A Preconditioning Paradox: Contrasting Effects of Initial Phyllosphere and Early Leaf Decomposer Microfungi on Subsequent Colonization by Leaf Decomposing Non-Unit-Restricted Basidiomycetes
Responses to reviewers’ comments and suggestions
Reviewer 1:
The title of the article is not clear. It is recommended to rewrite the title of the article.
We have changed the title to:
A preconditioning paradox: contrasting effects of initial phyllosphere and early leaf decomposer microfungi on subsequent colonization by leaf decomposing non-unit-restricted basidiomycetes.
The abstract should be rewritten by detailing the aim and concept of the manuscript. The abstract should state briefly the purpose of the research, the principal results and major conclusions.
It has been rewritten as:
Abstract: Fungal interactions during leaf decomposition can facilitate or inhibit subsequent colonizers via interference or changes to substrate quality. This experiment focused on whether preconditioning of leaf litter by microfungi that were confined to one leaf (Unit-Restricted) made leaf litter less likely to be colonized and decomposed by basidiomycetes that bind litter into mats (Non-Unit- Restricted, NUR) than non-preconditioned litter. Leaves of Manilkara bidentata in litterbags were preconditioned by incubating them for 0, 1, 2 or 3 months in flat baskets 10 centimeters above the forest forest floor to prevent colonization by basidiomycete and other NUR fungi. Preconditioned and non-preconditioned leaves were simultaneously transferred to 5 replicate basidiomycete fungal mats of Gymnopus johnstonii for 6 weeks. Both attachment by basidiomycete fungi via root-like structures and percent mass loss after 6 weeks decreased significantly with increasing preconditioning times of 1-3 months, consistent with our substrate degradation hypothesis.. In non-preconditioned leaves, more basidiomycetes attached to non-irradiated than to irradiated leaves which suggests phyllosphere microfungi facilitated basidiomycete colonization via substrate changes. While basidiomycete colonization was initially facilitated by phyllosphere fungi, we inferred that degradation of resource quality led to fewer fungal attachments and less mass loss after 1-3 months of preconditioning. The data indicate that there is a 1-month time window for basidiomycete fungi to incorporate fallen leaves into their litter mats.
Introduction is very general and need to be elaborative to explore the actual philosophy to design the experiment. The introduction is insufficient to provide the state of the art in the topic.
Response: We asked for clarification on this, but in the meantime, received helpful suggestions from reviewer 3. We have therefore revised the introduction according to the 3rd review.
In lines 153 and 154 specify the name of the radiation equipment used, as well as the model and the exposure time
Response: The radiation equipment used was custom-built, so it does not have a make and model number. The exposure time is irrelevant since we report the amount of radiation absorbed by the sample in rad units. “The rad is a unit of absorbed radiation dose, defined as 1 rad = 0.01 Gy = 0.01 J/kg.[1] It was originally defined in CGS units in 1953 as the dose causing 100 ergs of energy to be absorbed by one gram of matter.” It is not dependent on the type of material irradiated.
The reworded sentences are:
This removed most of the microorganism present on the leaf litter without changing the leaf composition and structure [18]. We used a custom-built 137Cs irradiator at the University of Puerto Rico Medical Sciences Campus, with absorbed doses of 3640 rad and 4090 rad on the first and second exposures, respectively [8].
In section 2.6. Measurements, clarify which variables were monitored, units, and reference.
Sentences were augmented as:
We harvested the litterbags after 6 weeks and counted the number of discrete attachments (i.e., hyphal strands, rhizomorphs and cords) connecting the leaves in the litterbags to the white-rot basidiomycete litter mat below directly in the field. We accomplished this by slowly rolling the litterbag beginning at one edge, counting each attachment as it broke [13].
Mass loss (g)
Percent of leaf area decomposed by white-rot basidiomycetes was estimated after spreading leaves on a 1 cm gridded background, counting the squares more than 50% covered by leaves, then the squares more than 50% covered by bleached leaf surface to obtain cm2 total leaf area and white-rot leaf area, respectively [26], then calculating % white-rot as cm2 white-rot/cm2 leaf area x 100.
