Secondary Succession after Slash-and-Burn Cultivation in Papuan Lowland Forest, Indonesia

Round 1

Reviewer 1 Report

Compared to other parts of the tropical world the is little information on secondary succession in southeast Asia. Data based papers like this one there for are welcome.

The study evaluates plant richness, diversity and densities, and soil conditions, in 5 secondary forest (SF) locations and 1 primary forest  (PF) location. In each of these locations 5 (SF) - 6 (PF) plots are installed. Plot size is quite small size, and increasingly smaller for the smaller size class plants. Additionally species are compared to lists of endemics, conservation value and invasives. Results are discussed in terms of how they fit into a swidden agriculture system, and in terms of conservation value. Plots have remnant trees and all kinds of planted species.

Although I like the topic and applaud that this has been done I have a number of issues with the manuscript

1. The plots are very small. For instance in a secondary forest the seedlings are measure in 2x2 m per plot, giving a total of 5 times 4 = 20m2 of seedlings per secondary forest location, and 24m2 for the primary forests location.
2. The plots are clearly pseudoreplicated, as the figure There are in total 6 locations and in every location (which I assume of one forest location with the same history and age), there are 5-6 small plots. This is fine to get a good view on these locations but they cannot be used as replicates in an analysis. This means that some figures are okay (fig2), while others are not (fig4, fig 5). At least not if you would like to generalize the results.
3. Many plots have remnant trees, and planted crops of various kinds. This is unclear where it is and how it impacts the results. In such small plots the impact of an individual remnant tree can be very large.
4. The units used and categories used are not clear in several figures (fig 2, fig 3, fig 4. Please adjust and include this in the figure headings.
5. Area of fig 4 is not correct. The mean is not per 0.3 or 0.36 ha. The mean value can be correct (for the whole) but then the error bars should not be there.
6. all SF are very young. You could then use these four SF as replicates. Then you could test is PF would fall into or outside the category of the SF and that would be a good test. Of course you would have to correct for plot size differences.
7. Fig 3 is using all the small and individual plots (maybe just the species presence? and including also all the smaller categories?) and because of the small size an remnant tree or planted tree will have a large impact and make that the small plots will be different.  In which of these this is the case?  
8. Soils are also pseudoreplicated, so be careful with interpretations other than that one location (one forest) is different from another one.
9. The pseudoreplication is mentioned somewhere in the discussion, but only slightly and between brackets. I think this is not correct as this is really a major issue in this study. As long as you describe the plots all is fine, but testing making it different.
10. The discussion has a lot of real life comments on the forests and the local situation and that by itself is interesting. However, a bit more critical review of your own data results may be justified.
11. I also see that nearly all studies are from the region. If in the intro you mention that there is so much out there for other regions, maybe you could add a bit more discussion here of your or your region results compared to other regions.

 

Qualitatively I like the study, but quantitatively some improvements would be needed (see comments above).  I like your plea for including SF into conservation schemes and programs.

 

Author Response

Reviewer 1

Compared to other parts of the tropical world the is little information on secondary succession in Southeast Asia. Data based papers like this one therefore are welcome.

RESPONSE: Thank you for appreciating the value of our study.

The study evaluates plant richness, diversity and densities, and soil conditions, in 5 secondary forest (SF) locations and 1 primary forest  (PF) location. In each of these locations 5 (SF) - 6 (PF) plots are installed. Plot size is quite small size, and increasingly smaller for the smaller size class plants. Additionally species are compared to lists of endemics, conservation value and invasives. Results are discussed in terms of how they fit into a swidden agriculture system, and in terms of conservation value. Plots have remnant trees and all kinds of planted species.

Although I like the topic and applaud that this has been done, I have a number of issues with the manuscript

1. The plots are very small. For instance in a secondary forest the seedlings are measured in 2x2 m per plot, giving a total of 5 times 4 = 20m² of seedlings per secondary forest location, and 24m² for the primary forests location.

