Sieving out the big picture: Collecting microvertebrate fossils

Dr Steven C Sweetman (UK)

Ask any palaeontologist, professional or otherwise, to name the first fossil vertebrate or vertebrate group that comes to mind and the chances are that the majority will come up with something like the charismatic dinosaurs, Dimetrodon (Fig. 1), the saber-toothed ‘tiger’ or some other large and spectacular creature from the past.

The chances of anyone coming up with, for example, albanerpetontids (Figs. 2 and 3), an extinct (Middle Jurassic to Pliocene) group of newt-sized, superficially salamander-like amphibians, are probably next to nil. Indeed, who except specialists have ever heard of the Albanerpetontidae?

However, an understanding of the small animals that lurked in the shadow of the large and generally better known beasts with which they coexisted can often shed valuable light on ancient ecosystems and palaeobiology, and provides insights that cannot be obtained from study of big beasts in isolation.

Despite this, the discovery of beautifully preserved dinosaur and large fossil mammal remains, particularly in the badlands (Fig. 4) and tar pits of North America, has quite naturally generated much more public interest than the discovery of microfossils.

As a result and until comparatively recently, the majority of funding for scientific research was channelled towards macro rather than micro remains, and they do of course make much better museum exhibits (Fig. 5).

The purpose of this article is to provide some tips on how to collect microvertebrate fossils (Fig. 6). A second article will discuss a search that took place for Early Cretaceous tetrapods using the techniques discussed in this article during which, over a period of just four years, as many new species were found as have been found during the preceding 180 years of collection by traditional methods.

First, a little background information. The use of wet-sieving techniques to obtain the remains of small fossil vertebrate dates back to at least the mid nineteenth century in Europe. However, it was not until the 1960s and 70s that methods were developed that facilitated isolation of large quantities of microvertebrate remains and funding became available for their collection. This came about as a result of an explosion of interest in Mesozoic mammals, most of which were poorly known, very small and often represented only by isolated teeth.

At that time, now well-known palaeomammalogists, such as Malcolm McKenna (American Museum of Natural History), William Clemens (University of Kansas and later University of California, Berkeley) and Richard Fox (University of Alberta) were working on the Late Cretaceous mammals of North America. Their primary objectives were to gain a better understanding of mammalian evolution and taxonomy. However, as data accumulated, it also became clear that mammals and other microvertebrate remains could also be used in biostratigraphy and for palaeobiogeographical, palaeoecological and palaeoenvironmental reconstructions.

At first, the methods employed to collect small vertebrate remains were basic at best. Surface prospecting was used to locate sites yielding small bones and teeth, and samples from them were passed dry through a sieve using window screen as mesh. This is the wire mesh commonly used in many parts of the world to keep mosquitoes and other insects out of houses, and usually has openings greater than or equal to 1.1mm. Alternatively, if a water source was available, samples were washed in boxes using the same screen material.

However, at the most basic level, samples were simply dumped in burlap bags and washed in these. Although large quantities of matrix could be processed in the field using these techniques, the large mesh size used and the open weave of burlap resulted in the loss of many of the smallest mammal fossils and, indeed, those of all other groups present. As a result, only a crude picture of faunal diversity was obtained.

Among the early workers, including Richard Estes (University of California, Berkeley), who worked on the non-mammalian components of the vertebrate assemblages recovered from bulk screening projects, there came a realisation that sieve mesh size needed to be reduced. With this and with improved screening methods, some involving laboratory-based equipment, many more taxa were recovered, allowing quantitative as well as qualitative analyses of the data collected.

Equipment

Work on microvertebrate assemblages continues apace and the number of new small taxa discovered each year continues to increase. So, how are these remains recovered and studied, and is the recovery and study of microvertebrate remains something that can be undertaken by private collectors and not just professionals? The answer to the last part of the question is emphatically, “yes”. As for the first part, some basic equipment is required, and this inevitably involves a not-inconsiderable initial financial outlay. However, once you have it, you tend to have it for life.

Undoubtedly, the most expensive two items of equipment you will need is as good a quality, low power, binocular microscope as you can afford and, by choice, a powerful but adjustable fiberoptic light source (Fig. 7). Once the microvertebrate collecting bug bites, you will spend hours peering down your microscope, both during the process of recovering fossils and during subsequent study of them.

The better the optics and ergonomic design of the instrument, the more can be observed and the less strain there will be on both body and eyes. A magnification range of about x5 to x75 will cover most of your needs, but cost savings can be made if you opt for a narrower magnification range of, say, x10 to x50. This should be sufficient for most, but will not permit examination of the finest detail on very small specimens. A good hand lens of about x10 magnification is also useful when working in the field.

To collect bulk samples for processing, you will need ‘large’ sample bags for small samples of up to about 10kg in weight, heavy duty buckets for larger samples up to about 25kg and large containers for storing even larger samples. A good quality, framed backpack can also make the job of getting samples from the point of collection to either a vehicle or your processing station a lot less physically demanding than hand carrying them. In addition, you will need tools to extract the samples in the field.

Of course, the tools needed will depend upon the size of samples you intend to collect and the nature of the rocks concerned. Items to consider are a shovel and spade, a metal detectorist’s long-handled trowel, pointed and square ended rock picks, chisels of various sorts, a strong wide-bladed knife (check local restrictions on carrying such an item), a large flat ended screwdriver … the list could go on, but these are a start.

The equipment and techniques required to process samples will depend upon their size, the nature of the matrix concerned and how much time you are willing to spend standing over the sieves. Those described here are most suitable for home or laboratory use, not for large-scale processing in the field. They are not suitable for the recovery of very small specimens (of less than about 350 microns [µm]), although they could be adapted for this purpose.

For small samples up to about 10kg in weight – the best way to start – a nest of hand sieves is sufficient. However, for larger samples, some form of ‘automated’ machine is much better and will free you to spend time retrieving fossils from residues obtained from previously processed samples or to study already isolated fossils. In either case, the mesh size of the sieves used for bulk processing is critical. Too large a mesh size and valuable fossils will be lost; too fine a mesh size and processing and fossil recovery times will become unacceptably long. For most purposes, a sieve with a mesh size of between 350µm and 400µm will be fine enough to retain most potentially useful small vertebrate remains.

Trial and error has shown that the vast majority are found in residues with a particle size of between 0.5mm and 1.18mm, but it is necessary to divide the residues into several fractions before attempting to separate fossils from them. This both saves time and ensures that small fossils are not obscured by larger particles during the sorting process. The largest mesh size used will depend on the nature of the matrix being processed. If this contains large clasts of any kind, it is best to screen these out at an early stage, both to reduce the volume of residue and to avoid them, so far as possible, damaging small vertebrate remains during the sieving process.

To provide flexibility, the following is a suggestion for hand-sieve mesh sizes: 10mm; 2.5mm; 1.18mm; 500µm; and 355µm. The brass variety, which is 200mm in diameter and 50mm deep, and manufactured by Endecotts Ltd, has proved to be well made and durable. If cost is not an issue, then the 300mm diameter sieves made by the same company provide the de luxe option.

As indicated above, large samples are best processed using a bulk processing machine.

Several designs have been described in the literature and a variation of one of these is shown here. This uses recirculated water and, depending upon the type of matrix being processed, permits processing of samples of up to about 50kg at a time (Fig. 8).

It incorporates a purpose-built sieve with a mesh size of 330µm. Since the illustration was prepared, the sprinkler system has been modified somewhat – the mist sprinkler has been replaced by an oscillatory sprinkler operating at right angles to the other. This was found to give substantially faster run times, with no appreciable additional damage to fossils.

Another very effective ‘wash box’ design is that used by palaeontologists working at the Royal Tyrrell Museum of Palaeontology in Alberta, Canada (Fig. 9). In this system, the sieve or sieves are suspended in a water bath, at the base of which are circular garden sprinklers. However, in this case, the sprinklers are used to generate air bubbles from a compressed air system. The bubbles interact with the base of the samples in the sieves, preventing clogging and further speeds up run times while being gentle on the retained fossils.

Finally, other items of equipment you will need (Fig. 7) include:

• Containers for residues.
• A sorting tray for use with the microscope.
• Fine paint brushes to aid sorting of residues and for handling fossils.
• Small glass or clear plastic bottles with caps for the storage of fossils.
• Labels to identify the contents and its origin.
• A fine pin holder and pins for various uses.
• A small cabinet of some sort to house specimens.

Some specimens are best observed and manipulated under the microscope when temporarily attached to a pin or similar object. Perhaps the most convenient device is a stub of the sort used in scanning electron microscopy. These can be stored individually or in boxes containing multiple stubs. Specimens are attached to the stubs using special adhesive discs and can be removed using readily available solvents. The aluminium stubs can be scored with identification marks using a pin and pin holder. Stubs, boxes and related supplies can be obtained from Agar Scientific (go to www.agarscientific.com to view their online catalogue).

Methods

As with all fossil collecting, it is important that the landowner’s permission is obtained before samples for processing are collected and that recognised codes of good collecting practice are followed, even if the fossils are not readily observable in the field. Furthermore, microvertebrate remains often occur together with scientifically important fossils of other kinds, and care should be taken to ensure that such specimens are not damaged or destroyed while collecting samples. Well-recognised safety procedures should also be observed.

The equipment described above is primarily suitable for processing silts and clays, and it is important to ensure that the horizons sampled will break down in water. Some even apparently well-consolidated strata will break down in this way, but it is necessary to break clay mineral bonds before samples are processed. To do this, they should be thoroughly dried before processing. If time is no problem, samples can be stored under cover outside (Fig. 8) or placed in a greenhouse or something similar. For smaller samples, you may wish to risk the chef’s wrath and use an oven.

An electric fan oven is best with the temperature set as low as possible (gas ovens are generally not suitable). Beware of rushing the process because this can shatter fragile fossils. Even when dry, some samples will not break down, in which case it may be possible to use something as simple as washing-up liquid or chemicals (such as hydrogen peroxide) to achieve this. Guidance on methods such as these can be found online.

So far as the choice of potentially productive horizons is concerned, you can be almost sure that if surface prospecting reveals small vertebrate remains (Fig. 10), substantial quantities of teeth and bones will be recovered from residues obtained after sieving.

However, very small vertebrate remains (often the most interesting) are often hard to detect in the field, in which case, it may be a matter of taking pot luck and sampling a particular horizon blind if you think it may produce fossils of the kind you are seeking. The pot luck approach (well almost) was adopted in the study to be outlined in a forthcoming article and stunning results were achieved. However, doing some preliminary research to identify potentially productive horizons and taking trial samples from these may save a lot of wasted time and effort.

Once you have collected samples and dried them, they can either be processed by hand or using a bulk processing machine (Fig. 11).

If using the latter, do not overload the machine and monitor it from time to time. If using recirculated water, it will be necessary to rinse the residue in fresh water before it is removed from the sieve for drying. Once dry, pass the residue in small batches through a nest of sieves comprising all or some of those outlined above, depending on the variety of particle sizes present. Then label and store each fraction separately in preparation for the isolation of fossils.

If processing the samples by hand, put small quantities in the 10mm sieve with it placed above the 2.5mm sieve and below this, place the 355µm sieve. Experience has shown that trying to use the full nest during the wet sieving process is too cumbersome and that the use of sieve vibrating machines to speed the sieving process usually results in unacceptable damage to specimens. Soak the sample by placing it in a container with water sufficient just to cover it.

A suitably modified plastic dustbin with a lid diameter larger than the base diameter is ideal for this (Fig. 12).

When the sample starts to break down, gently agitate it and/or use a hose to complete the sieving process. Dry as above and then pass the fraction retained by the 355µm sieve through a nest comprising the 1.18mm, 500µm and 355µm sieves. Store as above.

Material coarser than 2.5mm can be sorted very quickly by eye. The remainder must be sorted, grain by grain, under the microscope. In some cases, this process can be speeded by secondary processing of the residues using acids, usually dilute acetic or formic, to remove unwanted calcareous fossils. (One very well-known shark expert was once heard to say deliberately in the hearing of an equally well-known mollusc expert that his favourite sound in all the world was that of mollusc shells fizzing in acid.) However, be careful to ensure that you are not destroying previously unrecorded material of scientific significance that, apart from its palaeontological interest, may also help with biostratigraphy and palaeoenvironmental reconstructions.

Also, if substantial quantities of low density material (such as plant remains) are present, it may be possible to remove the bulk of this by simply panning it off. It can always be kept for future examination.

High density liquids can also be used to speed fossil recovery, but the non-hazardous substances available are very expensive and, therefore, beyond the reach of most private collectors.

An outline of the sort of material you might be able to recover is provided in Fig 13, but will, of course, depend upon the horizons you sample and the palaeoenvironments they represent. In the simplest terms, freshwater and terrestrial assemblages are generally obtained from lacustrine and fluvial deposits, but these are also found in lagoonal and some near-shore marine deposits, where they occur together with marine forms. However, there are exceptions, which will be explored in the next article.

Other articles in this series:
Sieving out the big picture
In the shadow of the Isle of Wight dinosaurs