Science
Science is a systematic enterprise that builds and organizes knowledge in the form of testable explanations and predictions about the universe (Science 2012). Testable predictions are more commonly known as hypotheses. Through rigorous testing of scientific hypotheses science incrementally builds knowledge on the foundation of previously conducted research.
Here is a very simplistic example of how science proceeds through the process of hypothesis testing.
1. Previous Research
We know from previous research that trees convert carbon dioxide, water, and light energy to carbohydrates and oxygen by the process of photosynthesis. We also know that some of these carbohydrates are subsequently transported from the leaves where photosynthesis occurs through the phloem, or inner bark, to the roots. The roots must then break down the carbohydrates so that they can survive, grow, and function. Through the process of respiration roots convert carbohydrates and oxygen to water, carbon dioxide, and energy, essentially reversing the reaction of photosynthesis to utilize energy initially captured from the sun. Some of the carbon dioxide respired from roots is released into the soil, and some of this is gradually diffuses from areas of high concentration in the soil to areas of lower concentration in the atmosphere. The movement of carbon dioxide from the soil surface to the air occurs constantly in forested ecosystems, and is called soil CO2 efflux.
This brief paragraph summarizes the results of countless experiments performed by many different scientists over 100's of years that have rigorously tested, confirmed, and reaffirmed each step in this sequence of events.
2. Creating a New Hypothesis
Given only the background above, it may be unclear whether trees are able to store carbohydrates for any significant length of time. Obviously they must supply the roots with enough carbohydrates to last them through the dark night when photosynthesis cannot occur, but are they able to store more carbohydrates than this?
This question is a hypothesis. Often hypotheses are phrased in negative terms known as null hypotheses. Our null hypothesis is thus:
Trees are unable to store sufficient carbohydrates to support root respiration for longer than one day in the absence of photosynthesis.
3. Testing the Hypothesis: An Experiment
Once a null hypothesis has been developed, an experiment is designed to test that hypothesis. Experiments are methodical trial and error procedures carried out with the goal of supporting, falsifying, or establishing the validity of a hypothesis (Science 2012). Experimental designs vary widely depending upon the hypothesis being tested. A good experiment is designed with a strong understanding of the system being tested, so that all the possible outcomes are predicted and interpreted prior to beginning experimentation. If an experiment only has two possible outcomes, then one will refute the null hypothesis, and the other will support it.
A simple experiment can be designed to test the null hypothesis posited above. Trees can be grown in a growth chamber, where it is possible to control light. By turning the lights off for several days, photosynthesis cannot occur. It will then be possible to determine how much carbohydrate trees had stored, since we know they cannot make any more through photosynthesis.
There are several different variables that could be measured to test the null hypothesis once the lights are turned off. It is possible to directly measure soil CO2 efflux itself. This procedure conveniently does not require any part of the trees to be destructively harvested, which allows measurements to be repeated over time.
The experimental procedure is to turn off the lights and measure soil CO2 efflux over a period of several days. If it remains unchanged, then we will reject the null hypothesis because it will be clear that trees do have sufficient carbohydrates stored to support root respiration for longer than one day in the absence of photosynthesis. If, however, soil CO2 efflux declines, then it will lend support to the null hypothesis. Note that this is phrased as supporting, rather than proving the null hypothesis. Proving a hypothesis conclusively requires many experiments conducted by different scientists before results are widely accepted as fact and incorporated into our understanding of the system in question.
4. Analyzing the Results
The experiment described above was conducted in December 2008. Here are the results shown in graphical form:
These results seem to support our null hypothesis when analyzed. Efflux does go down when the lights are turned out. However, the trend is more complex than our simplistic hypothesis predicted. Efflux gradually declines, but does not cease entirely. This would seem to indicate that some carbohydrates were stored in the roots, but not enough to allow respiration to continue at the same rate it had when the lights were on and new carbohydrates were available every day. Further, we can see that rates do not immediately recover when the lights are turned back on. While our null hypothesis was supported, and we know that carbohydrates recently fixed through photosynthesis are needed to maintain root respiration rates, we have raised several new questions.
This example is a simplistic representation of the analysis step. Often statistical methods are employed to make coherent inferences from complex data in as objective a manner as possible.
