Why does the Mediterranean Basin lack a littoral effect in the incidence of fleshy fruits?

@tonyrebelo @ludwig_muller @jeremygilmore @benjamin_walton @botaneek @adriaan_grobler @yvettevanwijk1941

In the southern parts of southern Africa and Australia, there is a noticeable pattern involving fleshy fruits in the littoral zone (https://www.merriam-webster.com/dictionary/littoral).

Many spp. of plants, representing many families, bear fleshy fruits on beach dunes, rocky promontories, and other littoral environments (within a few hundred metres of the sea) in

  • the southwestern Cape of South Africa, and, to a lesser degree,
  • southwestern Western Australia, and southern South Australia and Victoria.

This tends to contrast with comparable substrates farther inland, where fleshy fruits tend to have minor incidence, even within the same genera.

Particularly striking examples of this pattern include:

However, there is no similar pattern in the Mediterranean Basin.

The only species with fleshy fruits that is consistently associated with the littoral is Corema album (Empetraceae). A typical specimen and situation is seen in https://upload.wikimedia.org/wikipedia/commons/0/0c/Camari%C3%B1as_cerca_del_cementerio_de_los_ingleses.png

This species has a rather restricted distribution in the western parts of the Basin (https://www.inaturalist.org/taxa/326752-Corema-album). Indeed, this distribution does not technically fall within the Basin in the first place, because the coast involved is that of the Atlantic, not the Mediterranean Sea.

As far as I know, there is also no littoral effect on the incidence of fleshy fruits in climatically comparable parts of California and Chile.

Here is a hypothesis in explanation of the littoral effect.

What sets Australia and southern Africa apart ecologically, among the various continental regions with a mediterranean-type climate (https://www.researchgate.net/publication/228019813_Mediterranean_Ecosystems), is the nutrient-poverty of the soils, and the widespread sandiness at the coast.

Fleshy fruit-pulp tends to be rich in potassium, and not other nutrient elements (https://www.jstor.org/stable/2844617 and https://www.jstor.org/stable/2389867).

The sea is a source of affluence of potassium in particular, because of the tendency for aerosols to blow inland from sea-spray (https://www.mdpi.com/2073-4433/9/8/298/htm#:~:text=Marine%20aerosols%20are%20comprised%20of,condensable%20atmospheric%20gases%20%5B2%5D.).

It can therefore be argued that fleshy fruits represent an adaptation to a relative surfeit of potassium, in the sense of an ample ratio of this cationic element to other nutrients such as phosphorus and zinc.

This surfeit would hypothetically not arise in the Mediterranean Basin, because

The sea-swell is also greater in coastal California and Chile than those in the Mediterranean Basin. However, here too, a surfeit of potassium may not arise, because the coastal substrates are likely to be even nutrient-richer than those in Europe and North Africa.

Posted on November 21, 2022 08:58 PM by milewski milewski


Posted by milewski 4 months ago (Flag)

The other way to look at this is that the Cape and Australia burn, so bird dispersed fruit are rare and not a high priority dispersal mode. By contrast rocky outcrop forests, afromontane forest and strandveld have lots of bird-dispersed fruit. There are a few Fynbos excepions, but they prove the rule: Cassytha - a parasite killed by fire, that survives in rocky unburned patches and spreads by growing from these, that needs to recolonize other potential rocky refugia before the next fire eliminates it. And tortoise berries (where water is more important as an attraction than nutrients), rather than bird fruit.
Bird fruit dispersal is expensive, and best found where dispersal to bird perches that are fire-safe habitats (and often isolated and remote) is a premium. Strandveld is characterized by dune landscapes, which provide both summit perches and fire safe swales, as well as being a linear feature, often fragmented by rocky coasts.
So if it is expensive and there are no specialized habitats that can be targetted, then why should it be a viable strategy? Cheaper dispersal strategies make sense.
Wind and dunes go together, but in substantial areas the predominate southeaster wind direction is off-shore or long-shore, so is marine nutrient enrichment really so significant? Surely it is the wilder shores with larger wave action that will have more nutrients dispersed inland (rather than a gentle bath like the Med)?

