Interactive comment on “ Can a bog drained for forestry be a stronger carbon sink than a natural bog forest ? ”

The manuscript by Hommeltenberg et al. reports on two consecutive years of carbon dioxide flux measurements made above two bog forest sites. Thereby, the authors compare a natural bog forest with a drained bog forest, which are in near proximity of eachother. Such “paired” sites are rare and strongly needed to further understand carbon dioxide exchange of bog forests. Moreover only few studies of bog forest exists for the temperate region. Therefore the represented manuscript is of great interest for the readers of Biogeosciences. Unfortunately, the authors present carbon dioxide fluxes only, whereas methane fluxes are likely to contribute considerably to the total carbon budget of such big ecosystems. However, the calculations made at the end of the manuscript, considering past and current land use are strongly needed and contain valuable information.

The manuscript Hommeltenberg et al. reports very interesting dataset of two-years eddy covariance measurements of CO2 fluxes conducted at two different forested peat bogs with different land use history, but with the same genesis, climate and age. The paper is of high scientific importance, as comparative studies of pristine and nearbylocated natural forested bogs are rather rare in literature. The authors came to conclusion that 44-years old forest plantation of Picea Abies and Pinus silvestris on a previously drained bog are stronger CO2 sink then the near-pristine forested bog, with a big age span of Pinus mugo (from sapling to 150 yr). The differences in age of both ecosystems (the natural one have already reached a climax state, while the drained one is still in a phase of dynamic growth) should be well addressed in the paper, as it might have impact on the analyses performed and conclusions. Although analyses are performed based only on two-years of measurements, which might lead to mislead-C619

