CHAPTER VII
Dispersal in and around The West Indies
Patterns of dispersal geographically restricted only to the Antilles are well
known to the student of distribution, but it nees be the biogeographer's
viewpoint that in no case may even patterns of the kind be handled without
regard to the rest of the world. Patterns of the kind belong to the whole of
the Americas in the first place, which means to the earth by implication, and as
such - the biogeographer can only maintain - they must be approached.
The principle - right fundamental for a sequential appreciation of dispersal
in all its phases - that everything interlocks with everything else consequently
applies to the biogeography of the sector in and around the Caribbeans, and
holds good even in cases when differential formmaking was established in the
Antilles and/or on the continent nearby in ages so far remote as to preclude the
possibility of retracing today the details of the "tracks" responsible for the
original exchanges between the mainland and the Antillean islands; and it might
accordingly seem possible to handle the dispersal in the latter without regard
of dispersal elsewhere. This case is verified for example, when the picumnine
genus Nesoctites turns up in Hispaniola away from the rest of the
subfamily on the mainland, and the "track" that put it there ca no more be
pinpointed step by step today than can the channel responsible for
Picumnus' turning up simultaneously in Brazil and Indochina. Even more
striking in this same connection is the dispersal of two related avial living
familites, Todidae all in Greater Antilles, and Momotidae all in Latin America
1.
Trusting the most conservative estimates of time demanded for the
differential segregation of groups in this taxonomic rank, we cannot accept a
date much later than Late Cretaceous/Palaeocene (age of fossilization![1]),
which is to conclude that a critical study of the dispersal of archetypal
todid/momotid stock necessarily brings into play palaeogeographic factors that
lie by now wholly out of our grasp 1. All we know is that we face in these
"modern" families chips out of the same ancestral block which - strewn in origin
jointly over sectors of the world which we today severally understand as
Caribeans and Tropical America - eventually "ripened" (cf. Ameiva and Cnemidophorus) into the
"modern" families. These being the elements in our hands as regards the
"todid/momotid problem" it seems rather reasonable that: i) What is "caribbean"
or "continental" of these two "families" is at bottom a single "proble" for the
competent student of distribution; and not to become the subject ever of
aprioristic speculation on the ground that some of these birds are today
"insular", but others not. For all that we may reasonably accept, the birds
that are today "insular" had sires that were "continental", and the oher way
around; ii) It is wholly out of the question - this plainly being the case -
that constructive inquiry may be made to rest on academic discussions of
"lanbdbridges", "winds", "causal transportation", and the like; or on
assumptions beforehand that the Todidae became Momotidae at the result of
"colonizing flights" out of the Antilles into continental America; or that the
Momotidae were turned into Todidae by "jumps" out of continental America into
the Antilles. Discussions and assumptions of the kind are necessarily
very wide of the mark of nature, which is here unreservedly to affirm
that their results are not genuine part of sceince. Science is concerned on the
contrary with ancestors of joint todid/ momotid genetic, and distributional,
potential as the starting point of constructive inquiry; iii) Naturally, Todidae
and Momotidae are to be studied together if we intend to understand either, and
both throughout their history in time, space and form. This can only mean,
finally, that there is no biogeographic problem with the Antilles that is
not also a problem with the rest of the world.
I should think these conclusions only embody truisms, and I am reasonably
sure that these truisms are safe beyond a range of "zoogeographic" artillery.
This is my own opinion, but having had far better opportunities of hearing the
"zoogeographic" contrary the reader may perhaps not be inclined to accept these
truisms as plausible. Anticipating this possibility, I venture to state that to
draw the line beteen, e.g. Mayr's understanding [3} of Antillean dispersal, and
what I understand of it myself is easy. To discriminate quickly and finally we
no more than to take due account of concrete examples out of the records of
life. If the "zoogeographer" is right, the facts in the record shall bear him
out; if not, no one will have permanent reasonh of bemoaning his eventual
dismissal from the halls of knowledge.