Posted by tonyrebelo 4 months ago (Flag)


You have made good points. It is so valuable to have different minds, with fresh perspectives.

Here are a few thoughts (in no particular order), which may partly modify your contribution to the conceptual framework.

It is true that, in South Africa, fleshy fruits are associated with protection from wildfire. However, this association is looser in the Mediterranean Basin and California. In the former, there are various spp. of fleshy fruits on flammable shrubs and trees such as Phyllirea (https://en.wikipedia.org/wiki/Phillyrea_angustifolia), Arbutus, Juniperus, and Myrtus; and it is hard to think of any analogue to fire-free strandveld or forest on the mainland there, although there are analogues on e.g. the Canary Islands. In California, there are many genera with fleshy fruits in chaparral (e.g. Arctostaphylos, Heteromeles, Berberis, Malosma), and most or all are flammable, surviving fires more intense than those typical of fynbos.

The association concerned, in South Africa, is one of afromontane and afrotropical affinity. This biogeographical extension of equatorial floras into the temperate zone does not apply to any of the other continents concerned, except for the fire-free forest of the Atlantic archipelagos west of Iberia (the vegetation of which perhaps hints at what the mainland was like before desiccation and anthropogenic modification by fire).

One of the ways in which this shakes out is that Ericaceae include significant producers of fleshy fruits in the Mediterranean Basin, California, and Australia, whereas in southern Africa one does not see 'erica' and 'fleshy fruit' in the same sentence.

Putting this another way: an odd aspect of the southern tip of Africa is the dichotomy between pyrophilic and pyrofugic vegetation, which correlates with a dichotomy between temperate-zone and tropical floras. The fleshy fruits occur mainly in the pyrofugic/tropical flora.

Onshore vs offshore/longshore winds: this is indeed food for thought. I now see that the windiness is perhaps less important than the swell. The Mediterranean Sea presumably has relatively little swell. The southern African coast has major swell. Swell produces breaking waves, and it is the spume from these breakers that charges the littoral air with aerosols, even on calm days. Thank you for this.

Come to think of it: is there any good surfing, anywhere in the Mediterranean?

Your point re costs of fruiting is also thought-provoking. However, it does not seem self-evident that fleshy fruits are intrinsically particularly costly. I guess it depends on the resources considered, and a 'cradle to grave' accounting. I suspect that reproduction is always costly in some sense, and that what various plants do is to choose affordable currencies given their particular habitats, niches, and trade-offs.

Your further thoughts?

Posted by milewski 4 months ago (Flag)

Just to point out that the Cape may be poor in bird species, and especially bird frugivores, but it is noteworth that the Cape Flora endemics include:
Fynbos: 2 nectarivores, 2 insectivores, 2 seedeaters,
Renosterveld: 1 insectivore (Agulhas Lark)
Strandveld: 1 frugivore (Cape Bulbul) and a near-Cape endemic Cape Francolin.

It is also noteworthy that Cape Town gardens are filled with Strandveld birds, Bulbuls being ubiquitous (although summer rainfall and forest birds are moving into treed and summer-watered landscapes).

Posted by tonyrebelo 4 months ago (Flag)

Ectomycorrhizal associations occur in Kunzea (Bellgard 1991, https://agris.fao.org/agris-search/search.do?recordID=US201301753143). I predict that Kunzea pomifera lacks ectomycorrhizae.

The same paper is a reference for the lack of ectomycorrhizae or VAM in Persoonia, which instead has 'vesicular associations'.

Posted by milewski 4 months ago (Flag)

Kunzea pomifera occurs as a prostrate shrub, on Kangaroo Island and Yorke Peninsula, at the tip of Eyre Peninsula, in the Murray valley, and in southwestern Victoria. Its habitat is sand dunes and among rocks, "often close to the shore".

Birds eat the fruits.

The fruit is a reddish, succulent berry, diameter about 0.8 cm, turning bluish black at maturity.

Ref.: Wrigley J W and Fagg M (1993) Bottlebrushes, paperbarks and tea trees, and all other plants in the Leptospermum alliance. Angus & Robertson, Sydney.

