Developing Methods To Estimate Pod-drop and Seed-shatter In Canola

Rob H. Gulden; University of Manitoba Project Code: 3.4.2 Final Report: April 2013 Canola is a crop with high potential for seed-shatter and pod-drop, however, at this time no effective and proven tools to accurately and consistently estimate losses are available. Researchers at the University of Manitoba conducted a three-year study to evaluate different methods of estimating pod-drop and seed-shatter. The study showed the catch tray and visual ratings were the most time efficient methods evaluated, however the catch tray method is the only one that can estimate both pod-drop and seed-shatter. The vacuum method provided a good measure of total harvest losses. A new force gauge method for measuring pod retention has the potential to be a tool for rapidly assessing pod drop in canola, however refinement and validation across a more broad range of germplasm and different environments is required. Canola is a crop with high potential for seed-shatter and pod-drop, which can cause substantial losses at harvest. Seed-shatter and pod-drop, which are both agronomic and economic issues, vary among genotypes and are influenced strongly by environmental conditions. As direct combining becomes more prevalent, the need for pod-drop and seed-shatter resistant canola varieties increases. However, at this time no effective and proven tools to accurately and consistently estimate pod drop and pod shatter are available. Researchers at the University of Manitoba conducted a three-year study from 2010 to 2013 to evaluate different methods of estimating pod-drop and seed-shatter. The objectives of the study were to compare visual rating and catch tray or vacuum based methods, assess the potential use of digital images and evaluate a new pod retention resistance method. They also wanted to determine if there were difference in pod-drop and seed-shatter between open-pollinated and hybrid canola varieties. Field trials were conducted in 2011 and 2012 at two research farms in Manitoba, Carman and Kelburn. Eight different canola varieties (four hybrid and four open pollinated) were seeded at two target densities (40 and 120 seeds/m2). In the catch tray method, trays were placed in each plot at the early pod filling stage and removed immediately before direct-harvest. Seed-shatter and pod-drop (full pods) were measured and losses determined by seed weight and as a percentage of total yield. The vacuum method was used to determine total canola harvest loss including both pre- and post-harvest losses. A wet-dry vacuum cleaner collected samples and canola seeds recovered were measured as weight of seeds and as a percentage of total yield. Researchers also conducted a visual estimation

of seed-shatter in a standing canola crop, and evaluated the use of digital camera and image analysis computer software for estimating losses.

Fig. 1. Trays used to determine pod-drop and seed-shatter in canola before harvest. In 2012, catch trays were modified by adding a top section to reduce possible predation and / or the effect of adverse weather conditions. Researchers also developed a new pod retention resistance method specifically for this study to estimate pod-drop in canola. Both field and greenhouse experiments were conducted using a force gauge to determine the relationship between pod retention force and other harvest loss measurements. Data was collected from a number of different pods at designated plant positions (main rachis or branches) and rachis portions (middle of the top half or middle of the bottom half), and represented the force required to break the pedicle of the pod from the rachis, the weakest point of attachment of the pods. Pods on which pod-retention resistance was measured were collected to determine total pod weight and seed weight in each pod. In the field study, pod-retention resistance and pods were collected one week before combining and on the day before combining the crop. In the greenhouse, data and pods were collected at five weekly time intervals immediately before harvest.

Fig. 2. Digital force gauge tool in action in the field.

The results of the study showed that the catch tray method was the only method that allowed for the simultaneous, independent determination of both pod-drop and seed-shatter. The catch tray and visual ratings were the most time efficient methods evaluated, however visual ratings only provided a good measure of seed shatter. The vacuum method provided a good measure of total harvest losses immediately after harvest, however it is unable to distinguish between pod-drop and seed-shatter. For most varieties, pre-harvest yield losses determined using the catch tray method agreed well with total pre- and post-harvest losses determined using the vacuum method. These experiments were direct-harvested at a very late stage to be able to determine differences in pod-drop and seed-shatter among varieties and therefore it is not surprising that most of the harvest losses were incurred before harvest.

Fig. 3. Examples of the visual estimation of seed-shatter in two varieties. The percentage of shattered pods that were still attached to the plant was rated 70% (left) and 5% (right). The new pod retention resistance method provided quantitative data on pod-drop potential and interesting insights into the pod-drop phenomenon. The study showed that pod-retention is not affected by moisture content and overall, pod-retention resistance was remarkably consistent among varieties over years and locations. This new force gauge method for measuring pod retention has the potential to be a tool for rapidly assessing pod drop in canola, however refinement and validation across a more broad range of germplasm and different environments is required. The study also showed that the digital image analysis, which in a greenhouse experiment seemed promising, was more challenging than expected in the field. The current image analysis software did not have the capacity for shape and size recognition, and it was not possible to adapt this software for rapid pod or seed recognition from digital images obtained from the field. Therefore, this method was too laborious and difficult to develop for rapid and efficient data generation at this time. Generally the study showed that hybrid and open-pollinated varieties had similar seed shatter and pod-drop losses when measured with a catch tray. However, when the measures were different, canola hybrids generally exhibited more seed-shatter than pod-drop compared to open-pollinated varieties.

Scientific Publications Cavaileri A, Lewis DW, Gulden RH (2014) Seed-shatter and Pod-drop are not closely

related in Brassica napus. Crop Science 54:1184-1188.