Vast is indeed the arrary of instances useful to test the issue here
pointedly beforfe us, the difficulty being now - as ever - in selecting for
immediate use only a strict minimum of evidence of the sort nevertheless that
drives clear to the jugular of the matter. Freely using my own judgemnet -
which may not be the best, I am afraid - still it seems to me that the
plant-family Magnoliaceae offers excellent material as the test-case we here
need. This group turns up among the earlies angiospermous fossils which is to
say that it dates (age of fossilization!) from Early Cretaceous, i.e., some
150,000,000 years back. That this family was in existence by age of being at a
much earlier age goes without saying. Additionally to a factor of great age,
the Magnoliaceae are of major interest for us right here because we have of them
reliable data from the West Indies [4], nor are plants of the kind trashy
ubiquists with active "means of dispersal", come today and gone tomorrow. They
belong to the landscape, as we might say, and the land is perforce old there,
where they are at home. The record of Magnoliaceae in the sector of our present
interest stands in these terms:
I) Magnolia (West Indies only)
1) M. cubensis - Cuba (higlands of Oriente: Pico Turquino, Sierra de Nipe;
also Trinidad Mts. in Las Villas); 2) M. domingensis - Hispaniola
(hills of Haiti); 3) M. pallescens - Hispaniola (Dominican Repubvlic:
Cordillera Central at ca. 2100m. (7000 ft.)); 4) M. hamori - Hispaniola
(Southwestern Dominican Republic: highlands of Barahona Province at ca 1350 m.
(4500 ft.)); 5) M. emarginata - Hispaniola (Haiti: northern and central
highlands at 900 - 1300 m. (ca. 3000 - 4300 ft. )); 6) M. ekmanii -
Hispaniola (Haiti: southwestern peninsula, on Massif de la Hotte at 1200 m. (ca.
4000 ft.)); 7) M. splendens - Portorico (Sierra de Luquillo, El Yunque,
etc. at ca. 1000 m. (ca. 3300 ft.)); 8) M. portoricensis - Portorico (
Mt. Bajaza, Cerro Gordo, etc., at about 1000 m. (ca. 3300 ft.)).
II) Talauma (West Indies only)
1) T. minor(T. truncata) - Cuba (highlands of Oriente: Pico
Turquino, Sierra de Nipe, Sierra de Moa); 2) T. dodecapetala - Lessar
Antilles (highlands of Guadeloupe, Dominica, St. Vincent).
Thisa is the record authenticated by recent reliable work 1 and I venture
categorically to affirm in its light: 1) The localities stressed by these ten
species are uniformly well known centers of relictural endemism and
form-making. No naturalist at all familiar with life in the West Indies,
whether botanist or zoologist, can be found to maintain that these are "casual"
records; ii) The form-making in Hispaniola, Portorico and Cuba cannot be
explained having reference to secondary factors of chance-dissemination,
ecology, climate, etc. It represents the ultimate meeting of already ancient
cores of survival following an history that can certainly not be later than late
Cretaceous/Very Early Tertiary at the latest; iii) To match centers of this
notable kind in the Greater Antilles, T. dode capetala delivers into
our hands also a record out of the hoary cores of life extant in the Lesser
Antilles, particularly the islands of Guadeloupe, Dominica and St. Vincent.
This record is worth underscoring because it alone promptly disposes of the
oft-heard "zoogeographic" claim that if some of the Greater Antilles might,
perhaps, have had former connections with the mainland, still the Lesser
Antilles are themselves assuredly "oceanic". Talauma does plainly
show that this is false, because one of the best known centers of relictual
endemism in Cuba (the highlands of Oriente) does actually interplay in its ranks
with Guadeloupe, Dominica and St. Vincent. 2
Of course, vouched for by one shred of Talauma evidence, the
antiquity as a biogeographic center of Guadeloupe, Dominica, St. Vincent, and,
subsidiarily, St. Lucia and Martinica, is none the less verified by the sizeable
list of endemisms belonging to these islands in general, and in particular. I
note here, but to implement the data from Magnolia and
Talauma, that the lesser Antillean Core, generally represented by the
islands just mentioned, harbours endemic bird-species and genera, as follows: 1)
Guadeloupe - Dendroica plumbea, Melanerpes lherminieri; 2)
Dominica - Amazona arausiaca, A. imperialis, Cyanophaia bicolor (this
last common to Martinica)'; 3) St. Vincent -Amazona guildingi, Catharopeza
bishopi. Jointly shared by these islands and St. Lucia/Martinica are three
local avian genera, Cinclocerthia, Ramphocinclus, and
Cichlherminia (this one extending to Montserrat); and St. Lucia has for
its own two monotypes, Leucopeza semperi and Melanospiza
richardsoni. Even discounting records unsuppported by actual specimens to
the utmost, still it seems sure (2:p176) that these islands harboured to within
rather recent hisorical times between four and siz endemic parrots.
Some of the forms mentioned (e.g. Cinclocerthia/Ramphocinclus,
Melanospiza) have notable significance, and unquestinable position as
relics. Adding to this old avian stock and endemisms peculiar to these islands
in point of plant-life, insects, onychophores, herpetofauna, etc., the total
score is an imposing one, vouching for the sector of the map later become the
Lesser Antilles having received its normal quota of "modern" life neither much
earlier, nor much later, than the rest of the New World at large. Cuba that up
to the Mid-Miocene was not the single island it is today, only a galaxy of
lesser discrete islands; Panama and Costarica that throughout the Tertiary
repeadtely had insular status to match the Antilles as of today; Java, an island
far remote from the Caribbeans but in many respects of its making (marked recent
volcanism excepted) suggestive of Cuba; etc; do furnish a background of
reasonable biogeographic comparison with the Lesser Antillean Core and the
Antilles generally.
The 3-d nature
mentioned in discussion of Comparative Biogeography can be tested fully (no
matter the temporal discord) with C(AB) areas above:
What the Lesser Antilles are today, Panama/Costrica were in the past, and may
well be tomorrow and so Cuba and Java again; telling evidence how utterly
futile is the "zoogeographic" approach that begins buzzing around for honey with
the question: Is this today an island? Or is this not today an island? And
where could the "raft" or the "colonizing flight" themselves land to the
exclusion of the "landbridge"? I do not scruple affirming that this
approach, and all it entails, is strictly nonsense. It rests on streams of
turbid compilation dissociated from a sufficient knowledge of the records of
life: and having thus stated my own side of the question I suggest that
dissenters - for the good of science - do prove making precise reference to the
records of life that their own side of the question is better supported than
mine.
Of course, the ancient bonds between Lesser and Greater Antilles
displayed by Talauma, and handsomely confirmed by the endemisms
peculiar to the Lesser Antillean Core, are not the exclusive property of the one
magnoliaceous plant. Connections of the very same scope also turn up in the
rank of the following turdid species: Myadestes genibarbis - Highlands
of Jamacia, solitarius; highlands of Hispaniola (local on the lowlands
of Samana Peninsula; cf. Todus angustirostris), montanus;
highlands of Dominica, dominicanus; higlands of Martinica,
genibarbis s.s.; highlands of St. Lucia, sanctae-luciae;
highlands of St. Vincent, sibilans.
This record is precise and full, nor does it mean anything at all (cf Chapter
V) that it is pinpointed with "races" rather than with "species". Faced by it,
the innocent "zoogeographer" risks rushing along to claim that, oh!.... yes ....