Posted by milewski 4 months ago (Flag)

Various fleshy fruits in mediterranean-type climates are adapted to reptiles, rather than birds, for seed-dispersal.

The diaspores of these plants are provided within reach of reptiles that forage on the ground, by emerging just above ground level (e.g. Gethyllis), by falling off when ripe (e.g. Muraltia spinosa and probably certain ericas), or by being borne on infructescences that sag on maturity (e.g. Haemanthus and Zantedeschia).

@tonyrebelo Zantedeschia aethiopica is a candidate for seed-dispersal by agents other than the usual birds (e.g. Pycnonotidae, Collidae), based on the sagging of the stalk and the partial covering of the ripe fruits. Is there any literature on this?


Posted by milewski 4 months ago (Flag)

On 20 June 2000, I wrote to Teo Maranon (https://www.irnas.csic.es/en/teodoro-maranon-arana/) in Sevilla, Spain, about the flora of fleshy fruits in Spain and other regions of the Mediterranean Basin.

He replied:
"Dominant fleshy fruits woody genera in West Mediterranean (from my experience) are: Arbutus, Pistacia, Olea, Crataegus, Viburnum and Myrtus, and the climbers Smilax, Hedera and Lonicera. These are mostly pre-mediterranean (sensu Herrera 1992). On the contrary dry-fruited abundant genera, usually favoured by disturbance are Cistus, Halimium, Rosmarinus, Phlomis, Thymus, Lavandula and Helichrysum. (Genera names taken from the list in Arroyo and Maranon, 1990, J. Biogeog. 17: 163-176)"


Posted by milewski 4 months ago (Flag)

Zantedeschia aethiopica
Never seen birds take them, but then fruit are obviously sought after because ripe fruit dont last long. The plants are heavily predated by porcupines - in areas where the porcupines have been poached (using snares) the populations of lilies just explodes (& see ***): it would be a delicious irony if porcupines are important dispersers. [by the by, the plants are so full of oxalates a nibble is like chewing needles - does not bother porkies! - I have never tried the fruit but presumably they are oxalate free. But the fruit are orange and solid [like chickpeas] when ripe: not "inviting to try" imho).

https://www.inaturalist.org/observations?nelat=-33.943&nelng=18.467&place_id=any&subview=map&swlat=-33.951&swlng=18.4608&taxon_id=56006&verifiable=any - zoom in: note the occurrence inside and outside the exclusion plots. Some of the pictures illustrate this.

Posted by tonyrebelo 4 months ago (Flag)

Many of the plants with fleshy fruits in the Mediterranean Basin are shared with temperate Europe farther north, e.g. in Germany, where they grow in abandoned orchards (Milton et al., 1997). Examples are Prunus mahaleb (https://www.inaturalist.org/taxa/126808-Prunus-mahaleb), Prunus spinosa (https://www.inaturalist.org/taxa/55811-Prunus-spinosa), Cornus sanguinea (https://www.inaturalist.org/taxa/155124-Cornus-sanguinea), Hedera helix (https://www.inaturalist.org/taxa/55882-Hedera-helix), Sambucus nigra (https://www.inaturalist.org/taxa/765394-Sambucus-nigra), Crataegus monogyna (https://www.inaturalist.org/taxa/51147-Crataegus-monogyna), Ligustrum vulgare (https://www.inaturalist.org/taxa/119805-Ligustrum-vulgare), Rosa canina (https://www.inaturalist.org/taxa/55884-Rosa-canina), Rosa rubiginosa (https://www.inaturalist.org/taxa/78886-Rosa-rubiginosa), Asparagus officinalis (https://www.inaturalist.org/taxa/60265-Asparagus-officinalis), and Rubus (https://www.inaturalist.org/observations?place_id=any&taxon_id=47544&view=species), at least at generic level.

