Number crunching in palynology

Simulated pollen counts
Simulated pollen counts

Our understating of vegetation in the past, and how it has changed through time, comes mainly from the examination of macrofossils (e.g. wood and leaves) and microfossils (e.g. pollen and spores) found in the sedimentary record. The potential for microscopic fossils to provide an insight into past vegetation change on a landscape scale was pioneered by von Post (Von Post, 1916, reprinted 1967) and has been subsequently used to understand changes in regional floras (Godwin, 1956), and address conservation issues (Willis et al., 2007). Analysis of fossil pollen and spores (palynology) is now widely used on late Quaternary timescales to answer ecological questions linking vegetation and wider environmental/climatic change; these include:

  • Has there been a change in major vegetation type (biome)? For example a change between woodlands and grassland vegetation.
  • How have the ecosystem dynamics altered? For example the presence or absence of fire.
  • How has the diversity within the ecosystem changed? For example increase or decrease in sample richness.

Palynological analysis relies on obtaining a sub-sample of the pollen contained within the sediment at a specific depth (time) which allows the vegetation at that time to be reconstructed. This sub-sample is known as a pollen count. To build up a picture of vegetation change through time it is necessary to generate a sequence of pollen counts. The size of the sub-sample (pollen count) required from any particular depth (time period) is dependent on the nature of the vegetation association being investigated and the ecological question being addressed . For example, the amount of pollen analysed to determine if the vegetation was predominantly wooded or grassland is different to that required to provide information on the biological diversity within the vegetation assemblage.

Discussed below are some of the conventions related to choosing an appropriate pollen count size within palynology, with particular reference to the challenges of dealing with diverse tropical floras.

For a novice palynologist, the question: “how many pollen grains do I need to count?” is one of the first asked; however, the answer is not as simple as might be expected. Palynologists have attempted to broach this question since the origins of pollen as a proxy; however, a definitive answer has not yet been found (Hill, 1996 and Rull, 1987). A value of 300 fossil terrestrial pollen grains is widely accepted as an ‘industry standard’ count size (Moore et al., 1991); however, some caution is required when applying 300 grain pollen counts to different vegetation types and for addressing different ecological questions. Although the 300 count size is used by the majority of palynologists, not every palynologist abides by it. Some palynologists prefer to find a statistically significant number of pollen to count (Maher, 1981 and Maher, 1971) and/or to take into account local and regional vegetation characteristics (Rull, 1987).

I believe that there are three key issues which need to be considered with selecting an appropriate count size as each sample. Firstly, each sample is different in terms of taxonomic composition; i.e. variation in richness (number of taxa present), evenness (abundance of each taxa within a sample) and diversity (combination of richness and evenness). Secondly, further complications are inherent because, different species disperse pollen in different ways and produce different quantities (Sugita, 1994). Thirdly, and possibly most importantly, palynologists also need to think critically about what question they are trying to answer with the pollen count data. Therefore, ideally when deciding on an appropriate pollen count size vegetation assemblage characteristics, environmental conditions and the scientific question should all be considered .

Varying richness, evenness and subsequent diversity between samples should cause palynologists to treat all samples individually, leading to an appropriate count sizes (Rull, 1987 and Van der Knaap, 2009). Plant diversity varies within environments, with tropical ecosystems generally having a higher biodiversity than temperate ecosystems. The range of floristic  diversity should be represented in the pollen from the different ecosystems, with those with a higher plant diversity having a higher pollen diversity and vice versa. The changes between floristic diversity and the consequent pollen diversity should determine to the palynologists the apprpriate count size for each sample. With varying levels of floristic diversity within any study site through time the question of even having a appropriate count size for a single sequence is brought into question. It may be preferable to move away from the ‘industry standard’ model and move towards a system that allows the palynologist to choose different count numbers for different diversities in individual samples?

If the diversity is high within a sample (as is often the case when working in the tropics) then it may be appropriate to increase the pollen count size. However, is it always worth counting more, or is the extra pollen counted just an extra pressure on the palynologists time? The answer to this, of course, depends on what the scientific question being asked.

Dependent on what the palynologist is using the data for, larger counts can mean the difference between usable results and data that cannot answer the research aims. It is consequently critical for a palynologist to establish the question they are trying to answer early on and design a sampling strategy capable of answering it. Thinking about a suitable count size is one of the hardest problems a palynologist will have to approach. 

My preliminary research into statistically simulating pollen counts suggests that there is no simple answer, indeed to get a accurate picture of change through a sedimentary sequence it may be most appropriate to assess count size on a sample-by-sample basis, i.e. taking into consideration richness, evenness and diversity as the count progresses. Balancing the time invested in each sample by the palynologist against the ecological information extracted is fundemental to obtaining detailed robust records of vegetation change. For a novice palynologist choosing a suitable count size could mean that their time is spent efficiently, creating data that is usable and that was counted to a statistically appropriate number. Over the course of my PhD I hope to develop a statistical protocol for optimising the efficiency of pollen conting for different vegetation types. Watch this space for further information over the next couple of years…!


Godwin, H (1956) The history of the British flora:  A factual basis for phytogeography.CambridgeUniversityPress,Cambridge.

Hill, T.R. (1996) Statistical determination of sample size and contemporary pollen counts, NatalDrakensberg, South Africa. Grana. 35 (2), 119 – 124.

Maher, L.J. (1981) Statistics for microfossil concentration measurements employing samples spiked with marker grains. Review of Palaeobotany and Palynology. 32, 153 – 191.

Maher, L.J. (1972) Nomograms for computing 0.95 confidence limits of pollen data. Review of palaeobotany and palynology. 13, 85 – 93.

Moore, P.D., Webb, J.A. and Collinson,M.E.(1991) Pollen analysis. Blackwell Scientific, London.

Rull, V. (1987) A note on pollen counting in palaeoecology. Pollen et spores. XIXX (4), 471 – 480.

Sugita, S. (1994) Pollen representation of vegetation in Quaternary sediments: theory and method in patchy vegetation. Journal of Ecology. 82, 881 – 897.

Van der Knaap, W.O. (2009) Estimating pollen diversity from pollen accumulation rates: a method to assess taxonomic richness in the landscape. The Holocene. 19 (1), 159 – 163.

Von Post, L. (1916, reprinted 1967) Foresttree pollen in south Swedish peat bog deposits. Pollen et spores. 9, 375 – 401 (English translation by M.B. Davis and K. Faegri from original 1916 lecture; Introduction by K. Faegri and J. Iversen)

Willis, K.J., Araujo, M.B., Bennett, K.D., Figueroa – Rangel, B., Froyd, C.A.and Myers, N. (2007) How can a knowledge of the past help to conserve the future? Biodiversity conservation and the relevance of long- term ecological studies.  Philosophical transactions of the royal society B: Biological Sciences. 362, 175 – 186.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: