University of Hawaii, Honolulu, USA

Background and objectives
Plant-parasitic nematodes exist in clusters in the soil. The variance among population densities is usually greater than the mean of the population densities. This information is important in developing appropriate sampling schemes and control measures. Knowing the distribution of migratory endoparasitic nematodes such as Radopholus similis inside host roots is important for many purposes such as prediction, quarantine, and control. R. ;similis is found in root and shoot tissue of Anthurium andreanum, complicating its detection and the design of management tactics [1]. The objective of these experiments was to determine the distribution of R. ;similis in the roots and shoots of A. ;andreanum.

Materials and methods
In experiment 1, 25 A. ;andreanum cv. Midori plants growing in clay pots filled with volcanic cinders were inoculated with 2500 mixed life-stages of R. ;similis in 5-ml aliquots. The plants were grown for 30 ;days and then five plants were randomly harvested every 3 ;days. Roots and leaves were separated from the stem. Leaves were discarded. Each root was sequentially numbered with the lowest, oldest root from each plant labeled as 1. The stem was divided into 2-cm long sections and also sequentially numbered. Each root and stem section was assayed individually in a mist chamber. The numbers of nematodes recovered after 3 ;days was recorded. Frequency distributions were calculated and data analysed to determine the type of distribution.

In the second experiment, Midori plants were inoculated with 2500 mixed life-stages of R. ;similis and grown in the greenhouse for 80 ;days. Five plants were then harvested every 3 ;days. Leaves were removed and discarded. The shoot and root tissues was chopped into 1-cm long pieces, composited, and samples consisting of 1, 2, 3, 4, or 5 ;g were placed in a mist for 3 ;days to collect nematodes. Each plant was individually assayed. Data were used to calculate optimal sample size.

Results and conclusions
R. ;similis occurred in clusters in the roots and stems of Anthurium. The variance always exceeded the mean. The variance to mean was 49.54 to 24 in the roots, 17.44 to 7 in the stem tissue, and 43.90 to 19 in the entire plant. One root or stem section may have contained no nematodes, whereas a nearby root or stem section contained many. The greatest nematode populations were always found in the middle-aged roots. Because of the aggregated distribution of the nematode in the plant, larger composite samples were better than the smaller samples. Larger composite samples gave higher nematode population densities but not necessarily increased precision. A 4-g sample mixture of root and stem tissue had the lowest variance among the sample size tested.

Because R. ;similis aggregates in the roots and stem of Anthurium, it is necessary to collect appropriate tissue to assay for detection. For quarantine detection purposes, root samples should be collected from middle-aged roots. The oldest roots are often too decomposed to support nematodes, whereas nematodes in the youngest roots of a plant may not have reached high population densities. It is also appropriate to assay stem tissue. When plants are assayed for detection or management purposes, the sample should be sufficiently large not to miss nematodes and provide false negatives. Reliable population density determinations of this migratory endoparasite depend upon proper sampling methods just as for nematodes in soil.

1. Wang KH, 1996. Genetic and Cultural Control of Anthurium Burrowing Nematode, Radopholus citrophilus. MS Thesis, University of Hawaii, Honolulu.