Posted by milewski 4 months ago (Flag)

On the Agulhas Plain, vine thicket dominated by Sideroxylon inerme has the greatest incidence (dominance in terms of canopy or foliage cover) of plants with fleshy fruits. It is well-supplied with potassium by marine aerosol (enhanced by fog, settling on the tall shrubs by virtue of the shapes of the crowns), while having a substrate (consisting of calcite, aragonite, and quartz) that is not felsic but is inherently poor in phosphorus and micronutrients. It is protected from wildfire (by landform, climate, clumpiness of the woody plants, succulence of some of the smaller plants, and herbivory in the gaps). It thus has accumulation of humus, recycling potassium with minimal loss to volatilisation or leaching. It contains several parasitic plants, but negligible nitrogen-fixing plants. Vegetation growth is in equilibrium with herbivory, which recycles potassium faster than phosphorus (animals build bines but pass potassium in urine). The main plants all regenerate vegetatively, maintaining biomass with minimal need for incorporating potassium into growth of foliage. Sideroxylon inerme is extremely long-lived, and hardly ever observed as a seedling or sapling.

Posted by milewski 4 months ago (Flag)

One aspect of the littoral effect, in the case of southwestern Western Australia, between Albany and Israelite Bay:

Gavicalis virescens (https://www.inaturalist.org/taxa/370319-Gavicalis-virescens) eats fleshy fruits (e.g. Chenopodium spp.) as well as nectar and insects, and is one of the more frugivorous of the small meliphagids.

This bird is widespread in Australia. However, is in this region it is more or less restricted to the littoral.

Posted by milewski 4 months ago (Flag)

@tonyrebelo @jeremygilmore
I happen to have an individual of Zantedeschia aethiopica in full fruit in my garden at the moment.

The flower-stalks have sagged after the flowers wither.

Initially, the infructescence, resting on the ground, consists of small, green, developing fruits completely hidden in a green flask (a rolled, leaf-like spathe).

When some of the fruits, still at ground level, are full-size and ripening, the green flask = spathe opens partly, its inner surface not green.

The fruits are greenish-yellow at full size and just before full ripeness, and may be succulent at this stage, but are obviously not ready to be eaten.

The fully ripe fruits are shrivelled and yellow-orange, and naturally detach, so that they can easily be shaken out of the flask. They are moist but not succulent, and disintegrate easily if manipulated.

There is certainly no pre-ripe display.

It is possible that there is significant lipid content in the fully ripe fruit, which has the consistency of a soft aril, although it contains the seeds.

I just ate two of the ripe fruits. The fully ripe (shrivelled, yellow-orange) fruit, with its tender, pasty pulp, tastes mildly sweet, with no bitterness, sourness, or astringency. There are about five seeds, which found myself spitting out rather than swallowing (even though I swallowed the thin fruit-skin).

The seeds are pale and about the size of a sunflower kernel, although not oblong.

Initially there was no 'sting' of oxalate. However, after 20 seconds a faint spiciness started on the tip of my tongue, persisting for at least 15 minutes. This effect (consistent with oxalate) is strong enough that it would be unpleasant if I ate the fruits in any significant quantity.

In summary, this yellow, partly hidden fruit seems unpalatable to humans owing to oxalate, although it is not nearly as 'stinging' as a vitaceous fruit I once ate on the Athi-Kapiti plains in Kenya, which shocked my mouth with its 'electric' effect, and ensured that I have never had any subsequent inclination to sample the fruits of anything resembling a wild grape.

Posted by milewski 4 months ago (Flag)

Referring to our discussion earlier in this stream:

I have checked my notes, and apparently Phillyrea (Oleaceae) of the Mediterranean Basin, which bears fleshy fruits, is extremely flammable.

From the viewpoint of any plant ecologist from the Cape Floristic Region, this sounds odd, not so?

Do you know of any species, anywhere in southern Africa, that combines fleshy fruits with extremely flammable foliage?




By the way, Phillyrea latifolia seems to be heteroblastic. Do you know of any comparable plant in the Cape Floristic Region that has 'juvenile foliage'?

Posted by milewski 4 months ago (Flag)

How does one define and detect "extremely flammeable" ? Cannot say I know how to detect this, other than that the Gnaphalidae (at least Stoebe and Seripheum) seem to burn cool and smokey, whereas the Rutaceae (often in the same habitats, and esp. in rocky outcrops) burns extremely hot and clean.
But in the general fynbos matrix it is not easy to see during a fire which plants burn hot vs cold, or even clean vs smokey. The one exception we know of is Erica verticillata which smothers fire (and is very smokey) and raging fires going through, just become cool smoke until the other side, when they erupt again. Proposed as a mechanism to protect seeds in a summer flowering species with very very weak (and not well protected) serotiny.

