Flender/Flender Gear Units/Helical gearboxes H3
comparison of two varieties of Erwinia carotovora (.. carotovora andE.. atroseptica ) showed that there were few qualitative differences in

proles, but several small quantitative differencesin pathogen-host volatile signatures (. Volatile proles from potatoes (cv. Atlantic) infected with either Erwinia

carotovora var. carotovora orFusarium roseum var. sambucinum (fusarium dry rot) were studied (. Potatoes were inoculated and stored in plexiglass

containers and volatiles were trapped (Chromosorb . After thermal desorption, the volatiles were cryogenically fo- cused on precolumn (OV- and

then separated (Supelcowax . Many of the com-pounds reported earlier ( were present to varying degrees as background. The com-pounds pentane and dimethyl disulde were specic to both infections, and one additionalunidentied compound was unique to fusarium dry rot infection. 5 ASSESSMENT OF VOLATILE MONITORING The preceding literature indicates that volatile proles of diseased commodities are differ- ent from those of the intact commodities. In some cases, specic compounds associatedwith pathogenic activity have apparently been identied; however, pathogen specicityis not clear in general since many types of pathogen produce similar volatile metabolites,and the method of collection and analysis can inuence the results (. Furthermore, none of the work expressly aimed at disease monitoring in storage has managed to quantify either detected levels or rates of emanation. Thus, the question remains as to whether ornot techniques such as headspace analysis can be used in practice. This section attemptsto shed light on this question. Since no quantitative data are available, certain assumptionsare made and some operating conditions are discussed in terms of minimal detectablelevels of low-molecular-weight organic compounds. 6 Alvo et al. 5.1 Assumptions The rst assumptions are that the monitoring system is based on trapping headspace vola- tiles with an adsorbent, and that the