5. Subsequent Hypotheses
Most good experiments create as many new questions as answers they provide. Science is a repetitive, iterative process. We would likely want to design a new experiment using what we learned to better assess what is occurring with carbohydrate storage in these loblolly pine seedlings. We might destructively harvest some of them at intervals, and actually chemically analyze the carbohydrate levels in various parts of the tree. We might try to examine microbial communities in the soil, to see if some of the CO2 is coming from the decomposition of fine roots that have died without sufficient carbohydrates to support them. This is where a prior understanding of the system in question is critical. Without this understanding results can be misinterpreted, resulting in incorrect conclusions.
Sometimes in the course of experimentation we may find that our experiments are flawed in some way. This may be due to something new we learn about the experimental system itself, or simply due to problems becoming apparent as an experiment proceeds. In this example, it is possible that these trees would have had declining efflux rates even with the lights on since the study was done in December, when they are normally slowing their physiological processes as a result of winter weather conditions in their natural environment. Upon repeating the experiment it would be a good idea to add a control treatment where the lights are not turned off, so that we can compare those rates to the rates without light and thus rule out the possibility that seasonal variability was the cause of the observed decline.
We not only learned something from our experiment about carbohydrates in trees, but we also now have a better idea about what we don't yet understand. The process repeats with a greater knowledge of our system, new null hypotheses, and further experiments to conduct and analyze.
The Role of Research in Silviculture
Silviculture has been the subject of extensive research in the United States since the early 1900's. Silvicultural research is uniquely challenging for a variety of reasons.
- Forest ecosystems are very complex and spatially and temporally variable.
- Forest ecosystems are subject to disturbances and environmental conditions we cannot control.
- Trees are large organisms that can be difficult to manipulate, measure, and sample.
- Many relevant ecological processes occur below-ground or high in the canopy, making them difficult to quantify.
- Trees are long-lived organisms, requiring many years of data following some experiments.
- Research funding is often short-term, lasting only a few years.
- Land ownership changes over time, making it challenging to maintain long-term experiments.
Many researchers have spent their careers developing what we now know about silviculture in the face of these challenges. Much of what has been learned has already been incorporated into modern silvicultural practices. Here are but a few key examples of how research has affected operational silvicultural practices that will be covered further throughout this textbook.
- Tools for site selection allow the right species to be managed on the sites they are best adapted to.
- Tree breeding has dramatically improved growth rates, stem form, and disease resistance of forest plantations.
- Site preparation techniques improve access, increase survival and early growth, and ameliorate degraded sites.
- Density metrics, guides, and diagrams allow for informed management of stands to meet specific product objectives by:
- guiding planting density at establishment,
- allowing for planning of thinnings to best utilize site resources,
- minimizing trees lost to mortality due to competition.
- Forest herbicides selectively control non-crop woody and herbaceous species.
- Fertilizer prescriptions improve forest productivity and make otherwise unproductive sites suitable for forestry.
- Pruning techniques create clear wood and rapidly produce high-value sawlogs.
- Regeneration methods balance timber production with structural retention that improves ecosystem function.
- Best management practices protect water quality and ensure sustainable soil management.
- Silvicultural systems now exist that enable management of all major forest cover types of the United States to meet a range of landowner objectives.
Despite the progress that has been made, there remain many challenges that require further research. Climate change, invasive species, novel damaging insects and diseases, increasing demand for forest products from a declining land base, forest fragmentation, an increasing urban-wildland interface, and development of bioenergy markets all raise many questions as to how our forests may be best managed in the future.
Art
Art is the exercise of highly-proficient human skill honed by practice to create representations of the natural world (Art 2012). Despite the extensive research that informs silvicultural practices, much remains unknown. Where information is lacking, the silviculturist must rely on his or her own experience and understanding of the ecology of the forest to prescribe treatments that meet management objectives. Even if information is available in the literature, there are often multiple ways they may be operationally implemented. The art is in knowing the best and most efficient process to produce the desired changes in stand structure.