Interactive Comment
Full Screen / Esc

Printer-friendly Version
Interactive Discussion Discussion Paper ing, uncompleted conclusions, I consider this dataset long enough and scientifically extremely interesting to be presented in BIOGEOSCIENCE. However, this is very pity that the methane fluxes were not estimated for both sites based on EC or chamber measurements, as to have overall impression about a total carbon budget of both sites, this would be very crucial. As the authors are focusing only on CO2 budgets, I would not find it critical to include CH4 in a reviewed paper. The paper is very well structured and is written with very good, fluent language.
I have some concerns and a number of suggestions, that I believe will improve this manuscript once addressed.
Major comments: 1) The analyses of CO2 budgets for both sites are based on measurements conducted with two different EC approaches, which may biased the calculated CO2 budgets. Were these two systems compared at the beginning of experiment to demonstrate and prove that the fluxes are comparable? You may also add a sentence supported with some reference that these EC systems are working comparably. Although both systems have different advantages and disadvantages, the data loss should be higher in case of LI7500, especially during wintertime, but also in summer when raining, hence fluxes estimated for natural site for these periods of the year might be more biased as they are more related to modeled values.
2) The authors refer (Page 2194 lines 25-28, and page 2204, lines 25-30) to the old map of peat depth from 1940 concluding that due to anthropogenic activity the peat layer decreased by 1 meter within the last 70 years. And based on this assumption the estimated total carbon loss from this bog within 44 years between 1967 and 2011. Although this might be a truth, I would not trust so much the old maps, as we do not know how they were prepared and what is accuracy of a peat layer depth assessment. As this is uncertain I would use e.g. expressions as" potential peat loss" or "possible peat loss" on Page 2194 lines 25-28 If these values are comparable, then the analyses conducted at the drained bog would be more strong and trustful. This imply also later analyses conducted in chapter 3.5.
3)Footprint analyses were not performed for natural site (Page 2196 lines 24-26). As Authors do not mention about the height of the EC installation above canopy level (we know only the height of the towers) we do not have good impression about how big the footprint might be for the site. What is more, we do not know even what is the prevailing wind direction. The height of EC tower at the natural site is 6 meters (in average 4 meters above canopy level) while in the drained site -30 meters (9 meters above canopy level). Hence: 1) first specify what is your area of interest at natural site -looking on figure 1a we have bog-pine forest nearby the tower surrounded by single pine trees area-is this area also considered? If yes, please write it clearly in the methods section, otherwise, if we consider only bog-pine forest, then, according to my estimation, we have from 50 to 250 meters to the edge of this area. That means that your footprint should be bigger, assuming that EC system is installed 4 meters above canopy level. The same refer to the drained site, where we may suppose that EC system is 9 meters above canopy level and there is only 300-350 meters to the bog border in the west-east direction (but here luckily the footprint analyses were performed and data are filtered). Please deliver information about the footprint size and prevailing wind direction (wind rose) for both sites in the methods) Other minor comments and suggestions: Page 2193 lines 10-15 from the description of the natural site we know that the northern part of the natural site was affected by cutting and restored in 2001? How big was the degraded part of the peatland, how the site was restored and how far is this area from the footprint area of EC tower. This is just a question, If this restored peatland area may affect the measured fluxes? I think it might be useful for the interpretation of the data to add the main footprint area to the figures of both sites. Pages 2193, 2194, 2195, LAI -how LAI was derived? If it was derived just from measurements conducted with optical method (based on Sunscan DELTA-T system), as it is written in page 2195, then the values given in the paper are related to an effective Plant Area Index (PAI), which includes foliage but also branches and stems.
To derive LAI you need to subtract the area of branches and stems from PAI, or which is more reasonable -to change LAI to PAI in the paper. Another issue is, if the PAI values presented in the paper are related to trees only, or includes also sedges etc. The optical method used by the author assumes that PAR is measured just above the soil surface (in lowest part of the canopy) and above canopy level (where the reference PAR measurements should be done and not PAR should be used in the text and in figures. Page 2195 line 21, 1) I think it should be "indicating" instead of "integrating" 2) once analyzing the CS616 data You should consider that the sensors were originally calibrated for mineral soils, hence the measurements conducted in peat might be highly uncertain, especially in natural site, where the average WTD level was higher than 10 cm below the peat surface (3/4 of TDR was permanently in water) From the manual of the Campbell CS616 we know: " These coefficients should provide accurate volumetric water content in mineral soils with bulk electrical conductivity less than 0.5 dS m-1, bulk density less than 1.55 g cm-3, and clay content less than 30%." Page 2195 line 25 , 1) use 0.1 m instead of 10 cm; 2) use "thermistores probes" instead of "T-107-probes" and put T-107 to brackets. This is really a pity that peat temperature was measured only at one depth at Schechenfilz site, where there was observed rather high WTD level of a few cm below the surface. That means, measurements of T refer mostly for the water saturated peat layer for most of the analysed period. In fact, this may result in a low daily and seasonal variation of T, which might be used to Reco estimation. If yes, then I suppose that Reco fluxes might be underestimated for this site, what finally may bias the estimation of GEP.
Page 2196 line 16, data coverage of 91% for Mooseurach and 71% for natural site. . . is it correct, considering longer periods with gaps and more periods when Mooseurach station was not working correctly I would say that it should be in the opposite.
Page 2197 lines 18 -number of missing values might be slightly misleading, as only 9% of data are missing at drained site and the other 29% were rejected at this site because they were outside the footprint area Page 2198 line 10, -1) use 0.1m instead of 10 cm. 2) As mentioned before-I would be careful with using 10cm depth peat temperature to model Reco at natural site, as T sensor was in a water saturated peat layer by most of the year. How justify to use this T, considering that most CO2 is produced in the near-surface unsaturated peat layer  Fig.1. 1) I would increase slightly the font size of legend, 2) why there is a word "Key" before names of the sites? Is it colloquial name reflecting the shape of the site? If yes, I would recommend to not use it. Fig.3 . 1) Please add the height of air T measurements, 2) depth of soil T measurements, 3) use volumetric water content (%) instead of water content -here you should add that this is the average for 30 cm upper peat layer, 4) average WTD Fig. 4. 1) Are the figures 4a and 4b related to the same periods of 2011? I assume not, at the same PAR in winter and summer GPP should be different.
As it may lead to confusion, please explain it in the figure caption. 2) specify the height of Tair measurements. 3) use PPFD instead of PAR 4) why not to use 30 minutes averages, instead of averages of 100 half-hourly measured values? Fig.5 . please explain what you mean by "fraction of expected Reco" -is this REco normalized with temperature? Fig. 6 . I would suggest to use "period" instead of "year" Table 1. I would suggest to calculate sum of precipitation and average T and RH for 12 month periods analyzed in the paper between July 2010 to June 2011, July 2011-June 2012. This would be more useful for the interpretation of the data. Use RH, instead of rH Interactive comment on Biogeosciences Discuss., 11, 2189, 2014. C625