Posted by tonyrebelo 4 months ago (Flag)


Many thanks for the interesting details, of which I was unaware.

In my boyhood, I spent many a starry night sleeping, during hikes in the Cederberg, on a makeshift mattress of slangbos (Seriphium plumosum, https://www.inaturalist.org/taxa/431651-Seriphium-plumosum, https://www.inaturalist.org/taxa/431651-Seriphium-plumosum).

But I have never investigated the relationship to fire of this daisy, with its rather peculiar combination of 'leaflessness' and feltiness.

Posted by milewski 4 months ago (Flag)

And dont forget its acaricidal properties which make Slangbos ideal not only for makeshift beds, but also as stuffing for mattresses.
My question is - is the acaricidal properties related to the fire-cooling? Do the sticky waxy smelly features accomplish both ends?
And do birds use it to line their nests? - are they aware of the acaridal properties? - does this help to disperse seeds?

Posted by tonyrebelo 4 months ago (Flag)

All of these are good questions.

My initial guess would be that the main formula for reducing flammability in Seriphium plumosum is a combination of a 'woolly' foliage-form (https://pza.sanbi.org/seriphium-plumosum) that limits the flow of air (including oxygen) and a lack of any flammable hydrocarbons comparable with the 'aromatic oils' of many spp. of Myrtaceae in flammable vegetation in Australia, and Rutaceae in fynbos in South Africa. Many types of daisy are aromatic, but S. plumosum is not. The makeshift mattress is soft but not noticeably fragrant.

Posted by milewski 4 months ago (Flag)

I disagree. One of the best reasons for a Slangbos bed is the camphory odour. It is not very strong, but very noticeable and characteristic.
more here: https://pza.sanbi.org/seriphium-plumosum

Posted by tonyrebelo 4 months ago (Flag)

Lots of your references here relate to densification of Slangbos with bad management - hence one of its names: bankrotbos (bankruptcy bush) - and the need to treat the symptoms without apparently even bothering to understand or redress the underlying causes!

Posted by tonyrebelo 4 months ago (Flag)


I stand corrected on the question of the fragrance of Seriphium plumosum. It is more than 50 years since I slept on it.

Posted by milewski 4 months ago (Flag)


In https://pza.sanbi.org/seriphium-plumosum, it is stated that this species makes good kindling for fires, and can increase the domestic risk if planted in the garden in fire-prone areas. Is there any contradiction with the idea (above) that this plant retards rather than promotes combustion?

Posted by milewski 4 months ago (Flag)

Living vs dead plants. Dried plants are superb fire lighters.
But they are fire retarders, not excluders: they dont stop fires, just the fires burn slower and cooler than on average. In an open setting (fireplace) with lots of oxygen they burn very well, compared to dense stands among rocks - but they dont stop the fires, just attenuate it. In open veld individually among other plants their effect is hardly noticeable as fires move through.

Slangbos has surface seed banks. My theory is that it provides a cool fire to protect its seeds, whereas the Buchus with ant buried seeds incinerate Slangbos seed banks and stimulate their own deep seeds. But Slangbos still needs fire for recruitment in Fynbos, although removal of other plants mechanically or by overgrazing works just as well.
In rocky outcrops, there are two extreme stable states - as exemplified by Slangbos dominated outcrops vs Coleonema outcrops: it is a war.

Posted by tonyrebelo 4 months ago (Flag)


In the context of intensity of wildfire, it seems significant that one of the most consistent differences between kwongan and fynbos is in the incidence of Asteraceae. Kwongan, which burns with extreme intensity, is virtually devoid of daisies, and lacks asteraceous counterparts for even ericoids such as Metalasia.

Of all the daisies in South Africa, which would you regard as the most intensely flammable in the living condition?