Forest ecosystems, forest products markets, and societal factors that influence silvicultural decision making are all complex and multifaceted. Uncertainty abounds, and future risks to the success of silvicultural prescriptions are often unknown at the time they are first created. For example, pathogens that may cause damage may not yet be known or introduced. Forests grow slowly compared to agricultural crops, and thus today's silvicultural decisions may have consequences for decades or even centuries. Forestry is a long-term endeavor, and with longer time periods comes greater uncertainty. Despite the challenges presented by the very nature of forestry, we must still implement treatments that alter stand structures and functions to better meet management objectives.
Whatever treatment is prescribed and implemented today will limit the selection of future treatments in a stand (Smith et al. 1997). Skilled silviculturists must use caution when attempting new techniques so that they do not unintentionally limit the potential management actions that can be applied to a stand long after their careers have concluded. This is one reason that silviculture still tends to progress through minor modifications to the suite of techniques that have been gradually developed and implemented over centuries of forest management in Europe and then North America. Creativity may be better applied through the modification of existing tools to address a particular problem rather than through the wholesale development of new methodologies.
The Intersection of Art and Science
Conditionality
Conditionality is the principle that the probability of an event or result occurring depends, or is conditioned, upon a variety of causal factors. For example, the growth rate of a forest is conditional upon the amount of available nutrients and water on the site, the quantity of incident radiation (light) reaching the canopy, and the length of time between disturbances. All these factors vary at different spatial scales. Site quality varies within a region, while light varies between regions with different latitudes.
Further, species vary in their ecology, making them more or less capable of out-competing others and growing rapidly on a given site or in response to silvicultural treatments. For example, loblolly pine often does not grow any faster than slash pine when silvicultural inputs such as competition control and fertilizer application are limited on an average quality site. However, when silvicultural inputs are high, loblolly pine may grow at much faster rates than slash (Roth et al. 2007). If asked which grows faster, loblolly or slash pine, there is not a single and simple correct answer. Rather it depends on the conditions they are grown under.
Conditionality, more than any other factor, can create confusion for students studying silviculture or other ecologically-based disciplines. It is common to find two published research papers on the same topic that seem to produce contradictory results. However, the apparent conflict is often attributable to differences in important conditions that affect the results. What may be true in the boreal forest may not apply at all in bottomland hardwood stands of the South.
Because silviculturists work with many different species, ecosystems, management objectives, product markets, and societal constraints, it may at times be impossible to broadly generalize the stand dynamics or response to silvicultural treatments of stands that possess markedly divergent conditions. Thus, when trying to find answers to a silvicultural problem, it is always best to use research that is as similar as possible to the stand and other factors in question. Where such research is lacking, the forester must make an educated guess as to how their forest will respond to treatments based on experience and available research from other forest types. Thus art and science are blended by applying what science is available while simultaneously relying on proficient skill honed through practice to guide action where information is lacking. This is the inevitable product of the complexity and conditionality of forests and the social and economic factors involved in their management.
Anecdotal versus Empirical Evidence
When balancing the art and science of silviculture, it is often tempting to ascribe more weight to anecdotal evidence. Anecdotal evidence is obtained through personal experiences or accounts of observations from trusted colleagues. Empirical evidence, by contrast, is developed through the process of experimentation described previously.
Empirical evidence is produced through the collection and analysis of data. It is often the best foundation for inference provided the research was of high quality and the conditions of the experiment were sufficiently similar to the stand being managed. The peer-review process used for most published literature, while not perfect, is the best mechanism available for ensuring research is of high quality. Information found in methods sections can be used to assess the important conditions of the research such as species, geographic location, soil types, and stand age, structure, and function.
Over the course of a career, silviculturists develop a substantial bank of anecdotal evidence by examining the effects of their and others' prescribed management actions on individual stands. They also accumulate empirical evidence through continued review of the literature.
When anecdotal evidence and empirical evidence agree, it is easy to confidently prescribe treatments. However, at times anecdotal evidence and empirical evidence may not align, causing uncertainty. It is then important to consider sample size and conditionality when determining which information to rely upon more.
For example, the literature shows that the practice of windrowing, or piling all the slash and debris on a site following clearcutting into long mounds, has the potential to lower the next stand's productivity and uniformity by moving the organic matter, topsoil, and nutrients to the windrows (Morris et al. 1983, Tew et al. 1986, Fox et al. 1989).