In line 311 corresponding to the conclusions, restructure them according to the 2 hypotheses raised
Sect. 5, Conclusions – we have reorganized this by the two hypotheses, as suggested.
In the results section, it is recommended to clarify figures 2 and 3.
Figure 2 was deleted, as suggested by reviewer 2. Figure 3 (renumbered Figure 2), we expanded the caption for a) Number of basidiomycete fungal attachments (hyphal strands, cords and rhizomorphs) connecting leaves in litterbags to the white-rot litter mat below.
Reviewer 2:
Materials and methods. The part of the paper devoted to the collection of leaves, their preparation for field experiments, experimental options, processing of the obtained materials is written quite fully. However, there are also questions regarding the methodology of the work.
- Why did the authors choose Gymnopus johnstonii?
We added this sentence:
We selected G. johnstonii because it was the most abundant basidiomycete forming white-rot litter mats at our site [11, 13], it can be identified without basidiomes based on the white surficial mycelial fans [13], it causes rapid mass loss [12], and it forms discrete hyphal strands connecting leaves that are easily counted [13].
How and how many leaf bags were put on a one white-rot basidiomycete litter mat of G. Johnstonii?
5. We added this sentence: Five litterbags were placed on each mat.
How many such mats were there?
6. We modified the sentence to read: placing them directly on 6 white-rot basidiomycete litter mats of G.
7. What are "attachments" and how were they counted?
We augmented the first sentence as:
We harvested the litterbags after 6 weeks and counted the number of discrete attachments (i.e., hyphal strands, rhizomorphs and cords) connecting the leaves in the litterbags to the white-rot basidiomycete litter mat below directly in the field. We accomplished this by slowly rolling the litterbag beginning at one edge, counting each attachment as it broke [13].
How was the % white-rot calculated? All of these, undoubtedly, key questions for work under review and It is necessary to describe in detail these.
We modified the text as follows:
Percent of leaf area decomposed by white-rot basidiomycetes was estimated after spreading leaves on a 1 cm gridded background, counting the squares more than 50% covered by leaves, then the squares more than 50% covered by bleached leaf surface to obtain cm2 total leaf area and white-rot leaf area, respectively [26], then calculating % white-rot as cm2 white-rot/cm2 total leaf area x 100.
Explain how objective indicators of the processes of decomposition of plant residues they are and who is the author (s) of this method?
We measured both % white-rot and number of fungal connections between the litter mats and the leaves in litterbags.
For number of fungal connections versus mass loss, we added the following sentence:
We accomplished this by slowly rolling the litterbag beginning at one edge, counting each attachment as it broke [13]. Lodge et al. [13] found at our site that mass loss in senesced leaves at 14 weeks was significantly predicted by the abundance of fungal connections between the senesced litter cohort and forest floor at 7 weeks.
For % white-rot versus mass loss, we added citation [8] (Lodge et al. 2008), but further details are better left to Introduction and discussion as it is more complex. In the introduction, we say:
Basidiomycete decomposers can degrade lignin-like compounds and lignocellulose producing a white-rot, which accelerates decomposition [12, 15, 17]. Santana et al. [8] found in wet tropical forest that a white-rot basidiomycete that survived two rounds of gamma-irradiation in leaf litter in half of the microcosms containing leaves inoculated with early decomposer microfungi lost 21% more mass than leaves in microcosms containing only microfungi.
Most experiments, including Santana et al. 2005, Lodge et al. 2008, Osono, Cromack, Hintiika) have involved presence or absence of white-rot and effects on accelerating the rates of decomposition (15-22% faster with white-rot). However, Lodge et al. (2008) showed that leaf litter decomposed for 3 months on white-rot litter mats lost mass significantly faster than leaves placed off of white-rot litter mats (40% vs. 31% mass loss). We added this citation [8].
Unfortunately, some basidiomycete mycelia moved out from under the white-rot mat treatment and under leaf bags in the non-white-rot treatment, so the white-rot treatment had 60% white-rot whereas the non-white-rot treatment had 20% white-rot. Nevertheless, it showed a general correlation between % white rot and % mass loss.
A subsequent 3-month experiment in which the mycelia were ‘caged’ in or out is included in an invited, submitted book chapter for CRC press by Lodge and Cantrell on the roles of macrofungi. That experiment showed a 15.3% increase in mass loss rate attributed to % white-rot. This experiment in the book chapter can’t be cited as in-press yet, but the section is quoted below:
“Agaric fungi that produce white-rot have been shown to increase rates of mass loss in leaf litter in boreal, temperate and tropical litter (Hintiika, 1970; Santana et al., 2005; Lodge et al. 2008; Osono and Takeda, 2002, 2006). Experiments by Hintiika (1970) and Santana et al. (2005) used litter placed in microcosms while Osono and Takeda (2002, 2006) incubated sterilized litter inoculated with fungal cultures placed on agar. Santana et al. (2005) found that tropical leaf litter mass loss in microcosms containing a white-rot basidiomycete in addition to a dominant microfungal ascomycete was 22% higher than those with an ascomycete alone. Osono and Takeda (2002, 2006) found higher rates of leaf litter mass loss from decomposer basidiomycetes was generally higher than loss from ascomycetes and zygomycetes.
Field experiments to determine the effects of white-rot on leaf decomposition rates are more challenging. Lodge et al. (2008) conducted a field experiment at Sabana, Puerto Rico by placing 10 g dry wt of freshly fallen Dacryodes excelsa leaf litter enclosed in 1-mm mesh litterbags either on or off of white-rot leaf litter mats with 12 replicates per treatment and found 8.4% faster mass loss from leaves on the white-rot mats despite the invasion of the non-white-rot treatment by white-rot fungi and out-migration of basidiomycete mycelia from the white-rot treatment over the 3-month trial. DJL conducted a modified field experiment at Sabana in Puerto Rico as described here using plastic containers with the bottoms removed to restrict lateral movements of white-rot basidiomycete mycelia in the litter layer (Fig. 6). White-rot accelerated mass loss by 15.3% over the 3-month trial. Because the amount of white-rot varied, including the presence of some white-rot in the non-white-rot treatment, the slope of percent mass loss was regressed against percent of leaf area bleached by white-rot to estimate the effect of white-rot on mass loss (Fig. 6c).”
Results. Figure 2 should be removed, since all information about itis given in the text, and the figure itself is not informative.
We agree, and Figure 2 has been deleted and Figure 3 changed to Figure 2.
Section3.3. “Basidiomycete fungal attachments and mass loss in gamma-irradiated leaves vs control”. After getting acquainted with it, the question arises - does gamma irradiation affect the processes of decomposition and colonization of leaves or does it not?
We have added this sentence to the end of the paragraph:
Thus gamma-irradiation only significantly affected (decreased) colonization by basidiomycete fungi of non-preconditioned leaves (i.e., freshly fallen air-dried leaves), while % mass loss was unaffected with either 0 or 3 months of preconditioning.
Discussion. In the introduction, formulated two working hypotheses, and it would seem that the discussion of the results should include two relevant sections. In the form in which the discussion is written, it is perceived very difficult.
We added the following subheadings and rearrange the text under them:
4.1 Substrate effects – preconditioning by unit-restricted microfungi reduces litter quality
4.2. Interference effects – phyllosphere early unit-restricted fungi inhibit basidiomycete colonization
Conclusions. The conclusion should contain clear answers of the authors of the publication to the working hypotheses formulated by them. The proposed version of the conclusion, in my opinion, is rather an short version of the discussion, rather than a conclusion. It is necessary to more clearly formulate the main results of the study
We have shortened and rewritten conclusions as follows:
We negated our second hypothesis as we found no evidence that unit-restricted UR microfungi interfered with subsequent colonization by non-unit-restricted (NUR) basidiomycete decomposers. Contrary to our expectation, initial UR phyllosphere fungi present in freshly fallen leaf litter facilitated colonization by the dominant NUR white-rot basidiomycete fungus in wet tropical forest of Puerto Rico, consistent with our first hypothesis that alteration of the substrate by early decomposers affects colonization by later decomposer fungi. Longer periods of preconditioning by UR microfungi, however, resulted in fewer NUR basidiomycete decomposer fungal attachments and slower mass loss, which is also consistent with our first, substrate hypothesis. In tropical premontane wet forest such as our site there may be a very short period of 2-4 weeks after litterfall when NUR basidiomycete colonization is facilitated by the activity of the initial phyllosphere fungi but before degradation of resource quality via leaching and utilization of soluble carbohydrates. Future studies would benefit from using high-throughput sequencing to determine the dominant fungi, patterns of enzyme production and the timing of microfungal community turnover during the critical first 4-6 weeks after leaf fall.
Reviewer 3
General comments
The authors picked an interesting topic, analysing the successional effects of early fungal litter colonizers on late-successional white-rot Basidiomycetes. Fresh litter samples were collected, air-dried, and “pre-conditioned” for 0, 1, 2 or 3 months 10cm above the ground. By that, fungi already present in the phyllosphere and saprobic soil fungi arriving by spores were already able to colonize and degrade the litter samples. These pre-conditioned litter samples were then placed on mats of one specific Basidiomycete species, that apparently forms big cord-forming hyphal mats in this tropical forest in Puerto Rico. Additionally, gamma-irradiated samples were compared at time 0 and 3 months, to evaluate the effect of direct competition among early colonizing fungi and the Basidiomycete. The authors assessed litter decomposition and hyphal cord connections as response variables.
The results are a bit variable, as expected in a tropical diverse forest. Five replicates is a bit low, but since several time points were included it seems ok. But there clearly seems to be an effect of time. Also, it is interesting that non-sterile litter samples are preferentially colonized by the hyphal cords, indicating a facilitative effect among the fungal groups.
I have several comments regarding the experimental design, that needs to be described in more detail (see detailed comments below). Most specifically, regarding the analyses at different time points of pre-conditioning, it would be highly relevant to know whether the litter was put onto the mats at the same period, or with one month in between for each time point. In this case, small variations in precipitation or temperature may have caused variation observed over time, independent of pre-conditioning phases. In my opinion, for the result of the pre-colonizing effects observed this is crucial, litter should be brought to the ground during the same 6 weeks periods.
Response; Excellent comment. All the different treatment bags were placed simultaneously on 25 May 2012 on each mat to reduce the variance due to environmental and invertebrate variation on the forest floor. We modified the sentence in Methods to read:
“On 25 May 2012 we simultaneously placed all treatments including non-preconditioned leaf bags directly on white-rot basidiomycete litter mats in a replicated complete block design to reduce variation due to environmental and invertebrate heterogeneity in space and time. One litterbag from each of the 5 treatments was placed on each mat.”
As described in detail below, I would also highly suggest to more clearly phrase the ecological context. There is a bit confusion what is meant by early colonizers, and phyllosphere fungi. Also, I am personally not a big fan of the concept of UR and NUR, especially not this clear division among phyla. Other fungal groups than cord-forming fungi can spread with hyphae, though at microscopic level. And more importantly, in this specific case actually only one species was evaluated as NUR. I would suggest to better introduce differences among early and late colonizers, and refer to this Basidiomycete mat as a cord-forming Basidiomycete species.
These are all good points.We have added ‘cords’ to the description of this basidiomycete, (but not replaced the concept of UR fung).
Technically, G. johnstonii doesn’t form cords because there is not a differentiated outer layer, but their hyphal strands are the functional equivalent of cords, but we are willing to use this term in order to better make the connection to the relevant literature. In the 6-week time frame of exposure of the leaf litter to the cord-forming basidiomycete mats, other groups of fungi were not observed by Lodge et al. (2014) to connect the forest floor to freshly placed leaf litter in our forest when examined after 7 weeks (Lodge et al. 2014). These microfungi were observed colonizing litter cohorts via hyphae by 14 weeks and their hyphae together with those of both cord-forming and non- cord-forming basidiomycetes were also clearly visible making connections between litter cohorts separated by screens after 3 and 6 months in Lodge et al. 2014. Microfungi colonized leaf litter earlier than 7 weeks if leaves come into direct contact, but placing the leaves in litterbags minimized direct contact between litter cohorts. Consequently, colonization by cord-forming basidiomycetes was favored over direct colonization by microfungi, at least in the 6-week time frame of exposure to cord-forming basidiomycete litter mats.
Regarding phyllosphere vs early decomposer fungi, phyllosphere fungi that are already present in the freshly fallen leaves do persist and grow in the leaf litter at least through the first 3-6 weeks of decomposition, so they are part of the early decomposer assemblage. Other microfungi colonize via spores and are also part of the early decomposers. We added the following text to the introduction:
“Early leaf decomposers include phyllosphere fungi (endophytes and secondary pathogens) as well as decomposers not found in live leaves. Santana et al. [8] cultured vegetatively dominant fungi from Manilkara bidentata leaves that were decomposed under their source trees for 6 weeks in Puerto Rico and found that half were endophytic fungi of the phyllosphere and the others were decomposer fungi that colonized from the forest floor).”
Early versus middle and late fungal colonizers of decomposing leaf litter is not clearcut when making comparisons across ecosystems. We have modified the first sentence of the second paragraph of Introduction and added three more as:
“Direct colonization of leaf litter by NUR basidiomycete cord-forming fungi in humid tropical forests are difficult to classify by stage as they can occur early or late in decomposition depending on nutrient to carbon ratios [1, 11–16]. Basidiomycete decomposers can degrade lignin-like compounds and lignocellulose producing a white-rot, which accelerates decomposition [12, 15, 17]. Dominance of basidiomycete decomposers in middle or late decomposition stages in temperate and boreal forests is associated with increasing accumulation of recalcitrant substrates such as lignin following degradation of more labile components of litter by microfungi, especially Ascomycota [2,7]. Cord-forming basidiomycetes may colonize leaf litter earlier in wet tropical forests than in forests of higher latitude in part because they typically have sclerophyllous leaves that are generally high in recalcitrant compounds [11,12]. In addition, freshly fallen leaves in wet tropical forests often have low P concentrations that resemble those in wood, so cord-forming basidiomycete fungi that can import P accumulated from previously decomposed litter to build biomass in P-deficient litter have a colonization advantage over UR fungi colonizing via spores [11,12].”
Specific comments:
L26 Maybe change to “The data indicate that..”
Last sentence of Abstract changed as suggested
L51 I have the feeling that the division in NUR and UR is less relevant than the enzymatic capacity of the fungi – with Basidiomycete forming cords, but also being able to degrade lignin but also other complex C compounds.
Response: L50-52 said: ”Basidiomycete decomposers can degrade lignin-like compounds and lignocellulose producing a white-rot, which accelerates decomposition [12, 15, 17].”, which agrees with the reviewer that the enzymatic ability of basidiomycetes that form cords to break lignin and other complex C compounds is important. However, it may not be the only factor since P is a growth-limiting nutrient in tropical wet forests, and translocation of P from a previous food base as a limiting mineral nutrient has also been shown to be related to and predictive of mass loss. Fungal connectivity by basidiomycetes between litter cohorts was predictive of P accumulation in litter in our forest.. We have added after the cited sentence:
”Dominance of basidiomycete decomposers in middle or late decomposition stages in temperate and boreal forests is associated with increasing accumulation of recalcitrant substrates such as lignin following degradation of more labile components of litter by microfungi, especially Ascomycota [2,7]. Cord-forming basidiomycetes may colonize leaf litter earlier in wet tropical forests than in forests of higher latitude in part because they typically have sclerophyllous leaves that are generally high in recalcitrant compounds [11,12]. In addition, freshly fallen leaves in wet tropical forests often have low P concentrations that resemble those in wood, so cord-forming basidiomycete fungi that can import P accumulated from previously decomposed litter to build biomass in P-deficient litter have a colonization advantage over UR fungi colonizing via spores [11,12].”
L53 Just out of interest, was that an accidental finding that it survived? I guess the aim of the experiment was something else? In the context of the introduction this is a bit confusing.
