Physical seed damage, not rodent’s saliva, accelerates seed germination of trees in a subtropical forest
Data files
Jun 03, 2024 version files 116.63 KB
-
exp_1.xlsx
-
exp_2.xlsx
-
exp_3.xlsx
-
README.md
-
seed_trait.xlsx
Abstract
Many tree species adopt fast seed germination to escape the predation risk by rodents. Physical seed damage and the saliva of rodents on partially consumed seeds may also act as cues for the seed to accelerate the germination process. However, the impacts of these factors on seed germination rate and speed remain unclear. In this study, we investigated such impacts on the germination rate and speed (reversal of germination time) of four tree species (Quercus variabilis, Q. serrata, Q. acutissima, and Q. glauca) after partial consumption by four rodent species, through a series of experiments. We also examined how seed traits may affect the damage degree by rodents by analyzing the relationship between the germination rate and time of rodent-damaged seeds and the traits. We found that artificially and rodent-damaged seeds exhibited a significantly higher seed germination rate and speed, compared to intact seeds. Also, the rodent saliva on seeds showed no significant effect on seed germination rate and speed. Furthermore, We observed significant positive correlations between several seed traits (including seed mass, coat thickness, and protein content) and seed germination rate, but these seed traits had a positive correlation with the germination rate and speed. These correlations are likely due to the beneficial traits countering seed damage by rodents. Overall, our results highlight the significant role of physical seed damage by rodents (rather than their saliva) in facilitating seed germination of tree species and potential mutualism between rodents and trees. Additionally, our results may have some implications in forest restoration, such that intentionally sowing or dispersing slightly damaged seeds by humans or drones may increase the likelihood of successful seed regeneration.
README: Physical seed damage, not rodent’s saliva, accelerates seed germination of trees in a subtropical forest
https://doi.org/10.5061/dryad.3ffbg79qt
Description of the data and file structure
exp1 sheet
Experiment 1 data
groups show the different treatments:
(1) Naturally intact acorns without rodent handling (NT)
(2) Mechanically damaged acorns by people with normal saline (MJ)
(3) Intact acorns handled by L. edwards (LeT)
(4) Injured acorns by L. edwards (LeJ)
(5) Intact acorns handled by A. draco (AdT)
(6) Injured acorns by A. draco (AdJ)
(7) Intact acorns handled by A. chevrieri (AcT)
(8) Injured acorns by A. chevrieri (AcJ)
Q.v means Q. variabilis, Q.s means Q. serrata, Q.a means Q. acutissima, Q.g means Q. glauca;
time means germination time (days)
status means germination status; 1 means germinating, 0 means not germinating;
exp2 sheet
Experiment 2
groups show the different treatments:
(1) Naturally intact acorns (NT)
(2) Mechanically damaged acorns by people with normal saline (MJ)
(3) Injured acorns treated with mouth-rinsed saliva of L. Edwards (LeR)
(4) Injured acorns treated with mouth-rinsed saliva of N. fulvescens (NfR)
(5) Injured acorns treated with mouth-rinsed saliva of N. confucianus (NcR)
(6) Injured acorns treated with mouth-rinsed saliva of A. draco (AdR)
Q.v means Q. variabilis, Q.s means Q. serrata;
time means germination time (days)
status means germination status; 1 means germinating, 0 means not germinating;
exp3 sheet
Experiment 3
groups showed the different treatments:
(1) Naturally intact acorns (NT)
(2) Mechanically damaged acorns by people with normal saline (MJ)
(3) Injured acorns treated with gland-rinsed saliva of N. fulvescens (NfG)
(4) Injured acorns treated with gland-rinsed saliva of N. confucianus (NcG)
(5) Injured acorns treated with gland-rinsed saliva of A. draco rinsed (AdG)
Q.v means Q. variabilis, Q.s means Q. serrata;
time means germination time (days)
status means germination status; 1 means germinating, 0 means not germinating;
seed_trait sheet
Shows the traits of seeds.
Q.v_rate, Q.s_rate, Q.a_rate, and Q.g_rate means The germination rate of Q. variabilis, The germination rate of Q. serrata germination rate, The germination rate of Q. acutissima germination rate, The germination rate of*Q. glauca* germination rate;
Q.v_asin, Q.s_asin, Q.a_asin, and Q.g_asin means The germination rate of Q. variabilis underwent arcsin transformation, The germination rate of Q. serrata underwent arcsin transformation, The germination rate of Q. acutissima underwent arcsin transformation, The germination rate of Q. glauca underwent arcsin transformation;
Q.v_mgt, Q.s_mgt, Q.a_mgt, Q.g_mgt means Mean germination time (MGT) (days) of Q. variabilis, Mean germination time (MGT) (days) of Q. serrata, Mean germination time (MGT) (days) of Q. acutissima, Mean germination time (MGT) (days) of Q. glauca.
