Delayed postglacial colonization of Betula in Iceland and the circum North Atlantic

Reviewer #1 (Public Review):

This manuscript compiles the colonization of shrubs during the Late Pleistocene in Northern America and Europe by comparing plant sedimentary ancient DNA (sedaDNA) records from different published lake sediment cores and also adds two new datasets from Island. The major findings of this work aim to illuminate the colonization patterns of woody shrubs (Salicaceae and Betulaceae) in these sediment archives to understand this process in the past and evaluate its importance under future deglaciation and warming of the Arctic.

The strength of evidence is solid as methods (sedimentary DNA) and data analyses broadly support the claims because the authors use an established metabarcoding approach with PCR replicates (supporting the replicability of PCR and thereby proving the occurrence of Salicaeae and Betulaceae in the samples) and quantitative estimation of plant DNA with qPCR (which defines the number of cycles used for each PCR amplification to prevent overamplification). However, the extraction methods need more explanation and the bioinformatic pipeline is not well-known and needs also some further description in the main text (not only referring to other publications).

The authors compare their own data with previously published data to indicate the different timing of shrubification in the selected sites and show that Salicaceae occurs always like a pioneer shrub after deglaciation, followed by Betaluaceae with a various time lag. The successive colonization of Salicaceae followed by Betulaceae is explained by its differences in environmental tolerance, the time lag of colonization in the compared records is e.g. explained by varying distance to source areas.

However, there are some weaknesses in the strength of evidence because full sedaDNA plant DNA assessment, quality of the sedaDNA data (relative abundance and richness of sedaDNA plant composition) and results from Blank controls (for sedaDNA) are not fully provided. I think it is important to show how the plant metabarcoding in general worked out, because it is known that e.g. poor richness can be indicative of less preserved DNA and a full plant assessment (shown in the supplement) would be more comprehensive and would likely attract a larger readership.

Further, it would allow us to see the relative abundance in changes of plants and would make it easier to understand if the families Salicaeae and Betulaceae are a major component of the community signal. Further, the possibility to reach higher taxonomic resolution with sedaDNA compared to pollen or to facilitate a continuous record (which is different from macrofossils) is not discussed in the manuscript but should be added. Also, the taxonomic resolution within these families in the discussed datasets would be of interest, also on the sequence type level if tax. assignments are similar.

Another important aspect is how the abundance/occurrence of Salicaceae is discussed. Many studies on sedaDNA confirm an overrepresentation of this family due to better preservation in the sediment, far-distance transport along rivers, or preferences of primers during amplification etc. As this family is the major objective of this study, such discussion should be added to the manuscript and data should be presented accordingly.

I also miss more clarity about how the authors defined the source areas (refugia) of the shrubs. If these source areas are described in other literature I suggest to show them in a map or so. Further, it should be also discussed and explained more in detail which specific environmental preferences these families have, this is too short in the introduction and too unspecific. Also, it would be beneficial to show relative abundances rather than just highlighted areas in the Figures and it would allow us to see if Salicaeae will be replaced by Betulaceae after colonizing or if both families persist together, which might be important to understand future development of shrubs in these areas.

The author started a discussion about shrubification in the future, but a more defined evaluation and discussion of how to use such paleo datasets to predict future shrubification and its consequences for the Arctic would give more significance to the work.

Reviewer #2 (Public Review):

Harding et al have analysed 75 sedaDNA samples from Store Vidarvatn in Iceland. They have also revised the age-depth model of earlier pollen, macrofossil, and sedaDNA studies from Torfdalsvatn (Iceland), and they review sedaDNA studies for first detection of Betulaceae and Salicaceae in Iceland and surrounding areas. Their Store Vidarvatn data are potentially very interesting, with 53 taxa detected in 73 of the samples, but only results on two taxa are presented. Their revised age-depth model cast new light on earlier studies from Torfdalsvatn, which allows a more precise comparison to the other studies. The main result from both sedaDNA and the review is that Salicaceae arrives before Betulaceae in Iceland and the surrounding area. This is a well-known fact from pollen, macrofossil, and sedaDNA studies (Fredskild 1991 Nordic J Bot, Birks & Birks QSR 2014, Alsos et al. 2009, 2016, 2022) and as expected as the northernmost Salix reach the Polar Desert zone (zone A, 1-3{degree sign}C July temperature) whereas the northernmost Betula rarely goes beyond the Southern Tundra (zone D, 8-9{degree sign}C July temperature, Walker et al. 2005 J. Veg. Sci., Elven et al. 2011 http://panarcticflora.org/ ).

My major concern is their conclusion that lag in shrubification may be expected based on the observations that there is a time gap between deglaciation and the arrival of Salicaceae and between the arrival of Salicaceae and Betulaceae. A "lag" in biological terms is defined as the time from when a site becomes environmentally suitable for a species until the species establish at the site (Alexander et al. 2018 Glob. Change Biol.). The climate requirement for Salicaceae highly depends on species. In the three northernmost zones (A-C), it appears as a dwarf shrub, and it only appears as a shrub in the Southern Tundra (D) and Shrub Tundra (E) zone, and further south it is commonly trees. Thus, Salicaceae cannot be used to distinguish between the shrub tundra and more northern other zones, and therefore cannot be used as an indicator for arctic shrubification. Betulaceae, on the other hand, rarely reach zone C, and are common in zone D and further south. Thus, if we assume that the first Betulaceae to arrive in Iceland is Betula nana, this is a good indicator of the expansion of shrub tundra. Thus, if they could estimate when the climate became suitable for B. nana, they would have a good indicator of colonisation lags, which can provide some valuable information about time lags in shrub expansion (especially to islands). They could use either independent proxy or information from the other species recorded in sedaDNA to reconstruct minimum July temperature (see e.g. Parducci et al. 2012a+b Science, Alsos et al. 2020 QSR).

The study gives a nice summary of current knowledge and the new sedaDNA data generated are valuable for anyone interested in the post-glacial colonisation of Iceland. Unfortunately, neither raw nor final data are given. Providing the raw data would allow re-analysing with a more extensive reference library, and providing final data used in their publication will for sure interest many botanists and palaeoecologist, especially as 73 samples provide high time resolution compared to most other sedaDNA studies.