by all means! .... this is because "hurricanes" promoted "colonizing flights"
and "jumps", but to keep him out of danger ere he steps headlong into it, I here
observe: 1) Nothing can be claimed for the dispersal of Myadestes that
fails to answer also for that of Talauma, and the whole of the
endemisms belonging to the Antilles. All or nothing is the slogan, so
please refrain trying for only something. That will never do; ii)
Talauma rings the North Pacific from New Guinea and the Philippines to
Mexico, the Lesser Antilles and beyond into South America. The stations it has
in Cuba and the Lesser Antillean Core are accordingly but function of a massive
intercontinental dispersal, and the whole of this dispersal stands or falls
together. What is true of Talauma on the Arfak Mountains of New Guinea
must also be true by dispersal (i.e., by essentials of time and space)
of Talauma on the mountains of Dominica; iii) Myadestes,
currently accepted as an ally of the Hawaiian Phaeornis, ranges from
Alaska to the Bolivian yungas, and the lone form reported from
Venezuela, M. ralloides venezuelensis, has for its range: Colombia
(Meta: La Macarena, Cudinamarca, Norte de Santander) and Venezuela (Andes of
Tachira, Merida, Trujillo, Lara; cordilleras of the Distrito Federal).
Obviously, not being prepared to underwrite as good and true a "zoogeographic"
claim to the effect that La Macarena and Norte de Santander (see Chapter IV, V)
were colonized by "hurricanes" et. similia - considering said claim to
be misplaced in essence and by origin (i.e., strictly stemming from disregard of
elementary facts of nature) - I am even less prepared to accept a claim of the
sort as vaild in regard of the Antilles; iv) Indications are generally not
wanting for still other connections between the Greater and Lesser Antilles more
or less in Myadestes genibarbis' style. For example: Bonds' accurate
records [2} vouch for the finch Loxigilla portoricensis having ranged
within 50 years ago both on Portorico (portoricensis s.s.) and beyond
Anegada Channel on St. Chrisopher (or, St. Kitts; grandis extinct only
after 1880); for the woodpecker Melanerpes lherminieri (Guadeloupe)
being close to M. portoricensis (Portorico, Vieques); for the
hummingbird Cyanophaia bicolor (Dominica, Martinica) suggesting direct
alliance to Chlorostilbon (Portorico, Hispaniola, the Bahamas, Cuba
and Isle of Pines); for the hawk Buteo platypterus freely ranging with
different races from Cuba and the Isle of Pines to Hispaniola, Portorico,
Antigua, Dominica, Martinica, St. Lucia, St. Vincent, Grenada, Tobago; for
another hummingbird Orthorhynchus being distributed from St. Vincent to
the islands off Portorico (Vieques, Culebra,, Virign); etc.
"Moralizing" and "preaching" (or what sounds like it, anyhow) is currently
looked upon with disfavour in the best of scientific circles on the ground that
the lanugage of science consists of "facts". Well knowing how facts are on occasioin distorted to mean quite
the contrary of their intrinsic message (see Chapter III), and how wholly
brittle may prove the "statements of fact" even in the pages of "authoritative
literarture" (see end of Chapter III in particular), I do strongly believe that
the language of science does consist of carefully, sequential discussions of
pertinent evidence, or - in other words - of reason and common sense jointly
mustered to reach ever deeper -via the facts, or course - into the
vitals of nature. This discussion can happily be kept free from "moralizing"
and "preaching" if there is no need for it, but not otherwise. Knowledge is not an abstraction, but a psychological and
historical phenomenon rising and ebbinbg with the tides of times, and it is
strictly legitimate that the argument be conducgted with this also well
in sight, accepting "moralizing" and "preaching" as connatural to the ultimate
diffusion of better understanding. Here accordingly "moralizing" and
"preaching" a step farther still, I venture to invite
the reader to consider how easy it proves to shatter ageold "zoogeographic"
tales using but the club of a few assorted records, and how, starting with one
record, it is easy to find others quite as telling. The conclusion to present
"moralizing" and "preaching" needs of course be, that if old "zoogeographic"
tales still persist this can only be because the average naturalist is rather
too supine before "authority" for his own, and the common good.
This being the seventh chapter of this book, it is my distinct feeling that
whatever I am to deal with - inside and
outside the Antilles - can hardly return anything at all that the reader has
not seen, if in a different aspect, many a time
before. The zoogeographic score does not alter, and some new form of life comes
in, but the biogeographic shop always does its business at the old sign, and the
moment Myadestes or Talauma have gotten manicured and shaved,
the next customer can only take the chair for the same kind of service. Here,
for example, is with us a customer that, seemingly of the oddest, is in rfeality
but the ordinary run of the millo, as follows: Aratinga pertinax
(Psittacidae) - Panama (Chiriqui to the Canal Zone, islands in the Gulf of
Panama), ocularis; Colombia (Cordoba: Sinu Valley, eastward to
Northern Magdalena, southward to Santander), Venezuela (Zulia to Territorio
Amazonas), aeruginosa; Colombia (the eastern llanos from
Arauca and Meta to the Orinoco), lehmanni; Venezuela throughout to the
exception of Eastern Bolivar and Delta Amacuro, venezuelae; Venezuela
(Bolivar: Mt. Roraima district), British Guiana, chrysophrys; Venezuela
(Monagas:Barrancas to Delta Amacuro) eastward to French Guiana,
surinama; Northern Brazil (upper Rio Negro), chrysogenys;
Dutch Leeward Islands (Curacao; allegedly introduced in the Virgin Islands; St.
Thomas), pertinax s.s.Venezuela (island of Tortuga, west of Margarita),
tortugensis; Venezuela (Margarita and adjacent islets of Los Frailes;
Anzoategui, Northern Bolivar, Sucre), margaritensis; Dutch Leeward
Islands (Bonaire), xanthogenia.
Although records are extant in the literature [5] to the effect that the
Virgin Island population of pertinax s.s. is not quite true to the
birds of Curacao, still it has been constantly maintained in the pages of
standard referneces that this parrot occurs in the Virgin Islands only because
of "introduction". I do not know how this has
gotten universal credence when those who affirm it have no proof of the
"introduction" alleged 1,
and surely do know, somehow, of records such as these, for example:
1) Elaenia martinica (Tyrannidae) -...
2) Margarops fuscatus (Mimidae) - ...
3) Caprimulgus otiosus (Caprimulgidae) - ...
If one know casts a glance at the map (Fig. 80) displaying the distribution
of these birds jointly,
he will of course feel that there is nothing miraculous or unusual in A.
pertinax associating as one in range Curacao and the Virgin Islands. The
range of this parrot may be spectacular on account of its thus exhibiting a
marked disconnection, but it is clear out of the map that this disconnection
has place within a distributional norm of major interest to a circum-caribbean
circuit. Accordingly, this disconnection oversea is neither more
nor less wondrous than the breaks we have studied in previous pages, this time
of interest to continental South America overland. Concluding, the
subject of what A. p. pertinax did, or did not, between Curacao and St.
Thomas cannot be a major consideration with the informed student; we will
accordingly drop it here, well understanding that it is by no means sure that
this parrot was "introduced" to the island of St. Thomas. The contrary is most
certainly true.
Hardly inclined to waste time on aspects of dispersal that are, yes,
spectacular but nothing more at bottom, the inquistive student will be impressed
on the contrary (see Fig. 80) by the manner in which dispersal disports itself
along the whole of the insular chains fronting latin America (Cozumel - Grenada)
into the Antilles proper, not only, but does interplay on occassion (cf.
Aratinga pertinax, Caprimulgus otiosus) with the mainland. The
interplay that thus takes palce seems commonplace in regard of A.
pertinax, but is right challenging as concerns C. otiosus. Here
is "something" that does not agree at all with the geography now extant.
Dispersal that clings to an insular chain offshore, zigzagging - on occasion
- in and out the continent, eventually freely to stream - on occassion - tyo
reach genuine archipelagoes at sea is, of course, no longer new to us. This is
the dispersal we have identified already (see p. 492, 493) as parian or
horstian. It highlights the suture, as it were, of a land "in back"
and islands "infront", true in its course to a fundamental aspect of
geosynclinal activity. The islands "in front" are one entity, as such, and
so is one entity, as such, the land "in back"; but the two entites do stand
within a single one geological bond. Because of this bond dispersal
highlights, also necessarily, as essentially distinct the insular chain "in
front" at one hand; and the land "in backf" at the other hand. However, the two
also act as one in endless degrees of possible combination "in front" as well as
"in back" the like the attached diagram (Fig 81) generally indicates.