Response: Yes, it was accidental, and completely unexpected that a lone basidiomycete would survive two rounds of gamma irradiation that had killed all other fungi. It turned out to be just as interesting as the original experiment since it showed higher rates of mass loss from the basidiomycete than the dominant early decomposer fungi, and that the mass loss effects of the early decomposers and the basidiomycete were additive (not synergistic).
L61 The term phyllosphere fungi here and further below needs some explanation. I am not sure what the authors refer to. This seems to be mixing up of different theories and effects.
This study as it is framed seems to be about successional colonization of litter (after falling on the ground) by different fungal groups. Indeed, early colonizers often use more simple sugars and belong to Asco- and also Mucoromycota. While late colonizers are less fast growing with the capacity to degrade more complex C sources, including lignin. However, the strict division into UR and NUR is not clear to me. Rather cord- or non-cord-forming. I guess also Ascomycota can spread by hyphae from one leaf to the next in close proximity. Especially Mucoromycota have very fast growing mycelia, though admittedly also produce large amounts of spores at the same time. Thus, I would suggest to be more careful with this clear division of ecological lifestyles, especially when broadly relating them to whole phyla.
The other point is the phyllosphere fungi, where it is not fully clear to me what the authors refer to. In my understanding that is endophytes of the phyllosphere, that have early successional advantages by being the first decomposers present in decaying leaves. But those fungi will in short time be replaced or complemented by saprobic soil fungi arriving either by spores or hyphae. So there needs to be in my opinion a clearer definition of phyllosphere fungi, and early litter colonizing fungi.
Response: We agree with this understanding of phyllosphere fungi (with the addition of some latent pathogens). Our irradiation only removed phyllosphere fungi, so the distinction is important.We added the following sentence to clarify that the phyllosphere fungi include both endophytes and latent pathogens in the live leaves, and that while these are complemented by other early decomposers, they still comprised about half of the early decomposers in our leaves that were preconditioned for one month.
“Early leaf decomposers include phyllosphere fungi (endophytes and secondary pathogens) as well as decomposers not found in live leaves. Santana et al. [8] cultured vegetatively dominant early decomposer fungi from Manilkara bidentata leaves that were decomposed under their source trees for 6 weeks in Puerto Rico and found that half were endophytic fungi and latent pathogenic fungi of the phyllosphere and the others were decomposer fungi that colonized from the forest floor.”
And in this context it is also relevant to define to which of these groups the “pre-conditioning” refers to.
Response: We edited this sentence to read:
We therefore preconditioned leaf litter by UR early decomposers for different lengths of time (0, 1, 2 and 3 months), before transferring the litter to basidiomycete litter mats formed by Gymnopus johnstonii Murrill) A.W. Wilson, Desjardin & E. Horak, and quantified the number of fungal root-like attachments (cords) to the littermat, percent white-rot and mass loss after 6 weeks.”
L66 For readers not fluent in taxonomy, I would suggest to add in brackets the affiliation of these species, either phylum, or life-history strategy
Response: We edited these lines as suggested to read:
“individual early decomposer fungi belonging to the Ascomycetes on subsequent colonization by basidiomycete fungi [10]. They found that Mycena sp. (Basidiomycota, Agaricales) caused greater mass in leaf litter preconditioned by Xylaria sp. and Ascochyta sp. (Ascomycota)”
L143 Were the treatments mixed up in the baskets? In the picture it looks like the bags were in contact with each other, so fungi may have spread and communities become more similar over time. Also, litterbags at one position may be exposed to the same airborne spores. Thus, it may be relevant to consider this as a block effect in the analyses.
Response: Yes, the treatments were mixed in the baskets, so they were treated as random effects in the analyses. The only block or pairing effects we examined were for the cord-forming mats they were placed on. While there may have been greater similarity in microfungi within baskets, there were not enough degrees of freedom to analyze for two block effects.
Generally, the experimental design should be explained in more detail. It is not clear how these litterbags were distributed in the nets above the ground. But also not how they were placed on the soil after the preconditioning phase. How were mycelial mats selected, how many were used etc.
