Parsing Long-Term Tree Recruitment, Growth, and Mortality to Identify Hurricane Effects on Structural and Compositional Change in a Tropical Forest

Round 1

Reviewer 1 Report

Natural disturbance events, such as hurricanes, damage trees and open up the canopy, and can lead to drastic reorganization of forest communities. Tree species differ in their response to disturbance both at the individual level (e.g., immediate damage and mortality) and population scales (e.g., enhanced fecundity, recruitment and tree growth in the high light conditions in the wake of a storm) so a key determinant of forest change are species differential demographic responses to hurricanes. In this manuscript, the authors aim to parse out the contribution of tree recruitment, growth and mortality on structural and compositional change in a tropical forest in Puerto Rico that has been exposed to two severe hurricanes, hurricane Hugo (1989) and H. Georges (1998). To do so, they rely on tree demographic data collected over 25 years at uneven census intervals (~5 yrs) from 85(95) 10-m diameter plots distributed across a 13-ha area that spans variation in topography.  

 

The paper addresses an important question and relies on a great dataset. However, it suffers from several conceptual, structural, and methodological problems that limit its scope and relevance.

 

Conceptual and structural issues

1. The introduction is short and offers no clear conceptual framework or hypotheses for expected changes. Hurricanes restart the succession process and favor pioneers and wind-resistant species. The authors have a long record of research in the area, including information on successional status and vulnerability to wind damage for all the tree species in the study. What hypotheses can they offer in terms of expected community change post-hurricanes?

 

2. The authors could also offer some hypotheses based on variation in demographic rates among topographic positions. Overall, the inclusion of topography is not motivated in the introduction. The effects of hurricanes on landscapes are complex and depend on topographic exposure, soils and species composition. The authors have information on all of these factors. What are their expectations with respect to damage and post-hurricane community trajectories for each of the sites? For each process (e.g,, survival, recruitment)? It is important to note that topography also influences species distributions. Are particular groups of species (e.g., pioneers, wind resistant species) more abundant at some of the sites?

 

3. The introduction also meanders into data error and filling issues that are best addressed in the methods section, or even in supplementary information.

 

4. At a more fundamental level the paper does not do what it says it will do. Reporting rates of recruitment, growth and mortality is not the same as assessing the contribution of these rates to changes in composition and structure. See Rozendaal & Chazdon (2015) for a very nice example of how to do this for say, basal area. The authors may also want to consider an Integral Projection Model to evaluate the effects of various demographic rates on population growth.

 

 

Materials and methods

1. The methods are incomplete in many places. For instance, the plots were initially sampled three months after H. Hugo and the post-hurricane census was started three months after the hurricane. The authors do not state when the post-hurricane census was completed or when the subsequent censuses were completed; Fig 4 suggests that the plots were re-censused in 1994, 1999, 20004, 2009 and 2014. This needs to be in the text not tucked away in a figure.

 

2. The rationale for grouping species into PFTs is unclear. It’s fine for a model but this is not a modeling paper and species responses to hurricanes are quite distinct. Is it a sample size issue? From my perspective, it does not make much sense to group trees, shrubs, and ferns; they are completely different from a demographic perspective. Most shrubs are multi-stemmed and respond to hurricanes in a different way than trees.

 

3. The diameter filling algorithm does not make much sense. Tree growth, mortality, and recruitment after a hurricane depend not only on the size and taxonomic identity a stem suffered but also on its competitive environment, which can be quite variable in space (Uriarte et al. 2004). It is also not clear how damage was considered in diameter filling. Data is missing for less than 5% of stems, I suggest leaving this out. It does not seem necessary, rests on weak logic, and detracts from the message of the paper.

 

4. Same is true for assessing diameter measurement error. One of the key drivers of the degree of shrinkage a stem suffers is how much damage it sustained. A large negative measurement usually indicates that a tree is moribund. How was this consideration included?

 

5. Diameter growth and probability of survival are not linear with respect to size. Growth usually levels off and mortality follows a u-shaped distribution (there is a huge literature on this). It’s fine to calculate RGR and mortality at the population level but making statement about the relationship between size and demography requires a different approach; these are individual level processes not population level ones.

 

6. What was the age of the forest? There are good forest age maps for the area. If the forest is an old secondary forest as the authors state, why does it have such a disproportionately high density of small stems pre-hurricane in 1989? This is particularly evident when comparing pre-hurricane and 2014 size distributions (Fig. 7), with a greater number of large stems and smaller number of small stems in 2014 than in 1989 pre-hurricane. This has at least three potential explanations: 1. the pre-hurricane forest had been logged and a large proportion of large trees had been removed relatively recently; 2. the forest is young but this is not consistent with the low overall stem density, and 3. The size distribution recorded pre-Hugo was not accurate. It would be good to know which one of these potential explanations is correct. It would useful to compare stem densities to those reported in other parts of the LEF pre- and post-hurricane to get a sense of potential drivers.

 

7. Palms grow in height more than diameter.

 

 

Discussion

1. The discussion is underdeveloped and does not consider the substantial amount of research that has been conducted on this topic in El Yunque and other cyclone prone areas. The authors should place their results in a broader context: what is the novelty of their study? Does it match results from other studies or not? I have added several references below.

 

2. Palm mortality of ~42% is hardly low. Most studies at the study site find much lower rates of palm mortality, even in the far more severe H. Maria. The authors suggest that this is because small trees suffer the brunt of the damage. This is not at all consistent with the biomechanics of wind tree damage observed in many studies in Puerto Rico and elsewhere. One potential explanation for this variation is small-scale wind blowdown during H. Hugo which may have resulted in large trees falling on smaller trees. Is there a spatial pattern to the distribution of high mortality? Such information would go a long way towards understanding the mechanism behind such extremely high rates of mortality.

 

3. Similarly, are increases in growth similar to what was observed after hurricanes elsewhere in the LEF? Recruitment?

 

4. Mortality increases after a disturbance for two reasons: 1. Delayed effects of damage and 2. As the canopy closes many of the new recruits die. The data show strong evidence for the latter but this is hardly discussed. There is also a substantial amount of work on this subject at the site which the authors should consider.

 

5. Diameter growth is not linear with respect to size and is also influenced by neighborhood crowding (i.e., light conditions) and strong asymmetries in competition (i.e., the effect of a large tree on a small one is far stronger than the other way around). It’s not surprising that a linear model does not capture the patterns.

 

6. Was the variation in demographic rates across demographic positions significant (Fig. 6)?

 

 

References

 

Beard, K. H., Vogt, D., Vogt, D., Scantena, F. N., Covich, A., Sigurdardottir, R., . . . Crowl, A. T. (2005). Structural and functional responses of a subtropical forest to 10 years of hurricanes and droughts. Ecological Monographs, 75(3), 345-361.

Canham, C. D., Thompson, J., Zimmerman, J. K., & Uriarte, M. (2010). Variation in susceptibility to hurricane damage as a function of storm severity in Puerto Rico tree species. Biotropica, 42(1), 87-94. doi:10.1111/j.1744-7429.2009.00545.x

Comita, L. S., Thompson, J., Uriarte, M., Jonckheere, I., Canham, C. D., & Zimmerman, J. K. (2010). Interactive effects of land use history and natural disturbance on seedling dynamics in a subtropical forest. Ecological Applications, 20(5), 1270-1284.

Comita, L. S., Uriarte, M., Thompson, J., Jonckheere, I., Canham, C. D., & Zimmerman, J. K. (2009). Abiotic and biotic drivers of seedling survival in a hurricane-impacted tropical forest. Journal of Ecology, 97, 1346–1359. doi:10.1111/j.1365-2745.2009.01551.x

Frangi, J. L., & Lugo, A. E. (1991). Hurricane damage to a flood plain forest in the Luquillo Mountains of Puerto Rico. Biotropica, 23, 324-335.

Frangi, J. L., & Lugo, A. E. (1998). A flood plain palm forest in the Luquillo Mountains of Puerto Rico five years after Hurricane Hugo. Biotropica, 30(3), 339-348. doi:10.1111/j.1744-7429.1998.tb00069.x

Imbert, D., Labbe, P., & Rousteau, A. (1996). Hurrican Damage and Forest Structure in Guadeloupe, French West Indies. Journal of Tropical Ecology, 12(5), 663-680.

Negron-Juarez, R., Baker, D. B., Zeng, H. C., Henkel, T. K., & Chambers, J. Q. (2010). Assessing hurricane-induced tree mortality in U.S. Gulf Coast forest ecosystems. Journal of Geophysical Research-Biogeosciences, 115. doi:10.1029/2009jg001221

Putz, F. E., & Sharitz, R. R. (1991). Hurricane damage to old-growth forest in Congaree swamp National Monument, South Carolina, USA. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere, 21(12), 1765-1770. doi:10.1139/x91-244

Rozendaal, D. M. A., & Chazdon, R. L. (2015). Demographic drivers of tree biomass change during secondary succession in northeastern Costa Rica. Ecological Applications, 25(2), 506-516. doi:10.1890/14-0054.1

Uriarte, M., Canham, C. D., Thompson, J., & Zimmerman, J. K. (2004). A neighborhood analysis of tree growth and survival in a hurricane-driven tropical forest. Ecological Monographs, 74(4), 591-614. doi:10.1890/03-4031

Uriarte, M., Canham, C. D., Thompson, J., Zimmerman, J. K., Murphy, S. F., Sabat, A. M., . . . Haines, B. L. (2009). Natural disturbance and human land use as determinants of tropical forest dynamics: results from a forest simulator. Ecological Monographs, 79(3), 423-443.

Uriarte, M., Clark, J. S., Zimmerman, J. K., Comita, L. S., Forero-Montana, J., & Thompson, J. (2012). Multidimensional trade-offs in species responses to disturbance: implications for diversity in a subtropical forest. Ecology, 93(1), 191-205. doi:10.1890/10-2422.1

Zimmerman, J. K., Comita, L. S., Uriarte, M., Brokaw, N. V. L., & Thompson, J. (2010). Patch dynamics and community metastability of a tropical forest: Compound effects of natural disturbance and human land use. Landscape Ecology, 25, 1099-1111.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This article is originated from the doctoral thesis (2021, Georgia Institute of Technology) of the first author, the co-authors are researchers and part of the thesis committee. This doctoral thesis is not cited in the references of this article (I suggest including it since it is an excellent reference).

This article is very well written, clear, relevant, has a good data curation, the statistical and graphical analysis is solid. Therefore, the evidence supports good conclusions.

The following suggestions are only for your consideration. The article, as presented is fine, but the scope of the introduction and discussion could optionally be reinforced.

Introduction:
It is worth framing this article on the topic of global warming, phenomenum that is increasing the frequency and intensity of hurricanes in the Caribbean. Some quotes supporting this trend and how this can affect the ecology of the Caribbean forests could be of help. Perhaps there is data to illustrate the  the extend of land and forests impacted by hurricanes in the Caribbean, or at least indicate the effect of some hurricanes. The idea is to demonstrate that this study allows us to appreciate the resilience and capacity of forest ecosystems to recover their structure and species composition over time, but the question is, how sustainable these natural processes are if the frequency and intensity of hurricanes increases?.

Conclusions.
It should be recognized that the article employs a legacy experimental design. This design could be improved for future studies, what do you suggest?
Given the increase in frequency and intensity of hurricanes, what do you anticipate for the Caribbean forests (e.g., carbon gains or losses, the species composition will be less diverse) and what lines of research can you suggest?

Methodology.

It is no clear why the extend of the study area in the original experimental design is only 13 hectares.  I anticipate that the impacted area by Huracan Hugo was much greater.  Is the selected area the most impacted area by Hugo, or why these 13 has were selected?  

Do you consider a plot of 78.5 m2 adequate for this analysis? what would be the advantages o disavantages of using less bout greater plots, example 20x50 mts plots, 1000 m2?

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx