How essential is the use of modern training sets, also called analogs, to interpret past microfossil assemblages when reconstructing past environments and climates?

Palaeoecology is a branch of study of the earth’s geological history concerned mainly with the interrelationship of organisms which coexist in a given environment. Geologists have proven that past events in a particular environment have an influence on the future events in the same environment. This means that palaeocological studies of the distribution of organisms in a given environment today and information on the factors influencing such a distribution can help develop structures of the said environments in earlier years (Seddon, 2011).

Remains of plants and animals that earlier occupied the eco-system can be found in sediments from water bodies such as lakes and oceans. An analytical study of these remains can be used to show the way ecosystems have changed and evolved over time and their adaptation to environmental changes such as climate change and those caused by disruptive human activities. In recent times, there has been an increase in human activities such as deforestation and settlement, which have an effect on the environment (Simpson et al., 2005). As a result, new land use laws were introduced in Europe following the realization that human activities were causing the destruction of several water bodies (European Union, 2000). Palaeoecological studies therefore are important since they help predict the effects of such human activities on the environment in the future. This information can then be used to develop policies on pollution and environmental conservation. Some examples of microfossils used in studies involving the reconstruction of environments include diatoms, pollen grains and foraminifera. Environmental histories are obtained by comparing the modern microfossil assemblages with the ancient ones (Scott et al., 2001).

The determination of the evolution and the development of past environments due to environmental changes through the study of fossils is an indication of the magnitude and duration for the change (Hann, 2007). Quantitative palaeoecological studies are based on the fact that changes in an ecosystem occur over time hence there is a relationship between historical events, the present and what happens in the future (Rymer, 1978). Analogue techniques have been used in these studies to determine effects of the previous environmental conditions. This is done through exposure of the present-day species to known historical environmental conditions commonly known as space-for-time substitution, or modern analogue technique.

This paper is aimed at showing that analogs are important in palaeoecological studies intended for reconstructing past environments and climates. The organisms used in this study are fossil diatoms. Diatoms are a species of algae, which are among the most commonly occurring sea weeds. They are organisms often consisting of one cell, but they may occur as a group. They are always sources of food for other organisms in the food chain. Their cells are protected by a silica outer covering. Remains of diatoms have been used in palaeoecological studies to show changes in the ecosystem over time. They are mainly used in water studies since they are sea weeds and hence can show changes in marine ecosystems. Their remains have been found as silica deposits in sediments from water bodies and dry lands owing to the silica in their cell wall. However, they have been found to be more applicable to studies involving marine systems owing to the biology of their deposits being well-developed in water bodies. Diatom fossils can be used to determine ages of organisms up to 100,000 years since they are excellent tools for studying environmental and ecological changes over time (Culver, 2003).

The aspect of similarity is very important when using analogue techniques. Factors are used to measure the likeness between remains of organisms and every single analog. One of such factors is a chord distance used mainly for compositional data since there is a good relationship between the signal itself and other interferences in the analysis (Simpson, 2009). Analogue matching is a measure of similarity used to determine the areas in a current system of water bodies that show greatest likeness to areas before their evolution and change due to environmental factors. Areas that show the highest degree of similarity to the desired component in the study organisms found in the deposits in various water bodies are referred to as modern analogues. Thus, the determination of the state of the environment in the selected study site before the influence of environmental and human activities depends entirely on the likeness of the analogue samples to the fossil samples (Simpson, 2009).

 The level of uncertainty is obtained from the variation between the reconstructed value and the analogue value, and this gives the idea about the response of the inhabitants to weather changes and varying conditions in water bodies. This method assumes that the unpredictability in both climatic and oceanographic conditions is similar, which may not be the case. The uncertainty value is also influenced by the content of like assemblages that constitute any given sample, but this may be corrected by sampling modern assemblages again in a way similar to the response surface approach. This method involves placing the fossils in a representative environment using the modern analogues. Several methods have been proposed regarding the number of analogues to be used in any study. The method is similar to the modern analogue technique (MAT) in that the most similar assemblages are used for the study, and they guide the reconstruction (Rull, 2010). The similarity is still determined using the chord distance just like in MAT (Guiot & Vernal, 2007). Waelbroeck et al. (1998), suggest choosing the analogues while the chord distance is still small thus ensuring that the analogue and the fossil samples are very similar. Gavin et al. (2003), on the other hand, used a receiver operating characteristic curve to determine the highest similarity between the analogue and fossil samples. The following figures show plots made after using the MAT method (Simpson, 2009):

The use of multiple methods is used when more adequate reconstructions are desired, and when transfer functions are inadequate in determining the assemblages. The MAT and revised analogue methods (RAM) have shown excellent performance when applied to the multiple method approach (Jackson & Williams, 2004). Modern analogue techniques therefore make it possible to infer environmental history through comparison of analogue and fossil assemblages.