treat means treatment, A means Q. variabilis, B means “Q. serrata, C means Q. acutissima, D means Q. glauca;
asin means Seed germination rate after arcsin transformation;
mgt means Mean germination time (days);
fresh_weight means Oak seed fresh weight (g);
coat_thickness means Oak seed coat thickness (mm);
protein means protein content (%);
fat means fat (%);
starch means starch (%);
tannin means tannin (%);
Cellulose means “Cellulose (%)”;
Caloric_value means Caloric value (KJ/g);
Caloric_value_per_seed means Caloric value per seed (KJ);
“dormancy” means dormancy status, 1 means dormancy, 0 means non-dormancy;
rodent_mass means rodent mass (g);
Code/Software
R version 4.3.2 was used for analysis.
Methods
To evaluate the effects of physical damage and saliva of rodents on the germination rate and speed of the oak seeds, we set up a germination experiment with intact and treated seeds. We planted the seeds in potted soil with each pot containing 5 seeds. The pots had a diameter of 12.5 cm and a depth of 11 cm. The seeds were planted at a burial depth of 2 cm, approximating the average burial depth of seeds hoarded by rodents (Xiao & Zhang, 2004). The germination progress of the planted seeds was regularly observed and recorded. Seed germination was defined as the emergence of any part of the seedling from the seed, based on the criteria outlined in a previous study by Izhaki et al. (1995). Germinated seeds were carefully removed from the pots to minimize their influence on the non-germinated seeds, by following the methodology described by Tang et al. (2007). We also recorded the germination time of each seed to examine the relationship between germination time and the traits (including seed mass, coat thickness, and protein content) of damaged seeds. In all experiments, injured oak seeds were referred to as those damaged by rodents. We also used the data from Experiment 1 to test Hypothesis 4, which postulates that different seed traits demonstrating tolerance or resistance to rodent predation might affect the germination of damaged seeds. Seed traits of four species of oak seeds are obtained from a previous study (Yang et al., 2018). We used the following criteria to measure the germination metrics of oak seeds such as germination rate and time:
1. Germination rates (GRs), measured as the proportion of germinated oak seeds among all tested seeds.
2. Mean germination times (MGTs), measured as the average germination time of all germinated seeds during the experimentation period. The reversal of MGTs is referred to as germination speed.
Experiment 1. Effects of seed damage on seed germination
To test Hypothesis 1, 2, 3 & 4 that artificial or rodent damage would speed up seed germination (higher germination rates but shorter germination time), we conducted experiments in Spring 2020, measuring the germination rate and time of partially damaged seeds of Q. variabilis, Q. serrata, Q. acutissima, and Q. glauca. These seeds were subjected to damage by three rodent species: A. draco, A. chevrieri, and L. edwards. Due to the dormancy of Q. acutissima, and Q. glauca oak seeds, we recorded their germination status weekly for 122 continuous days from January to May 2020. Prior to the experiments, the rodents were fasted for at least 6 hours and then given the oak seeds in individual plastic boxes (L × W × H = 90cm × 50cm × 40cm). Thus, both rodents and oak seeds were in the same plastic boxes for the experiment. Twenty oak seeds were given to an individual L. edwards due to its large body mass (> 200g), and ten oak seeds were given to an A. draco and an A. chevrieri, considering their smaller body size, at a time, based on their daily intake. The condition of the seeds was assessed based on feeding position and proportion. Seeds with half or more of the endosperm consumed were excluded from the germination tests based on feeding position and proportion. We used fresh intact seeds as the negative control group (NT group) serving as a reference to assess the natural conditions of seed germination without any interference from rodents or humans. This helps to establish a baseline of seed germination for making a comparison between intact and treated seeds by rodent or artificial damage. We artificially damaged seeds, in which we manually cut the seed with normal saline in half to simulate the physical damage by rodents without the effect of rodent’s saliva, as a positive control group (MJ group) to test the effects of physical seed damage and rodent saliva on germination. Fresh oak seeds, which were handled by rodents (L. edwards, A. draco, A. chevrieri), and those that were partially damaged, were grouped into the LeT, AdT, and AcT categories to test the effects of other unknown factors except for seed damage. Fresh oak seeds handled by rodents (L. edwards, A. draco, A. chevrieri), and those partially damaged were categorized into the LeJ, AdJ, and AcJ groups to test the effect of seed damage by rodents. Before planting the seeds into flower pots, all intact and partially damaged oak seeds were carefully cleaned with absorbent paper. The number of seeds involved in Experiment 1 was shown in supplement Table S9 and the eight treatment groups (NT, MJ, LeT, LeJ, AdT, AdJ, AcT, AcJ) were shown below:
NT, Naturally intact oak seeds;
MJ, mechanically damaged oak seeds artificially with normal saline;
LeT, Intact oak seeds handled by L. edwards;
LeJ, Injured oak seeds by L. edwards;
AdT, Intact oak seeds handled by A. draco;
AdJ, Injured oak seeds by A. draco;
AcT, Intact oak seeds handled by A. chevrieri;
AcJ, Injured oak seeds by A. chevrieri;
Experiment 2. Effects of mouth-rinsed saliva on seed germination
The goal of this experiment was to explore how the saliva of the four rodent species L. edwards, N. fulvescens, N. confucianus, A. draco affect the germination of Q. variabilis and Q. serrata seeds. The experiment was conducted in autumn 2021, aiming to test Hypothesis (3) that rodent saliva may accelerate seed germination. Due to the non-dormant characteristics of Q. variabilis and Q. serrata, seed germination occurred relatively quickly (Table S9), thus, from October to December 2021, initial monitoring of seed germination was conducted every 2-3 days, which was later adjusted to once a week until 80% of the seeds had germinated and no further germination was observed in the subsequent two weeks. We used an 80% germination rate as the threshold to end monitoring because the remaining seeds might fail to germinate due to various reasons, such as dormancy mechanisms and germination potential. Due to insufficient numbers of A. chevrieri, we had to substitute A. chevrieri with N. fulvescens and N. confucianus in this experiment.
Before conducting the tests, all rodents were deprived of food for at least 6 hours and we preheated saline water to a temperature of 37-38℃. We collected saliva by rinsing the mouths of each rodent with the pre-heated saline water, and this process was repeated at least 10 times per rodent to obtain a sufficient amount of saliva extract. The saliva extract volume from each rodent species (five individuals in total of the experiment) exceeds 10 ml, and saliva from an individual rodent was spread over ten seeds. We used cotton swabs to collect saliva and evenly spread it on the surface, ensuring that both sides of the seeds came into contact with rodent saliva. The number of seeds involved in Experiment 2 was shown in supplement Table S9 and the six treatment groups (NT, MJ, LeR, NfR, NcR, AdR) are shown below:
NT, Naturally intact oak seeds;
MJ, mechanically damaged oak seeds artificially with normal saline;
LeR, Injured oak seeds treated with mouth-rinsed saliva of L. edwards;
NfR, Injured oak seeds treated with mouth-rinsed saliva of N. fulvescens;
NcR, Injured oak seeds treated with mouth-rinsed saliva of N. confucianus;
AdR, Injured oak seeds treated with mouth-rinsed saliva of A. draco;
Experiment 3. Effects of gland-rinsed saliva on seed germination
The experiment is similar to Experiment 2 but we used saliva gland (N. fulvescens, N. confucianus, A. draco), aiming to test Hypotheses (3) in the autumn of 2022. We encountered a seed shortage and were unable to collect a sufficient quantity of seeds in the local forest, forcing us to reduce the number of seeds for five treatment groups from 50 to 25 this year. Due to the excessively large size of L. edwards (body weight > 200g), it was excluded from this test as it was difficult to be anesthetized. Saliva glands were surgically extracted from rodents for analysis. In each group of rodent species, we only euthanized one individual. This experiment involved one N. fulvescens, one N. confucianus, and one A. draco. Due to insufficient numbers of A. chevrieri, we had to substitute A. chevrieri with N. fulvescens and N. confucianus in this experiment. The salivary glands, including the parotid gland, submaxillary gland, and sublingual gland, were surgically removed. The collected salivary glands were then immersed in a preheated saline solution at a temperature of 37℃. The bottle containing the salivary gland diluent was gently shaken, and after 5 minutes, the diluted saliva liquid was dropped onto the surface of cut oak seeds of Q. variabilis and Q. serrata. The oak seeds were cut in half to simulate physical seed damage by rodents while ensuring that the embryo within remains viable. Our objective was to maintain consistent levels of damage across all seeds to evaluate their influence on germination rates under controlled conditions. After the experiment, all animals raised were treated harmlessly. The number of seeds involved in Experiment 3 was shown in supplement Table S9, and the five treatment groups (NT, MJ, NfG, NcG, AdG) are shown below:
NT, Naturally intact oak seeds;
MJ, mechanically damaged oak seeds artificially with normal saline;
NfG, Injured oak seeds with gland-rinsed saliva of N. fulvescens;
NcG, Injured oak seeds treated with mouth-rinsed saliva of N. confucianus;
AdG, Injured oak seeds treated with mouth-rinsed saliva of A. draco;