 

RESPONSE: The reviewer is correct that we sampled progressively smaller plots for smaller life-forms. Whilst the total area covered is small, we consider that our plots have value in being a much more extensive analysis of forest floristics than the majority of tropical forest studies that focus only on trees (usually ≥ 10 cm dbh).

 

1. The plots are clearly pseudoreplicated, as the figure. There are in total 6 locations and in every location (which I assume of one forest location with the same history and age), there are 5-6 small plots. This is fine to get a good view on these locations, but they cannot be used as replicates in an analysis. This means that some figures are okay (fig2), while others are not (fig4, fig 5). At least not if you would like to generalize the results.

 

RESPONSE: Thank you for your comments on this important point. In order to attend to this, Figs. 2 and 3 remain unchanged. In response to Reviewer 2’s comments, Fig 4 has now been removed. As Fig. 5 shows means for each of the forest ages, we consider it to be suitable and prefer to retain the table in order to facilitate comparisons between SF and PF as noted in in our response to point 6 below.

 

1. Many plots have remnant trees, and planted crops of various kinds. This is unclear where it is and how it impacts the results. In such small plots the impact of an individual remnant tree can be very large.

 

RESPONSE: Thank you for your comments on this important point. It can be seen from Fig. 4 that the contribution of large trees to the total density is very low. Generally, large trees (≥ 20 cm dbh) increased in density with fallow age, however, the fact that 4-yr-old forests are not different from any other SF in terms of large trees shows that the presence of large trees in this forest age is more variable than the others. There appears to be no clear pattern in terms of the presence of crop/cultivated species and location seems to be a more important factor than fallow age (Appendix). We have now also added a photo of the forests to help with interpretation (Fig. 2).

 

1. The units used and categories used are not clear in several figures (fig 2, fig 3, fig 4. Please adjust and include this in the figure headings.

 

RESPONSE: We have altered the Figure headings accordingly to increase the clarity of what the Figures are showing. Particularly Fig. 4 where we included that densities were multiplied accordingly so that the same area per lifeform is included – as this is an important point, we also included it in the Methods. Please inform us is further clarification is needed.

 

1. Area of fig 4 is not correct. The mean is not per 0.3 or 0.36 ha. The mean value can be correct (for the whole) but then the error bars should not be there.

 

RESPONSE: Thanks for noting these errors, However, In response to Reviewer 2’s comments, Fig. 4 has now been removed.

 

1. All SF are very young. You could then use these four SF as replicates. Then you could test is PF would fall into or outside the category of the SF and that would be a good test. Of course you would have to correct for plot size differences.

 

RESPONSE: We acknowledge the generally young age of our SF fallows and have added a sentence to the Discussion in this regard. In terms of comparison between PF and SF, we consider that the patterns show clear differences between these two classes in Fig. 3 (species accumulation curves), Fig. 4 (dendrogram) and Fig. 5 when looking at the table of Tukey test results.

 

1. Fig 3 is using all the small and individual plots (maybe just the species presence? and including also all the smaller categories?) and because of the small size and remnant tree or planted tree will have a large impact and make that the small plots will be different. In which of these this is the case?  

 

REPONSE: We have answered this under point 4 above.

 

1. Soils are also pseudoreplicated, so be careful with interpretations other than that one location (one forest) is different from another one.

 

RESPONSE: We acknowledge this useful point and have tried not to over-interpret our data.

 

1. The pseudoreplication is mentioned somewhere in the discussion, but only slightly and between brackets. I think this is not correct as this is really a major issue in this study. As long as you describe the plots all is fine, but testing making it different.

 

RESPONSE: Thank you for your clear recommendation.  We have now included a penultimate paragraph in the Discussion where we note how our study could be improved including the concern of pseudo-replication and that we need to overcome this in future studies.

 

1. The discussion has a lot of real-life comments on the forests and the local situation and that by itself is interesting. However, a bit more critical review of your own data results may be justified.

 

RESPONSE: As in our response above, we have now included a penultimate paragraph in the Discussion where we note how our study could be improved including the concerns of pseudo-replication, small plots and young fallow ages.

