The usefulness of Magnetic Resonance Imaging (MRI) to investigate anatomical and physiological changes during cotton seed and fiber development was explored in (Gossypium hirsutum L.) pilose, a monogenic mutant line and Texas Marker-1, a normal line. Pilose fibers are typically shorter than TM-1, but secondary wall development has been reported to be normal (Klhel et al. 1973). MRI is a non-invasive, non-destructive technique used in clinical studies for observing human anatomical features. It is also useful for plant anatomy and physiology investigations. Because MRI spatially maps changes in water concentration, it is useful in studying plant water relationships (Brown et.al. 1986, 1988). In MRI a sample is placed in a strong magnetic field and radio frequency waves are pulsed through the sample to measure both water distribution and the thermodynamic status (degree of water binding) of protons chiefly associated with water.

Detached cotton bolls were imaged at 8, 18 and 25 days past anthesis (DPA). MRI images were acquired at 63 MHz in a Siemens 1.5T magnetron clinical imaging system housed at the University of Missouri Medical Center. A gradient refocussed pulse sequence (GRASS) with a 40 flip angle, 400 ms repetition time, 6 s echo time, 12 acquisition cycles, 2 mm slice thickness, 50 mm field of view reconstructed on a 256 x 256 matrix.

Images clearly showed changes in anatomical features within the capsule (boll) during development, including carpels, vascular system, ovule, integument, embryo sac and embryo. Individual fibers could not be detected, however, fiber masses were clearly visible. Linter development after 8 DPA could be detected on ovules. Changes in both bound and free water were detected among tissues in the different developmental stages.

These preliminary studies suggest that embryonic development in TM-1 is complete after 25 DPA where Pilose was not. Furthermore, MRI images suggest that the termination of cotyledon expansion within the embryo sac occurs from the basal end of the boll first in TM-1 and last in Pilose. In addition, cross sectional scans through the boll show seeds to be randomly positioned in Pilose and more symmetrical in TM-1. Kobel et.al. 1973, suggest no difference in seed development between the two cultivars.

MRI may be useful in studying earlier stages of boll development because preparation techniques for light microscopy are often difficult and disturb the tissue. Because MRI is non-destructive, it is possible to study not only anatomical development but also plant-water relationships in intact plants. An additional study demonstrated the feasibility of using MRI to detect changes in seed oil content in developing bolls.