Response: The sentence in methods describing the 4 preconditioning baskets together with the photo seems clear to us, so we have not changed this: “Four baskets were built to expose bagged leaf litter to leaching by rain and colonization by unit-restricted microfungi for the three preconditioning treatment times.”
We agree that the methods regarding placement of the bags on the white-rot mats needed greater detail. We expanded the detail in experimental design for selection of littermats and placement as:
“On 25 May 2012 we simultaneously placed all treatments including non-preconditioned leaf bags and irradiated leaves directly on white-rot basidiomycete litter mats in a replicated complete block design to reduce variation due to environmental and invertebrate heterogeneity in space and time. One litterbag from each of the 5 treatments was placed on each mat. We selected G. johnstonii because it was the most abundant basidiomycete forming white-rot litter mats at our site [11, 13], it can be identified without basidiomes based on the white surficial mycelial fans [13], it causes rapid mass loss [12], and it forms discrete hyphal strands (cords) connecting leaves that are easily counted [13]. Only mats that were at least 0.5 m diameter were selected.”
Furthermore, I wonder how the time may be relevant. Preconditioning was done for 0, 1, 2 and 3 months. Which makes sense. But then it was placed on the ground during different times of the wet season? Or was the reconditioning phase started at different time points and meshbags brought out simultaneously?
See response above - they were all placed simultaneously.
L167 Why were the leaves spreaded on this grid? Please clarify
We agree that these methods needed greater detail. The augmented text reads:
“Percent of leaf area decomposed by white-rot basidiomycetes was estimated after spreading leaves on a 1 cm gridded background, counting the squares more than 50% covered by leaves, then the squares more than 50% covered by bleached leaf surface to obtain cm2 total leaf area and white-rot leaf area, respectively [26], then calculating % white-rot as cm2 white-rot/cm2 total leaf area x 100.”
L175 I have to admit that I never heard of the PAGE test before. I am not sure this is really the test you need (after a quick google recherche what that is), please correct me if I´m wrong.
In case of an ANOVA with observed heterogeneity of variances, I would recommend to either run a general least square method (gls() in R) including the weights function. Or in case you run an lme in R with block (mat) as random factor, you can include a weights function right in the model.
In both cases time can be modelled as a factor or continuous variable, as needed
Response: The reviewer is correct that a Page test is a non-parametric ANOVA-like analysis. Our number of replicates is low (5), which means that non-parametric analyses have more power to detect differences than parametric ones, and it is appropriate for our data. If we had more replicates then the methods found using Google would be useful. The Page test uses an ordered alternative hypothesis. Using the weights function in gls should work similarly if we had more reps. Besides the non-parametric Page test having more power given our number of replicates, the only other notable difference is that non-parametric tests with low numbers of replicates have been calculated to an exact probability whereas the parametric analyses have an approximate probability of < some value.
This is what I found using Google, with greater detail found using the link:
https://www.itl.nist.gov/div898/software/dataplot/refman1/auxillar/page.htmMar 8, 2013 — The Page test is a non-parametric test for analyzing randomized complete block designs. It is derived from a Spearman's rho rank correlation ...
L197 Please indicate what the error bars show. Since the weighing error of starting litter was already quite high (L 137)
In L 137, the 6 +/- 0.5 g does not indicate the error in weighing because we did not compare final weight to a standard initial weight of 6 g. We compared final net weight to the measured initial net weight of the leaves in each individually numbered bag. The variation is because whole leaves were used, and we needed whole leaves in a monolayer in each bag.
this is important information to display. Best would be to always show the actual data points.
L197: This was standard deviation of the mean for weight loss during preconditioning, from the caption for Figure 2. Another reviewer felt that the graph was not needed as the pattern and analysis were discussed in the text, so this figure was deleted and Figure 3 was renumbered as Figure 2. If the figure were to go back into the manuscript, the point is well taken that the data points should be shown.
Fig. 4 Please clarify that this is data without pre-conditioning. It would be actually also nice to see the non-significant data after 3 months, since this is an ecologically interesting result that there was no effect.
Response: This is now figure 3. The caption already says these are non-preconditioned leaves. We can provide the additional figure as supplemental online material.