 

1. I also see that nearly all studies are from the region. If in the intro you mention that there is so much out there for other regions, maybe you could add a bit more discussion here of your or your region results compared to other regions.

RESPONSE: We now include a paragraph comparing our results with those from other parts of New Guinea and also extending this, briefly, to other meta-analyses.

Reviewer 2 Report

The article “Secondary Succession after Slash-and-Burn Cultivation in Pa- 2 puan Lowland Forest, Indonesia” compares species richness and diversity between fallows and primary forests. These comparisons included different life forms such as trees, herbs, ferns, shrubs and lianas, and relatively young fallows (2-9 yrs old). I suggest that the authors consider the following commentaries:

1. Many studies compare primary vs. secondary forests, and consider that primary forests are really primary. However, since most of the forests of the world were disturbed at least once, these forests are really primary or are older secondary forests? What are the conditions or characteristics that these forests have to say they are primary? Older and bigger trees, high diversity, in inaccessible areas or unsuitable areas for agriculture? Maybe it would be useful in methods to say that these forests remained relatively undisturbed for decades, if this is the case.
2. In the Intro, the authors say that “local people return to these re-growth forests to implement the next farming cycle…” Do you have information that the sites selected for this study were used more than once by indigenous people? This is important because this information can affect the results. In this case, density and diversity of plants may be influenced only to the fallows age but also to the number of times a forest was slashed and burned.
3. Even though fallows were compared with primary forests, they will not necessary reach “primary structure” or a similar plant composition during succession. There are alternative pathways in succession due to the influence of different factors (for example, previous agriculture practices, species present in surrounding forests, climate and edaphic conditions, etc.). Additionally, primary forests not necessary represent a climax phase, since its structure may be influenced by the intensity and frequency of natural and anthropogenic disturbances. Thus, I suggest the authors to discuss these aspects in the Discussion.
4. In the last paragraph of the Intro, there are good predictions but explanations of the mechanisms that lead to those predictions are necessary. It was predicted that the density and species richness will increase during succession, but what are the processes involved? Plant facilitation, more time for plant recovery, competition or other species interactions?
5. Finally, I found the diversity analyses were redundant. For example, you can obtain relatively similar results in rarefaction and Shannon, but rarefaction is a more powerful analysis.

 

 

Author Response

Reviewer 2

The article compares species richness and diversity between fallows and primary forests. These comparisons included different life forms such as trees, herbs, ferns, shrubs and lianas, and relatively young fallows (2-9 yrs old). I suggest that the authors consider the following commentaries:

1. Many studies compare primary vs. secondary forests, and consider that primary forests are really primary. However, since most of the forests of the world were disturbed at least once, these forests are really primary or are older secondary forests? What are the conditions or characteristics that these forests have to say they are primary? Older and bigger trees, high diversity, in inaccessible areas or unsuitable areas for agriculture? Maybe it would be useful in methods to say that these forests remained relatively undisturbed for decades, if this is the case.

 

RESPONSE: We have added the following text to clarify the useful points raised by the reviewer: “These [secondary forests] were compared with forest that had been relatively undisturbed for decades due to its inaccessibility with no reports of ever being cultivation as it was in an inappropriate location; we hence call this primary forest.”

 

1. In the Intro, the authors say that “local people return to these re-growth forests to implement the next farming cycle…” Do you have information that the sites selected for this study were used more than once by indigenous people? This is important because this information can affect the results. In this case, density and diversity of plants may be influenced only to the fallows age but also to the number of times a forest was slashed and burned.

 

RESPONSE: According to local informants, these fallow forests were only used once for cultivation. We have clarified this in the following text: “All secondary forests were in areas that had only been cultivated once (according to local informants).”

 

1. Even though fallows were compared with primary forests, they will not necessary reach “primary structure” or a similar plant composition during succession. There are alternative pathways in succession due to the influence of different factors (for example, previous agriculture practices, species present in surrounding forests, climate and edaphic conditions, etc.). Additionally, primary forests do not necessary represent a climax phase, since its structure may be influenced by the intensity and frequency of natural and anthropogenic disturbances. Thus, I suggest the authors to discuss these aspects in the Discussion.

