the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Jan 2024
Divergent responses of evergreen needle-leaf forests in Europe to the 2020 warm winter
Abstract. Relative to drought and heat waves, the effect of winter warming on forest CO2 fluxes during the dormant season has less been investigated, despite its relevance for net CO2 uptake in colder regions with higher carbon content in soils. Our objective was to test the effect of the exceptionally warm winter in 2020 on the winter CO2 budget of cold-adapted evergreen needle-leaf forests across Europe, and identify the contribution of soil and air temperature to changes in winter CO2 fluxes in response to warming. Our hypothesis was that warming in winter leads to higher emissions across colder sites due to increased ecosystem respiration. To test this hypothesis, we used 98 site-year eddy covariance measurements across 14 evergreen needle-leaf forests (ENFs) distributed from north to south of Europe (from Sweden to Italy). We used a data-driven approach to quantify the effect of air and soil temperature on changes in net ecosystem productivity (NEP) during the warm winter of 2020. Our results showed that the impact of warming was different across sites, as in the lower altitude and lower latitude sites positive soil temperature anomalies were larger, while positive air temperature anomalies were larger in the northern latitude and high-altitude sites. Warming in winter led to a divergent response across the sites. Out of 14 sites only in 3 sites net ecosystem productivity declined in winter significantly in response to warming. In addition, we observed that in the colder sites daytime NEP (that is dominated by photosynthesis) declined with warming of the air in winter, whereas in the warmer sites daytime NEP increased with warming of the soil. This shows that warming of the air – if not translated into a direct warming of the soil– might not trigger productivity in winter if the soil within the rooting zone remains frozen. Forests within the same plant functional type category can exhibit differing reactions to winter warming and to predict their responses accurately it is crucial to account for variations in local climate, physiology, and structure simultaneously.
- Preprint
(3152 KB) - Metadata XML
-
Supplement
(2217 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2964', Anonymous Referee #1, 19 Feb 2024
The theme holds significant importance, as they aim to assess the impact of a warm winter on CO2 fluxes through the examination of flux tower data. The findings indicate that winter warming triggers a divergent ecosystem response. Additionally, observations reveal that in colder locations, daytime Net Ecosystem Productivity (NEP), primarily driven by photosynthesis, decreases with warming air in winter. Conversely, in warmer sites, daytime NEP increases with soil warming.
However, the paper exhibits several big issues:
The introductions lack clarity and a well-organized structure. Introductions should avoid incorporating captions.
The introduction lacks logical coherence, for example, Lines 78-83 dedicates excessive space to elaborating on physiological responses, which is not the primary focus of this study.
Table 3 could be presented graphically.
The analysis focuses solely on solar radiation and temperature, neglecting the consideration of moisture limitations. RF model should also consider moisture variables.
Furthermore, the potential collinearity between soil temperature and air temperature can impact the accuracy and reliability of the model results.
To enhance the manuscript, a concerted effort should be made to streamline the content, maintain logical progression, and incorporate necessary elements for a thorough analysis.
Citation: https://doi.org/10.5194/egusphere-2023-2964-RC1 -
RC2: 'Comment on egusphere-2023-2964', Jonas Lembrechts, 25 Feb 2024
This is an interesting and ambitious attempt to disentangle the impact of an extremely warm winter on forest productivity (the netto balance between respiration and CO2 uptake). Using high-resolution (temporal) data from multiple forest sites across Europe, linked to locally measured environmental conditions, the authors try to disentangle a) how local microclimatic conditions were different from baseline conditions the years before, and b) how these differences impact winter productivity.
While I applaud this effort, the paper might be falling short of answering the actual questions convincingly, perhaps largely due to the extreme complexity of the whole system. Effects are highly site-specific, and the analysis fall short – if I understand all correctly – of showing the direction of trends resulting from the differences in temperature between the warm year and baseline conditions.
I’m not sure the answer to the above is more analysis – the paper already has plenty of figures trying to make sense of the complex story – but perhaps a refocus towards figures that synthesize the actual relationship between NEP changes and temperature differences, both within and between sites is needed to bring the story. Also, the figures that are there might need some clarifications to make them more intuitive (see comments at the end).
I have a bunch of more detailed comments below, which I hope identify where for me as a reader the uncertainties arise.
L95: why does the risk for photo-oxidative frost damage increase? Is that due to the lost winter hardiness mentioned earlier? Might be good to make the reason explicit.
L99: ‘the interaction of light quality and photoperiod’: unclear to me what this implies
L103: ‘thus’: this word implies that the previous sentences explain why cold periods play an important role in forming the photosynthetic capacity, but to me there is a step missing: why are these pathways resulting in improved photosynthetic capacity. The previous paragraph hints to this, perhaps, by mentioning that a lack of these would result in damage (and thus reduced photosynthesis as a result). But all of this feels a bit implicit and disconnected. Just adding a sentence might already help.
L104-106: now here you continue with examples. Again, I feel the need for a better structuring of this introduction, disconnecting the theoretical cause-and-effect relationship from the examples.
L112: might there be a need to define respiration, or can we assume this concept to be sufficiently well-known?
L131-132: this sentence is a bit too dense in information for me to fully understand
L133-135: I feel like we’re missing some information to make the distinction between direct and indirect effects on respiration clear: what happens in the direct pathway, and can you give some examples on how the other factors are affecting respiration indirectly?
L140: to me the ‘on record’ doesn’t match too well with the cut-off of 1981, as intuitively I think we have older records than that (although not in that source). Perhaps rephrase to ‘in the last four decades’?
L150: soil temperature comes a bit out of the blue here – there has been very little discussion on how it can affect respiration in addition to air temperature. Similar for radiation, for which it’s even the first time the word is mentioned.
L192: ‘the forest’?
L194: what do you mean with climate variables that overlapped? Data availability within your dataset? Or overlapping values?
L197: for those sites that measured snow depth, could you make a test of the accuracy of the remotely-sensed snow depth measurements, as this is notoriously hard to get right?
L216: ‘were’
L235: there is a ‘the’ that doesn’t seem to fit there
L235: what exactly are these anomalies here?
L247: ‘lowest positive anomaly of 0.87°C’: what about the even smaller anomalies in Fig. 3, in FR-Bil and DE-RuW?
L254: first time soil temperature depth is mentioned – feels like more something for the methods
L263: 315 days? Table says 365
L272: ‘IT-SR2… shifted towards being a smaller source in winter’ Is that true? I thought that we saw in Supplementary Table 2 that the value actually showed it turned into a CO2 sink?
L274-275: why are the values described here different from those in Fig. 6? Are they describing something different? If so, why refer to Fig. 6?
L276: ‘increased significantly’ -> remind the reader that increasing here means a decrease in value? As the previous comments also reflect, this part was pretty hard to follow.
L285: ‘winter respiration fluxes’: remind the reader which panel in the figure that is
L324: where do we see this result?
L325: ‘high latitude-high elevation’: this is either/or, right, not both together? Unclear from the way it’s written
L329: ‘dominated’: is the difference that strong? Can you quantify that?
L333: what are the implications of those relationships in Table 3? They feel a bit disconnected from the story now (although, as I ask down below regarding Fig. 7 and 8, they might be critical)
L336-339: the role of LAI feels a bit like an afterthought now, and the way it’s currently analyzed (basically qualitatively rather than quantitatively) makes it hard to build strong conclusions on it. You could at least model the relationship between Tair-Tsoil and LAI?
L435: weaker snow buffering effect often results in lower winter temperatures, as snow tends to buffer against freezing temperatures. This makes sense, as snow is usually present when air temperature is below zero, and then soil temperature stays at zero. You see this in Lembrechts et al – Global maps of soil temperature – that soil temperatures in cold regions are usually higher than air temperatures. Here, you have a reduced snow cover, so the only way in which soil temperature can be higher, is if air temperature is also unusually high (i.e., positive).
L436-438: can you specify which sites this are? I’m getting a bit lost in trying to link the different figures on snow and temperature.
L449: Supplementary Figure 5 shows very little trend to me (the blue and red dots are basically randomly distributed?
L445 and following: in these paragraphs lies one of my main questions. Here indeed you are connecting everything together (soil temperature, air temperature, NEP, …). You have an analysis (the machine learning model) that does this as well, but I am missing figures and/or analyses of the relationships between these things. Now, it requires juggling of all figures and tables to connect everything together, but the machine learning model is only used to show the SIZE of the effect, not the actual relationships themselves.
For example, if you say: ‘warmer sites however (low altitude or low latitude sites) winter warming also increased the productivity and CO2 uptake’ then this should be supported with a figure showing delta productivity/CO2 uptake as a function of delta T in winter and background T, and their interactions (for example, a separate line for the relationship for warmer versus colder sites).
Similarly, if you say (L453-454) that when soil temperature reaches above freezing level, CO2 uptake increases, this should be shown by a figure showing daytime NEP as a function of soil temperature.
These are just two examples, but this comment is valid throughout the chapter. The main conclusions (the relationships between local climatic conditions and NEP are not really shown in the results, if I’m not mistaken.
L461: how does Fig. 7 show that baseline climate conditions are a good proxy for this? Do you mean that if you order sites from warm to cold that there is a rough trend emerging in the amount of variance explained? If so, then I’m not super convinced that 1) this is a ‘good proxy’, and 2) that it says anything on ‘how’.
L472: perhaps add half a sentence on what a higher Q10 means in practice for these soils.
L474: where do these labile C inputs have to come from?
L475: Supplementary Figure 3 is not mentioned in the results, so it’s very hard to link this to the story. In this figure, the differences between 2020 and reference are also very hard to spot. If the story in L445 and following is indeed true, then it should show up somehow in figures correlating NEP to temperature (or better perhaps, delta NEP to delta temperature)
Fig. 2 and Supplementary Fig. 1: x- and y-axis labels are not entirely intuitive, yet not explained
Fig. 3: anomalies are in °C?
Fig. 3: unclear from this figure which sites are high latitude or high-altitude sites (L 246)
Fig. 6: could you change the color scheme so it is white at zero?
Fig. 7: ‘overall variable explained’, shouldn’t that be variance?
Fig. 7: ‘three climatic variables’: why is there a fourth one – not mentioned in the legend – for the bottom panel? Even more confusing: it doesn’t seem to be mentioned in the main text on Fig. 7 either?
Fig. 7 and 8: isn’t there a correlation between Tsoil and Tair? If so, how can the model decide which of the two explains the variance (and have this sum up to 100%)? This is different from the way I’m used to variance partitioning,
Supplementary Fig. 1: SE-Ros or FI-Ros?
Supplementary Fig. 4: legend doesn’t seem to explain the figure
Figure numbers are not always in the right order throughout the manuscript, which muddies the water unnecessarily.
Kind regards,
Jonas Lembrechts
Citation: https://doi.org/10.5194/egusphere-2023-2964-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 393 | 75 | 20 | 488 | 27 | 11 | 6 |
- HTML: 393
- PDF: 75
- XML: 20
- Total: 488
- Supplement: 27
- BibTeX: 11
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1