BOTANY VOLUME 4 - ECOLOGY - 2007

12. FUNDAMENTALS OF PLANT ECOLOGY

12.4. Biological Variability

Ever since scientific biological observations and experiments began, Variability has hindered data interpretation. Ecology-oriented botany also strives for unambiguous, clear-cut results and tight functional correlations—ideally, the rigid cause-and-effect relationships familiar from the "exact sciences." Such an ideal can be closely approached in plant research provided certain initial and boundary conditions are met. To this end, researchers select plant material that is as uniform as possible (such as clones of a single genotype); all external variables, apart from the one under investigation, must be kept constant or non-limiting, and the results must pertain strictly to a defined range of the studied response. This is how maximum accuracy and reproducibility are achieved.

However, such results often cannot be extrapolated to natural conditions because exclusively single-factor dependencies either do not exist, or such prime "optimal" growth conditions simply do not occur, or the genotype in question plays no significant role. When observation and experimental conditions closely approximate natural ones, the results lack these drawbacks; yet, they are highly variable, making statistical Processing difficult, and often fail to yield mathematically unambiguous expressions. Ecological research constantly oscillates within this field of tension between precision and relevance. Sometimes the solution lies in combining these approaches. Whichever approach is chosen, sample-based findings must be scaled up to a real-world Setting—as large as possible and always incorporating data on biological variability (different genotypes or origins, varied habitats, different species) into the experimental or observational records. While this generally increases data scatter, it also significantly raises the labor investment.

According to current understanding, evolution is closely tied to biological variability. The survival success of a species in a changing environment is frequently determined by individuals that deviate from the norm (mean values). It is therefore desirable for experiments or observations to capture the widest possible range of variability. The broader the observational scope, the more meaningful and useful the results will be. The variability of manifestation forms and response types is of great scientific and practical significance and must not be confused with inaccuracies caused by flaws in measurement or observation techniques. On the contrary, biological variability is most clearly revealed through frequency distribution records.

Accounting for intra- and interspecific biological variability and the boundaries of a plant's living space demands higher material, temporal, and technical costs. Researchers must carefully weigh how best to obtain results: through in-depth studies of a small number of samples or less intensive studies of numerous samples. Maximally achievable measurement precision in itself is an insufficient criterion to sacrifice accounting for biological variability or natural-like growth conditions. A little information about many things is often more valuable than extensive (precise) information about very few things when it comes to validating existing theories or formulating new ones. Before setting up resource-intensive experiments, one should consider whether more affordable Methods could be employed to yield highly reliable, well-supported data that, while less precise, remain methodologically flawless in their documentation and statistically buffered against biological chance events.

Scientific ecology continually faces the challenge that while internationally recognized rules exist for the mathematical and statistical expression of results, none exist for real-world conditions, experimental/observational relevance, or object Selection. The latter always require clear interpretation, especially in experimental ecology. Aside from the necessity of capturing biological variability, five crucial criteria must be met to achieve realistic, interpretable, and—ideally—generalizable results:

✵ Integration of the plant with a specific or (ideally) natural soil environment (soil organisms, particularly mycorrhizal Fungi, availability of natural nutrients);

✵ selection of the plant at a developmental stage appropriate to the research objective;

✵ sufficient duration of observation;

✵ a representative climatic regime—i.e., differentiated in time (climatic cycles), space (SHOOT vs. ROOT), and, where applicable,

✵ neighboring plants (competition, mutualism).

Naturally, as in any natural science research, independent individual and/or spatial replications of experiments or observations are always necessary; however, the "precision" gained through Replication alone cannot compensate for shortcomings in the aforementioned five criteria or in biological diversity.