Early Land Animals
Bill Shear (Hampden-Sydney College, Hampden-Sydney, Virginia, USA)
Department of Entomology & Department of Biology, Ohio State University
23 November 1999
Some remarkable early terrestrial animals were found by paleobotanists (Birmingham Campus of SUNY) recently in the Devonian of New York. The evidence for the earliest terrestrial ecosystems is scattered and fragmentary - most of it is in the Afro-Arabian area of Gondwana and southern Laurentia (Early Paleozoic coordinates).
The earliest evidence is fossil spores. Sporopollenin is very resistant, including being resistant to bacterial attack. The Middle Ordovician of Arabia has spore tetrads, which are markers for embryophytic plants (correlates with terrestriality) - were dispersed as tetrads (groups of four) - modern liverworts still do this. Trilete marks indicate that a single spore was originally from a tetrad. Very few other remains are available. There are no macrofossils in the Ordovician, though, that contribute to our understanding of the earliest terrestrial ecosystems.
We see a shift from dispersed tetrads to dispersed individual spores (moss grade) by the Silurian. Tracheids are found in Silurian rocks, as well as plant cuticle, including cuticle from enigmatic plants called pneumatophytes, plants consisting of trunks or bundles of tiny tubes. By the Silurian, most evidence for terrestrial ecosystems is from either side of the Acado-Caledonide Mountains. Animal evidence begins to show up. Middle Silurian sediments (including the Rose Hill Formation of western Virginia) contains arthropod cuticle (Shear showed a Rose Hill setate podomere from some arthropod).
In the Upper Silurian, there is richer evidence of terrestrial plants and animals, including Cooksonia (basically living astroturf) - Cooksonia comes in vascular and nonvascular forms which are otherwise indistinguishable at the generic level. Looking at the type locality for the Ludlow in Wales (in town of Ludlow), we have a succession of Ludlovian rocks containing a fish bed which is traditionally chiseled way back into the outcrop, but has been recently bulldozed in order to more easily sample nowadays. This fish bed contains tea-leaf-plant-fragment hash and a nice eurypterid claw chelicera. It also has yielded a coalified spider-like arachnid, a trigonotarbid (an arachnid without spinnerets and no silk production and the abdomen is still segmented). This trigonotarbid is the earliest known land animal (Late Silurian). In the Upper Silurian of Scotland, there is a millipede-like fossil called Archadesmus which has verifiable diplosegments; so, millipedes were also early invaders of the land.
The Old Red Sandstone of southern Scotland was partly deposited near hydrothermal activity near the town of Rhynie. The paleo-hot springs would occasionally overflow into surrounding vegetated areas, resulting in the local terrestrial flora and fauna being preserved in fine-grained chert, the Rhynie Chert. Rhynie Chert has long been used by locals as building stone, and has been called noodle rock. The noodles in the rock are actually plant stems with wonderfully preserved details. The Rhynie Chert is ~405 my (Devonian), and is a potential gold (Au) field. In the Devonian, contrary to most reconstructions, the land probably had vast tracts of monospecific stands of plants, rather than the mixed diversity scenes we often see. The Devonian land plants probably used a turfing strategy. The Rhynie Chert also has animal fossils, including trigonotarbids (with compound eyes and book lungs preserved, as well as exceptional preservation of carapace and legs, etc.). There is an exceptionally good specimen in 1 translucent chip that is reminiscent of amber - the trigonotarbid has 3-D preservation that shows up beautifully in transmitted light.
Near Gilboa, New York state, a pump storage facility was recently being constructed, and the excavation uncovered a lens of dark gray shale packed with plant remains. The construction crew was good enough to stop and allow collecting by SUNY Birmingham paleobotanists for 1 day - they came with a pick-up truck and collected several tons of rock.
By the Mid-Devonian, vegetation was more layered than before - had tall trees to low shrubby things. The Gilboa material included trigonotarbid arthropods - only about a dozen are complete specimens. Trigonotarbids are fundamentally spider-like, but they lack many spider apomorphies.
Found a spinneret in the Gilboa macerations with spigots from which silk was issued - this was obviously from a real spider. Cellular impressions were found on this spiders cuticle. It has been reconstructed based on additional macerated material from this site to be similar to the most primitive living spiders. This Devonian spider is called Attercopus [Note: does anyone recognize the inspiration for the genus name of this fossil spider? Bilbo Baggins in The Hobbit used the name "Attercop" as an insult when confronting the mongo spiders of Mirkwood Forest!]
Attercopus has silk spigots that appear to have been scattered over the ventral surface of the abdomen, rather than concentrated at the posterior tip of the abdomen. Showed a chelicera of Attercopus - < 1/10 mm across - wonderful preservation; a spinneret for silk occurs on the chelicerae.
The 1st Paleozoic pseudoscorpions also occurs in this material, and is called Dracochela - a representative of a primitive clade of pseudoscorpions. Pseudoscorpions lack a stinger, and their relation to true scorpions is dubious.
Also have orobatid mite material in the Gilboa collections, some of which is very well preserved with setal hairs. Gilboa has also yielded centipede remains. There is also potential evidence of Devonian insects in this material - cuticle with coffin-shaped sockets reminiscent of bristletails - flightless insects living today; or, perhaps, they represent silverfish.
A later Devonian site near Gilboa (South Mountain locality) has arthropleurid leg material (forms which by the Late Carboniferous were 6' long and 2' across) and Attercopus spider cuticle & silk spigots and also a hemipteran insect-like head (there is no smoking gun for insect material - but there is mounting evidence for Devonian insects). This locality also has Archaeospermatopsis, probably the first tree - had spore & seed bearing forms, indistinguishable from each other at the generic level.
The Red Hill locality in central Pennsylvania has yielded a scorpion leg with individual setae in place, etc.; also part of a scorpion pincer; also large patches of unidentified arthropod cuticle.
Vertebrates were real latecomers to the land - they were more aquatic than terrestrial - they may still have had internal gills. They were capable of land excursions, but to what extent is unknown. Ichthyostega has 7 digits on each front leg and 8 digits on each hind leg (not the traditional 5 each). The earliest demonstrated tetrapod that was a herbivore is from near the Pennsylvanian-Permian boundary.
The earliest evidence for herbivory is probably Middle Carboniferous insects. There is a problem with looking at insects mouth parts in judging herbivory based on morphology - a praying mantis mandible & a locust mandible will not inherently indicate to you that the mantis is predatory and the locust is herbivorous. In the Middle and Upper Devonian, there is evidence of suctorial damage to plants (by aphid-like things?). Many mites are pollenivores. Spiderlings are also - they will make an orb web that catches pollen, and the web with its pollen is eaten.