon_id=698822). In West Coast National Park, this plant is a preferred food of the latter antelope (Oom Dawid Bester, pers. comm., please scroll to the end of https://thenaturecollege.com/team/).

Flammability and relationship to wildfire:

The foliage of P. tricuspidatus is very flammable (Oom Dawid Bester, pers. comm.), which is puzzling in terms of adaptation.

I accept that this species is part of a fire cycle in relatively open forms of strandveld, as in West Coast National Park. However, in the area of Nature's Valley, the plant is free of wildfire, except perhaps once per century or so. I crushed the leaves, but could not detect any resinous smell. So, overall, I do not see that this species can be regarded as a fire-adapted plant, in the sense of Australian counterparts with a 'mallee' growth-form.

The literature states that the wood is very flammable, hence the name 'candlewood'. How does this work (resin canals)?

Fruiting phenology:

At Nature's Valley, the plants were in fruit, nearly full-size but premature, greenish-yellow, with a few flowers still apparent.

The episodicity of fruiting is striking.

I have been in the habitat of P. tricuspidatus repeatedly over the years, but have not encountered the fruits in any numbers until this autumn of 2000, leading me to realise that this is a 'mast-fruiting' species. A reference on the forests of the southern Cape states that the species flowers every year but only produces fruit every 2-3 years; I suspect that thus understates the episodicity of the fruiting. I found that the season at Nature's Valley was one month or more later than at Gansbaai. In retrospect, I should have looked under the shrubs in West Coast National Park for fallen fruits, since I saw the plants in full fruit only at Langebaanweg, leaving me wondering whether, in this Park, its fruiting was over, yet to come that year, or a non-event here despite its mast-fruiting elsewhere.

The bright hue of the fruits is conspicuous, particularly when the fruits are massed (https://www.inaturalist.org/observations/10896818 and https://www.inaturalist.org/observations/28484135 and https://www.inaturalist.org/observations/15940165), as they are in a 'mast' year. I now realise that the fully ripe fruits go beyond yellow-orange, to orange with a tinge of red. There is no pre-ripe display.

I noted that P. tricuspidatus was in fruit near Sedgefield (location near https://www.inaturalist.org/observations/11025916) in mid-September 2000.

Seed-dispersal:

This remains unresolved.

I made a special trip to ask Jack Skead (https://en.wikipedia.org/wiki/CJ_Skead) which spp. of birds eat the fruits of P. tricuspidatus, but he did not know. I also asked Carl Vernon (https://www.tandfonline.com/doi/pdf/10.2989/00306525.2021.1960020 and https://www.facebook.com/victorianheritage1905/posts/6202846636466994), who had years of experience in the habitat of P. tricuspidatus, and he did not know, apart from confirming that he had never seen any bird taking any interest in the fruits.

Books on southern African trees state that this species is difficult to grow, being a poor germinator. Oom Dawid Bester agreed, telling me 'it is as if the seed is dead'.

Rowan (1983, https://www.tandfonline.com/doi/abs/10.1080/00306525.1984.9634592?journalCode=tost20) mentions that Phillips recorded Tauraco corythaix (https://www.inaturalist.org/taxa/7205-Tauraco-corythaix) as eating the fruits of the genus Pterocelastrus. However, this leaves me wondering about the differences between rostratus (https://www.inaturalist.org/taxa/592662-Pterocelastrus-rostratus) and tricuspidatus, and whether the fruits were eaten ripe or green. After all, T. corythaix eats not only fruits but also buds of Ficus and flowers of Cussonia.

Phillips also compiled an extensive list of the spp. of fruits eaten by Columba arquatrix (https://www.inaturalist.org/taxa/3012-Columba-arquatrix), and Pterocelastrus is not among them. However, this fruit-eating columbid deserves further investigation.

Pterocelastrus tricuspidatus has repeatedly been recorded in the diet of Loxodonta africana near Knysna. This includes ripe, fallen fruits (https://www.inaturalist.org/journal/milewski/61207-plants-eaten-by-the-savannah-elephant-in-the-cape-floristic-region-part-2#).

However, it is hard to see this proboscidean as the main disperser and sower of the seeds. This is because

  • germination of P. tricuspidatus has not been observed in the faeces, and
  • L. africana is unlikely to be sensitive to the orange/red hues of the ripe fruits, and probably forages olfactorily rather than visually, anyway.

At Fairhill Private Nature Reserve (https://fairhill.co.za/) near Gansbaai, I watched several individuals of Taurotragus oryx oryx eating the ripe fruits of P. tricuspidata. I examined the fresh faeces for seeds. I found many fully intact seeds of Euclea racemosa, which was also in 'mast-fruit' at this location at this time. The landowner, Val Deverson, confirmed that she had observed the animals eating the ripe fruits of E. racemosa. However, I failed to find seeds of P. tricuspidatus in the faeces.

Habitat:

The enigma of dispersal and sowing of P. tricuspidatus is intriguing, because this plant is relatively widespread, and dominant in considerable areas of forest and scrub, in southernmost South Africa. A guide-booklet for Swartkops Valley Bushveld (Urton 1993, https://books.google.com.au/books/about/Plants_of_the_Swartkops_Valley_Bushveld.html?id=dTVFAAAAYAAJ&redir_esc=y) lists P. tricuspidatus, Euclea racemosa, and Sideroxylon inerme, among others, as typical members of this vegetation type.

I do not recall seeing P. tricuspidata in Addo National Park, but I suspect that it occurs in Bontveld on relatively poor soils on slightly elevated ground here. Readers, please check https://www.researchgate.net/publication/47296499_A_systematic_check_list_of_flowering_plants_collected_in_the_Addo_Elephant_National_Park.

Pterocelastrus tricuspidatus is one of the few spp. shared between dry strandveld (https://journals.co.za/doi/pdf/10.10520/AJA00423203_2374) in West Coast National Park (average rainfall about 300 mm per year) and the Tsitsikamma forest (>800 mm per year, https://en.wikipedia.org/wiki/Tsitsikamma_National_Park#).

In West Coast National Park, Boucher and Jarman (1977, https://www.tandfonline.com/doi/abs/10.1080/00359197709519916) show it to be one of the main components of broad-leafed scrub on sand, extending on to granite but not on to limestone, where Searsia longispina instead occurs (https://www.inaturalist.org/taxa/593892-Searsia-longispina).

Comparison of fruit with that of Putterlickia pyracantha:

Putterlickia pyracantha (https://www.inaturalist.org/taxa/524315-Putterlickia-pyracantha) coexists with P. tricuspidatus in coastal scrubby vegetation (https://www.inaturalist.org/observations/140781514), and belongs to the same family (Celastraceae). It was concurrently in fruit at West Coast National Park and Nature's Valley in 2000.

There are many photos of the fruits of both P. tricuspidatus and P. pyracantha (e.g. https://www.inaturalist.org/observations/127498576 and https://www.inaturalist.org/observations/121068064 and https://www.inaturalist.org/observations/23509411) in iNaturalist. What is immediately obvious is that dehiscence is extremely expressed in the latter, but subtle and easily overlooked in the former.

Intrigued by this, I have compiled the following comparison between the fruits of the two spp.

The fruit of P. tricuspidatus (https://www.inaturalist.org/observations/141918224 and https://www.inaturalist.org/observations/141150869 and https://www.inaturalist.org/observations/140341231) is smaller, and usually less red, than that of P. pyracantha (https://www.inaturalist.org/observations/11091125 and https://www.inaturalist.org/observations/139127586 and https://www.inaturalist.org/observations/140150333 and https://www.inaturalist.org/observations/140643701).

Another difference is that the fruit of P. tricuspidatus is horned (https://www.inaturalist.org/observations/140571322 and https://www.inaturalist.org/observations/119132528), whereas that of P. pyracantha is smooth.

Each fruit contains only two seeds, compared to more than a dozen (wrapped in orange-hued arils) in P. pyracantha (https://www.inaturalist.org/observations/100394322 and https://www.inaturalist.org/observations/79527193).

The seeds are smaller than those of P. pyracantha, despite the difference in total number of seeds.

The fruit-wall is more succulent than that of P. pyracantha, and does not necessarily dehisce, even after one week in a packet of leafy sprigs.

The fruit is less affected by picking than are those of P. pyracantha: it retains its yellowish hue, despite drying out to non-succulence. By contrast, that of P. pyracantha dries and dehisces (https://www.inaturalist.org/observations/127662606 and https://www.inaturalist.org/observations/113436780 and https://www.inaturalist.org/observations/46214709), the outer surface of the capsule retaining a dull red hue (https://www.inaturalist.org/observations/140150330 and https://www.inaturalist.org/observations/96128646).

The aril is smaller (see second photo in https://www.steenboknaturereserve.org.za/wp-content/uploads/2021/05/Pterocelastrus-rostratus-PDF.pdf) than that of P. pyracantha (https://www.inaturalist.org/observations/133126093 and https://www.inaturalist.org/observations/141674995 and https://www.inaturalist.org/observations/82469682), and does not cover the seed entirely. The aril of P. tricuspidatus is more oily and less succulent than that of P. pyracantha.

The fruit-wall of P. tricuspidatus is only weakly succulent, loses some of its moisture if/when it dehisces, and contains arils that are non-succulent (although possibly moister than that of Acacia cyclops, https://www.agefotostock.com/age/en/Stock-Images/acacia-cyclop-aril.html). By contrast, in P. pyracantha, the fruit-wall is non-succulent (drying to definitely non-succulent when dehisced), but the copious arils are clearly succulent (and non-oily).

In P. pyracantha, the conspicuous hues (red outer surface of capsule, bright orange aril) and arillate structure of the fruit are clearly adaptive to dispersal and sowing by birds, such as Pycnonotidae (e.g. https://www.inaturalist.org/observations/35691747) and Coliidae (e.g. https://www.inaturalist.org/observations/135258890). Although the ripening process may seem complicated in transforming the fruit from https://www.inaturalist.org/observations/80707225 to https://www.inaturalist.org/observations/76068591, the identities of the agents of dispersal are fairly straightforward.

However, in the case of P. tricuspidatus, the nature of the adaptation may remain as puzzling today as it was two decades ago, when I wrote my field-notes. Does any reader know better?

Publicado el 20 de noviembre de 2022 por milewski milewski

Comentarios

Agregar un comentario

Acceder o Crear una cuenta para agregar comentarios.