Germination percent and rate for seeds ingested by Cuora amboinensis

To determine if seed germination was enhanced following ingestion by a turtle, we conducted germination trials of ingested seeds from fecal samples compared with control seeds collected from the forest. Control seeds were collected by finding mature fruit on the forest floor and removing the seeds from the pulp. Seeds of both types were rinsed with well water. Seeds were placed in germination bags which consisted of a paper towel measuring 22 x 11 cm with an 11 x 11 cm sheet of clear plastic beneath it and a clip holding the edge of the paper towel onto the plastic. Ten control or ingested seeds of a given genus or species were placed on the half of the paper towel seated on the plastic. The other half of the paper towel was saturated with well water and then folded over on top of the seeds on the plastic sheet. The plastic sheet with saturated paper towel containing the seeds was slid into a clear zip-top plastic bag (17 x 15 cm). We inflated germination bags by blowing into them so that the upper surface of the plastic bag was not resting on the top of the paper towel. Germination bags were randomly positioned in a box measuring 195 x 189 x 12 cm and placed on a table beneath a roof outside. The bags received ambient light but none received direct sunlight at any time of the day. Each morning we inspected seeds in germination bags. Germination of seeds of <i>A. teijsmannii</i>, <i>C. trifolia</i>, <i>D. serrata</i>, and<i> V. cofassus </i>could be seen unaided, but we examined seeds of <i>Ficus </i>spp. and <i>N. cadamba</i> using a microscope at 20x magnification, because of their small size. About 4 ml of well water was added to the upper surface of the paper towel every three days or as needed to keep seeds moist. For each of six plant species, we conducted germination trials on seven replicates of seeds collected from the forest (controls) and seven to 11 replicates of seeds that had been ingested by turtles. We documented date of germination for each seed and removed germinated seeds from the bag to prevent damage to the paper towel by the cotyledon. Trials were terminated after 30 days for each germination bag because pilot trials indicated that all of our target species were capable of germination within one to two weeks, fruiting periods varied between species necessitating seed collection when available, and because of limited space in the germination box. We compared germinability (final percent germinated) between ingested and control seeds for each plant species using a general linear model with a binomial distribution and a logit-link function. We compared germination rate (days to germination) for seeds ingested by turtles and control seeds for each plant species using a general linear model with a Poisson distribution. Statistical analyses were conducted in SAS (SAS Institute, version 9.4).