Posted by milewski 4 months ago (Flag)

Well Bietou goes into the category of those that dont want to burn. It survives the prescribed burns (managers dont want to risk a really hot fire) and then takes over. A "thicket" invasion of Fynbos during cool fires.

I cannot answer your question: "which are the most flammable alive?"

But it is worth pointing out that
** Renosterveld which burns well, is dominated by daises. (that often leaf-stink to some degree).
** Asteraceous Fynbos - which also burns well - is usually on the hotter, drier (and lower, and north-facing) slopes. I am not aware of any inherent difference in fire (other than flame length which is most impressive in Proteoid Fynbos) between Ericaceous, Proteoid, Restioid or Scrub Fynbos (although Grassy Fynbos burns more like grass fires than Fynbos fires: i.e. not nearly as terrifying).

Posted by tonyrebelo 4 months ago (Flag)

It sounds as if Dicerothamnus rhinocerotis may be the prime example of a flammable daisy, in the Cape Floristic Region. If so, this would be subject to the caveat that it was probably subordinate in the original, grassy situation with the full community of herbivores. It was a 'natural fireweed' in a community otherwise rather suppressive of fire (with palatable grasses and non-flammable bush-clumps).

If we provisionally accept that the abundance and diversity of daisies in the southwestern Cape (including karoo vegetation) correlates with relative downplaying of wildfire in the natural ecology, then this makes for a remarkable contrast with southwestern Australia.

Here, there are only two common daisies, and both are restricted to situations with minimal wildfire.

Olearia axillaris (https://en.wikipedia.org/wiki/Olearia_axillaris) occurs at the littoral, where fire tends not to penetrate because the adjacent beach is bare, and because the associated acacias are not particularly flammable.

Cratystylis conocephala (https://en.wikipedia.org/wiki/Cratystylis_conocephala) occurs inland, mainly in eucalypt woodland that is too sparse to carry fire.

It is noteworthy that neither species is sclerophyllous or even semi-sclerophyllous in the way of Metalasia muricata.

Posted by milewski 4 months ago (Flag)

But Metalasia muricata is a thicket-strandveld species. It is replaced in Fynbos by Metalasia densa, which is to all intents and purposes identical!

Posted by tonyrebelo 4 months ago (Flag)


There is a daisy similar to Seriphium plumosum in southern Australia, viz. Leucophyta brownii (https://www.inaturalist.org/taxa/323827-Leucophyta-brownii and https://en.wikipedia.org/wiki/Leucophyta and https://www.bgpa.wa.gov.au/about-us/information/our-plants/plants-in-focus/2559-leucophyta-brownie and https://plantselector.botanicgardens.sa.gov.au/Plants/Details/873).

See https://www.inaturalist.org/observations/143005214.

However, in Western Australia this species is so restricted, ecologically, that one encounters it in suburban cultivation more than in the wild.

There is no fire-free ecosystem in southwestern Australia in the sense that there is in part of the strandveld in the southwestern Cape. However, L. brownii seems specialised for the least fire-prone situation possible, being confined strictly to the littoral, and being protected by topography to boot (https://www.inaturalist.org/observations/141239538).

Instead of being an element of kwongan, it has virtually 'been driven into the sea' (https://www.inaturalist.org/observations/141317656), figuratively-speaking, in its pyrofugic dependence.

Posted by milewski 4 months ago (Flag)

Osteospermum moniliferum, the fruit of which is fleshy, occurs both on the littoral and far inland, e.g. in the Karoo (https://www.inaturalist.org/observations/49695964). The fruit-pulp is relatively thick and moist in the former situation, vs thin and meagre in the latter situation (Richard S Knight, pers. comm.). This intraspecific pattern conforms to the littoral effect.

Posted by milewski 3 months ago (Flag)

A species showing the littoral effect in Australia is Carpobrotus virescens.

This species has fleshy fruits (https://cambridgecoastcare.com.au/wp-content/uploads/2015/05/P140A-2014-x-Greg-Keighery-Carpobrotus-SNEC-REPORT2.pdf).

The following show the distribution:

Posted by milewski 3 months ago (Flag)

Add a Comment

Sign In or Sign Up to add comments