Based in part on the results of these experiments, most large forest land management corporations in the US South no longer windrow as a site preparation treatment. However, individual foresters working for these companies sometimes report that some of the best stands they have ever seen have been windrowed. If the site prep operator is careful not to disturb the mineral soil, windrowing can allow for easy access that results in a high quality planting operation.
These two sources of evidence (anecdotal and empirical) are in this case both correct, and yet they obviously contradict one another. This seemingly intractable discrepancy can be resolved by considering sample size and conditionality to determine which form of evidence should be given greater weight.
The studies listed above each took place on 3 different individual sites in the North Carolina Piedmont and Florida Flatwoods. If site is treated as our sampling unit, then this reflects a sample size of three sites across all the studies. The conditions of the studies include species (loblolly or slash-longleaf mixture), soils, and silvicultural treatments. These are relatively similar to the conditions for windrowed pine plantations across much of the US South, and are thus a sound basis for inference within this population.
So one site, or however many were observed as part of the anecdotal evidence, that grew well following windrowing must be compared with these three sites with documented effects of windrowing. In this consideration the nature of the anecdotal observations are also important. The literature is based on detailed measurements. If the anecdotal evidence is derived from simply driving past the stand, it would not hold as much weight as if the stands in question were cruised for a timber sale.
After considering the conditions in the literature (similar in this case), the sample size (three sites versus one), and the nature of the anecdotal evidence, the forester must decide which source of information to use for future management decisions. In this example the empirical evidence likely has more weight on an average site in the US South. Also, the convenience of windrowing must be balanced with the known potential for degrading the site. Given the potential severity of the consequences of windrowing, many foresters would decide to prescribe a different treatment to manage slash on the site.
In the example above the studies cited represented a small number of sites. However, there are some silvicultural experiments published in the literature that may span many sites. The weight of empirical evidence of such studies is very high compared with anecdotal evidence. Anecdotal evidence based on a single site may reflect what is essentially an outlier, or extreme example, compared with the broader population of sites that are better sampled in a large study.
This example makes more clear that the balancing of anecdotal and empirical evidence requires much consideration. However, it is a skill that can be improved through practice, field observation, and reading of the literature. Successfully integrating the art and science of silviculture is one of the discipline's greatest challenges, yet working towards this goal offers many rewards.
References
Art. (n.d.). In Collins English Dictionary - Complete & Unabridged 10th Edition. Retrieved August 15, 2012, from http://dictionary.reference.com/browse/art
Experiment. (n.d.). In Wikipedia. Retrieved August 15, 2012, from http://en.wikipedia.org/wiki/Experiment
Fox, T. R., L. A. Morris, and R. A. Maimone. 1989. The impact of windrowing on the productivity of a rotation age loblolly pine plantation. Pages 133-140 in Proceedings of the fifth biennial southern silviculture research conference. General Technical Report SO-74. New Orleans, LA: USDA Forest Service, Southern Forest Experiment Station, Memphis, TN. http://www.treesearch.fs.fed.us/pubs/1749
Morris, L. A., W. L. Pritchett, and B. F. Swindel. 1983. Displacement of nutrients into windrows during site preparation of a flatwood forest. Soil Sci. Soc. Am. J. 47:591-594. http://dx.doi.org/10.2136/sssaj1983.03615995004700030040x
Roth, B. E., E. J. Jokela, T. A. Martin, D. A. Huber, and T. L. White. 2007. Genotype x environment interactions in selected loblolly and slash pine plantations in the southeastern United States. Forest Ecology and Management 238:175-188. http://dx.doi.org/10.1016/j.foreco.2006.10.010
Science. (n.d.). In Wikipedia. Retrieved August 15, 2012, from http://en.wikipedia.org/wiki/Science
Smith, D. M., B. C. Larson, M. J. Kelty, and P. M. S. Ashton. 1997. The Practice of Silviculture: Applied Forest Ecology. Ninth edition. John Wiley & Sons, Inc., New York, New York. ISBN: 047110941X.
Tew, D. T., L. A. Morris, H. Lee Allen, and C. G. Wells. 1986. Estimates of nutrient removal, displacement and loss resulting from harvest and site preparation of a Pinus taeda plantation in the Piedmont of North Carolina. Forest Ecology and Management 15:257-267. http://dx.doi.org/10.1016/0378-1127(86)90163-5