In thus stressing - and so forcefully - an essential interplay in
horstian (or, parian) distribution of flesh and rocks it is
not my intention to imply that the whole of the islands between Cozumel and
Grenada, and Grenada into the Antilles proper is necessarily to be viewed as a
single geologic feature (i.e. one geosyncline). Geologists would not
approve of too broad a generalization, for, quite as we are particular in
matters of detail and proper terminology, so are they. However, and so by far
this time, dispersal of the type we have seen - and we are yet
abundantly to see - could not be were it not for an essential interplay, to
repeat, between one system of islands "in front" and land "in back"; which
interplay is characteristic in general of geosynclinal activity, and thus a
powerful molder of dispersal on the rebound. It matters not whether,
technically speaking, we deal with one geosyncline or a whole system of geosync
lines. The point is that flesh and rocks here act in unison in answer to a
generality common to geology and dispersal, as one. Whether this
genearlity has yielded in the Colombian range high mountains or in the Caribbean
sector only insular galaxies - as of today, for in the geological
future the situatioin may easily be reversed - is a secondary consideration,
spectaular as its effects may be for us today1. (1 Many of the readers,
and the North American ones in particular, have surely heard...)
Summing up, the patterns of dispersal we have seen displayed above do not
belong to the geography of today otherwise than in
point of record. They do belong to the palaeogeographies of the
Caribbeans,
and becasue of this they play havoc with (the geographic) set of land and
sea that happens to rule today. Had not a now vanished foreland of
Parialand once held sway; had not the Antillean chains once been
underwater; etc; we would not have today under the eye patterns of distribution we have sween and will yet see.
This is the fundamental consideration in Caribbean and Antillean
dispersal; and I do suggest that those who may intend to deny that had better
critically and carefully digest the records of life in preparation to their
going to print. Details to discuss and to settle remain plentiful; the main
outlines seem rather clear right now.
Parian dispersal has not gone undectected by naturalists other than
myself, nor can it be missed by anyone at all familiar with American
distribution. In one of his contributions on the herpetofauna of British
Honduras [7] Schmidt speaks for example of a certain type of fauna "Peculiar to
the keys" offshore, which is a qualification transparent enough. Bond does not
fail to comment in his study of the avifauana of the Bay Islands off Caribbean
Honduras [8] as follows: "The Bay Islands are evidently the tops of a submerged
mountain range, which, through isolation from the mainland, have retained a
number of Central American species, which have entirely or most disappeared from
Central America proper and which appear as relicts, not only on the Bay Islands,
but also on islands off the coast of Yucatan". Thus
very well informed in general, Bond does not fail to follow the statement just
quoted with examples exhibiting distribution far-flung along the "tops of a
submerged mountain range", including, e.g. Myiarchus tyrannulus insularum
of the Bay Island but "Remarkably similar" to M.t.nugator of the
Southern Lesser Antilles (Grenada, Grenadines, St. Vincvent); a close repetition
of the "similarites" between races of Troglodytes musculus on Cozumel
off Yucatan and the Southern Lesser Antilles again (St. Vincent/St. Lucia);
etc.
Hardly credible, but nevertheless quite true, Bond's work as a professed
"zoogeographer" [e.g.,9] actively spreads the creed of "zoogeography" by winds,
hurricanes, causal agencies and the like, supinely toeing the Matthewian line1.
It is a curious question how a naturalist who had of life in the Caribbeans
matchless knowledge; who knew that dispersal there made much better sense if -
wholly against current geography - the
islands of San Anres and Providencia were associated on score together with the
true Antilles; who had clear perception of
distribution: "Along the tops of a submerged mountain range" generally to
parallel the shore of the main, and understood these "tops" as relic-stations;
who had, in sum, faithful, exact understanding of masses of details - some of relevant potential scope - and returned in every
cases excellent analyses of affinites, "similarities", etc., from the very
standpoint of rigorous biogeography;