 

REPONSE: Thank you for these useful points. We have added a couple of sentences to the Discussion with an appropriate reference that read: “It is also worth considering that there are alternative pathways in succession due to the influence of factors such as previous agriculture practices, the species present in sur-rounding forests, and edaphic conditions [47]. It is also important to study a large number of primary forest stands due to high heterogeneity in structure and diversity influenced, at least partly, by the intensity and frequency of disturbances.”

 

1. In the last paragraph of the Intro, there are good predictions but explanations of the mechanisms that lead to those predictions are necessary. It was predicted that the density and species richness will increase during succession, but what are the processes involved? Plant facilitation, more time for plant recovery, competition or other species interactions?

 

RESPONSE: We have now added that this is hypothesised to be “due to the greater amount of time for accumulation of species” and that the increase in soil fertility will be due to “nutrients are returned to the soil from regrowing vegetation.”

 

1. Finally, I found the diversity analyses were redundant. For example, you can obtain relatively similar results in rarefaction and Shannon, but rarefaction is a more powerful analysis.

 

REPONSE: We have now removed the Shannon diversity analysis (Fig. 4) as it is considered redundant by the Reviewer.

 

Reviewer 3 Report

The authors’ manuscript “Secondary succession after slash-and burn cultivation in Papuan lowland forest, Indonesia” clarified the change of community structure, composition, and plant diversity along secondary succession. It is important to understand the community changes and soil fertility during forest succession, which can provide positive suggestions for local forest managers and biodiversity conservation. However, this manuscript is still existing some problems. It needs a minor revision before publication.

line 17. “vegetation richness” should be “plant species richness”.

line 20. I want to know the vegetation types of primary forest in your manuscript, such tropical rainforest or tropical lower mountain evergreen forest.

line 67. in this part, the authors need introduce more details about primary forest, such as vegetation, species composition.

line 81. I has a confusing about plot size in your study, such as same size between primary forest and fallow forests.

line 93. “as noted above”, and I don’t find above life form classification, such as lianas, ferns,and herbs. Please add this content.

line 138-141.I don’t understand 0.3Ha. The authors mean that fallow forests and primary forest have different sampling plot size.

Author Response

Reviewer 3

The authors’ manuscript “Secondary succession after slash-and burn cultivation in Papuan lowland forest, Indonesia” clarified the change of community structure, composition, and plant diversity along secondary succession. It is important to understand the community changes and soil fertility during forest succession, which can provide positive suggestions for local forest managers and biodiversity conservation. However, this manuscript is still existing some problems. It needs a minor revision before publication.

line 17. “vegetation richness” should be “plant species richness”.

RESPONSE: Changed as requested (and also at four other points in the manuscript for consistency).

line 20. I want to know the vegetation types of primary forest in your manuscript, such tropical rainforest or tropical lower mountain evergreen forest.

RESPONSE: In line 20, we have already written ‘lowland tropical forest’ but we have changed to ‘lowland evergreen tropical forest’ for clarity, and also changed this in the ‘Study area’ section.

line 67. in this part, the authors need introduce more details about primary forest, such as vegetation, species composition.

RESPONSE: We have referred readers to other studies by the team that outline typical species composition for primary forests in this region which we hope is suitable.

line 81. I has a confusing about plot size in your study, such as same size between primary forest and fallow forests.

RESPONSE: Plot size is identical between forests – but the number of replicates differs. We have added the text “In all forests,…” at the beginning of the relevant sentence to clarify this point.

line 93. “as noted above”, and I don’t find above life form classification, such as lianas, ferns, and herbs. Please add this content.

RESPONSE: The classification according to lifeform has now been noted before the sampling scheme to clarify, and the sentence referred to in this comment has now been removed.

line 138-141.I don’t understand 0.3 ha. The authors mean that fallow forests and primary forest have different sampling plot size.

RESPONSE: In response to Reviewer 2’s comments, Fig 4 (and its legend) has now been removed.