Biological Variance in Agricultural Products Theoretical Considerations

The food that we eat is uniform neither in shape or appearance nor in internal composition or content. Since technology became increasingly important, the presence of biological variance in our food became more and more of a nuisance. Techniques and procedures (statistical, technical) were developed. The most widely used are sorting or grading on large-scale operations. Statistical rules were developed on the sampling size to obtain a reliable mean value. Experimental design strategies were developed for efficient research. All the techniques used are rather empirical and only loosely connected to the theoretical and fundamental knowledge we do have. An effort is made to approach biological variance from a new perspective. The nature of biological variance is traced back to its origin in the processes of production and conversion, primarily driven by the "random" differences in climate and soil. By modelling the dynamics on the level of the individual units that constitutes a batch, rather then modelling the mean value for the batch itself, more fundamental models can be developed. These more fundamental models are generic in nature and describe changes in product properties for all kinds of circumstances like growing area, seasonal effects, harvest maturity and storage temperature. INTRODUCTION As long as man exists on this earth, he has eaten his daily food to stay alive and to survive. In the very early days, when gathering our food by hunting and searching, the variation in kind, amount and quality was tremendous. Mankind adapted to that situation and developed skills to distinguish between edible and inedible products. After a while, mankind settled down in small settlements and started farming. The era of agriculture began. The variance in quality and properties of our daily food gradually decreased to a bearable and manageable level. In the recent past, the effects of biological variance were avoided as much as possible by sorting and grading the products solely based on external properties. Statistics provided sufficient and reliable means of experimental design to decrease the effect even further. So, we learned quite efficiently to live with the problem without too much trouble. But still we are faced with the hidden and not (and certainly not fully) understood rules of biological variance. In more recent times, driven by the shift in importance from external properties (shape, colour, size) to internal properties (sugar content, taste, vitamins, potentially health promoting compounds), more problems have been encountered even in apparently similar batches of produce. Many studies have been devoted to the causes and effects of biological variance in agricultural products. Especially the growing of food has attained much effort, as that is the realm where technology fails most in controlling the onset of biological variance (see for some examples General: Tijskens et al. 2000). Altogether, the issue of biological variance has so far mainly been dealt with by technological means, without too much understanding of the rules and changes that govern the processes that cause biological variance. In this lecture, a new approach is proposed, that is based on the understanding and modelling of the behaviour of individuals in a population, rather then starting from the behaviour of the mean values of batches. Some theoretical examples are deduced and shown, that have however, a strong implication for practice. Proc. 8 Int. CA Conference Eds. J. Oosterhaven & H.W. Peppelenbos Acta Hort 600, ISHS 2003 662 ORIGIN AND SOURCES OF BIOLOGICAL VARIANCE To understand how biological variance works, how it affects the perception of quality in our food and the handling of our food commodities, we have to understand what it is and where it comes from. In a previous lecture (General: Tijskens et al. 2000), an attempt was made to review these sources and the nature of biological variance. The results are briefly summarised as follows: • Variance in any product is part of the production process, whether natural (growing) or technological (food processing). Without production of a commodity or conversion of one commodity into another, no variation will be present. Variance in commodities is a direct consequence of variations in the producing sequence. • Usually, effects of external factors, different in space and/or time initiate this variation. Examples of the most frequently responsible external factors are sunlight, daily and seasonal temperature changes, rainfall, soil type and its spatial distribution, fertilisers and their spatial distribution. Superimposed on this type of variance, is the variance that originates from internal factors as e.g. genetic differences in the producing species. • The search for the variations in the primary production process brings us to the full area of growing and cultivation. The effects of various circumstances, especially weather, climate, soil type and fertilisation have been studied for already a long time. It covers a list of over 1 million references. The literature references, provided in Tijskens et al (General: 2000), constitute merely a short list of efforts. The biological variance occurring in agricultural products, caused by external factors, is generally subdivided into three categories, reflecting more the type of observation rather than the fundamental cause of the differences. For the time being, we will stick to this classification: − Regional variation: difference between growing areas caused by soil composition, fertiliser availability and different climatic conditions due to location (e.g. latitude, vicinity of sea, ocean, mountain, desert), different cultivation treatments by growers during the cultivation period, (plowing, fertilisation, and irrigation). − Seasonal variation: The most important driving force is here the difference in weather over the different years. Indices to optimise the harvest time belong to this category. − Batch variation: the variance between the individuals in a batch. The source of this type of variation is the result of the same driving forces as for the previous two origins of variation, but on a far smaller scale. At the moment of harvest, not all individuals in a batch are at the same stage of development and exhibit a distribution in maturity. • Technical variance in machinery and processing equipment will lead to different holding and passing times (Residence Time Distributions) which on their turn will induce variance in the products under treatment. WHAT IS BIOLOGICAL VARIANCE? Biological variance is difficult to describe in a generic way. The Webster dictionary defines variance as “the fact, quality, or state of being variable or variant”, and “an instance of variableness: a degree of difference”. This definition does bring us not much further than what we already guessed. Some analogy with the definition of quality by Kramer and Twigg (General: 1970) is evident: “the composite of those characteristics that differentiate individual units of a product, and have significance in determining the degree of acceptability of that unit by the buyer”. So, we could define variance as “the composite of properties that differentiate individual units of a product”. This definition contains several interesting aspects and consequences: • Variance can only occur in a population of individual units (entities). Without individual units in a greater set or part of the total population (e.g. batch, growing lot, season, partial population, total population etc) variance does not exist. This aspect brings in the vast knowledge on population dynamics and the manifold procedures developed by statistics to describe variance, population dynamics and statistical