Approaches to pollen taxonomic harmonisation in Quaternary palynology

Birks, H. John B.; Bhatta, Kuber P.; Felde, Vivian A.; Flantua, Suzette G.A.; Mottl, Ondřej; Haberle, Simon G.; Herbert, Annika; Hooghiemstra, Henry; Birks, Hilary H.; Grytnes, John-Arvid; Seddon, Alistair W.R.

doi:10.1016/j.revpalbo.2023.104989

Cited by 8 publications

(6 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…López-Sáez et al, 2014;Flantua et al, 2016), examining human cultures and food production based 205 on anthropogenic indicators (e.g. Flantua and Hooghiemstra, 2023) or the assessment of rates of vegetation change during the Holocene (e.g. Mottl et al, 2021).…”

Section: Discussionmentioning

confidence: 99%

“…Mottl et al, 2021). With such a large database now publicly accessible and open access workflows to process and standardise large compilations (Flantua et al, 2023;Vidaña and Goring, 2023), countless opportunities for further study are available, including global and hemispheric syntheses.…”

Section: Discussionmentioning

confidence: 99%

“…The latter served as a training data set for palaeoclimate reconstructions for comparison with palaeoclimate models (Herbert and Harrison, 2016). This version of the database has been used in multiple wide-ranging studies to date, including global overviews of rates of change 70 processes (Mottl et al, 2021) and pollen taxonomic harmonisation processes (Birks et al, 2023). Here we explore this new constituent database in depth and suggest directions for future research.…”

Section: Introduction 40mentioning

confidence: 99%

See 2 more Smart Citations

The Indo-Pacific Pollen Database – a Neotoma constituent database

Herbert,

Haberle,

Flantua

et al. 2024

Preprint

View full text Add to dashboard Cite

Abstract. The Indo-Pacific Pollen Database (IPPD) is the brainchild of the late Professor Geoffrey Hope, who gathered pollen records from across the region to ensure their preservation for future generations of palaeoecologists. This noble aim is now being fulfilled by integrating the IPPD into the online Neotoma Palaeoecology Database, making this compilation available for public use. Here we explore the database in depth and suggest directions for future research. The IPPD comprises 226 fossil pollen records, most postdating 20 ka, but some extending as far back as 50 ka or further. Over 80 % of the records are Australian, with a fairly even distribution between the different Australian geographical regions, the notable exception being Western Australia, which is only represented by 3 records. The records are also well distributed in modern climate space, the largest gap being in drier regions due to preservation issues. However, many of the records contain few samples or have fewer than 5 chronology control points, such as radiocarbon, luminescence or Pb-210 for the younger sequences. Average sedimentation rate for the whole database, counted as years per cm, is 64.8 yr/cm, with 61 % of the records having a rate of less than 50 yr/cm. The highest sedimentation rate by geographical region occurs on Australia’s east coast, while the lowest rates are from the Western Pacific. Overall, Australia has a higher sedimentation rate than the rest of the Indo-Pacific region. The IPPD offers many exciting research opportunities, such as examination of human impact on regional vegetation, contrasting first human arrival and colonisation, and assessment of rates of vegetation change during the Holocene. Merging the IPPD into Neotoma also facilitates inclusion of data from the Indo-Pacific region into global syntheses.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introduction 40mentioning

confidence: 99%

See 1 more Smart Citation

The Indo-Pacific Pollen Database – a Neotoma constituent database

Herbert,

Haberle,

Flantua

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…We matched pollen and spore names to accepted plant names from Plants of the World Online 86 using the R package taxize (version 0.9.100) 87 . As pollen identification is often limited taxonomically 88 , we use two methods to improve the thoroughness of our approach. Both approaches, standardization-1 and standardization-2, removed the categories of indeterminate pollen and spores and taxa that indicated uncertainty at levels higher than family.…”

Section: Methodsmentioning

confidence: 99%

Floristic homogenization of South Pacific islands commenced with human arrival

Strandberg,

Steinbauer,

Walentowitz

et al. 2024

Nat Ecol Evol

View full text Add to dashboard Cite

The increasing similarity of plant species composition among distinct areas is leading to the homogenization of ecosystems globally. Human actions such as ecosystem modification, the introduction of non-native plant species and the extinction or extirpation of endemic and native plant species are considered the main drivers of this trend. However, little is known about when floristic homogenization began or about pre-human patterns of floristic similarity. Here we investigate vegetation trends during the past 5,000 years across the tropical, sub-tropical and warm temperate South Pacific using fossil pollen records from 15 sites on 13 islands within the biogeographical realm of Oceania. The site comparisons show that floristic homogenization has increased over the past 5,000 years. Pairwise Bray–Curtis similarity results also show that when two islands were settled by people in a given time interval, their floristic similarity is greater than when one or neither of the islands were settled. Importantly, higher elevation sites, which are less likely to have experienced human impacts, tended to show less floristic homogenization. While biotic homogenization is often referred to as a contemporary issue, we have identified a much earlier trend, likely driven by human colonization of the islands and subsequent impacts.

show abstract

“…This is because pollen grains are often morphologically similar, especially if belonging to closely related species. Significant effort has been spent towards improving taxonomic resolution, identification consistency between analysts, and identification accuracy for pollen with very similar morphology [ 4 , 5 ]. That being said, expecting consistent results between different analysts should not necessarily be expected [ 6 , 7 ].…”

Section: Introductionmentioning

confidence: 99%

Pollen identification through convolutional neural networks: First application on a full fossil pollen sequence

Durand,

Paillard,

Ménard

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

The automation of pollen identification has seen vast improvements in the past years, with Convolutional Neural Networks coming out as the preferred tool to train models. Still, only a small portion of works published on the matter address the identification of fossil pollen. Fossil pollen is commonly extracted from organic sediment cores and are used by paleoecologists to reconstruct past environments, flora, vegetation, and their evolution through time. The automation of fossil pollen identification would allow paleoecologists to save both time and money while reducing bias and uncertainty. However, Convolutional Neural Networks require a large amount of data for training and databases of fossilized pollen are rare and often incomplete. Since machine learning models are usually trained using labelled fresh pollen associated with many different species, there exists a gap between the training data and target data. We propose a method for a large-scale fossil pollen identification workflow. Our proposed method employs an accelerated fossil pollen extraction protocol and Convolutional Neural Networks trained on the labelled fresh pollen of the species most commonly found in Northeastern American organic sediments. We first test our model on fresh pollen and then on a full fossil pollen sequence totalling 196,526 images. Our model achieved an average per class accuracy of 91.2% when tested against fresh pollen. However, we find that our model does not perform as well when tested on fossil data. While our model is overconfident in its predictions, the general abundance patterns remain consistent with the traditional palynologist IDs. Although not yet capable of accurately classifying a whole fossil pollen sequence, our model serves as a proof of concept towards creating a full large-scale identification workflow.

show abstract

Approaches to pollen taxonomic harmonisation in Quaternary palynology

Cited by 8 publications

References 36 publications

The Indo-Pacific Pollen Database – a Neotoma constituent database

The Indo-Pacific Pollen Database – a Neotoma constituent database

Floristic homogenization of South Pacific islands commenced with human arrival

Pollen identification through convolutional neural networks: First application on a full fossil pollen sequence

Contact Info

Product

Resources

About