Evaluation of seed-dispersal services by ants at a temperate pasture: Results of direct observations in an ant suppression experiment

Ant dispersal Ants disperse seeds of many plant species adapted to
myrmecochory. While advantages of this ant–plant mutualism for
myrmecochorous plants (myrmecochores) have been previously studied in
temperate region mostly in forests, our study system was a pasture.
Moreover, we used a unique combination of observing the effect of
ant‐activity suppression on ant dispersal and comparison of the
contribution of ant and unassisted dispersal to the distance from mother
plant. We established plots without and with ant‐activity suppression
(enclosures). We offered diaspores of a myrmecochorous (Knautia arvensis),
and a non‐myrmecochorous (Plantago lanceolata) species in a choice test
and followed ants carrying diaspores during days and nights (focus of
previous studies was on diurnal dispersal). We measured frequency and
distances of ant dispersal and compared them with unassisted dispersal
recorded using sticky trap method. The dispersal frequency was lower in
enclosures (3.16 times). Ants strongly preferred diaspores of the
myrmecochore to non‐myrmecochore with 586 and 42 dispersal events,
respectively (out of 6400 diaspores of each species offered). Ant
dispersal resulted in more even and on average longer distances (maximum
almost tenfold longer, 994 cm) in comparison to unassisted dispersal. Ant
dispersal altered the distribution of distances of the myrmecochore from
roughly symmetric for unassisted dispersal to positively skewed. Ants
dispersed heavier diaspores farther. Ants dropped the majority of
diaspores during the dispersal (which reduces clustering of seeds), while
several (11%) were carried into anthills. Anthills are disturbed
microsites presumably favorable for germination in competitive habitats.
Ants provided non‐negligible dispersal services to
myrmecochorous K. arvensis but also, to a lesser extent,
of non‐myrmecochorous P. lanceolata. Unassisted dispersal A
sticky trap of the same size as the experimental plots of the cafeteria
experiment (3 m × 3 m) was used to record unassisted dispersal (i.e.,
without any biotic dispersal agent, mostly by gravity and additionally
wind). We used linoleum covered with insect glue to trap the falling
diaspores. Five full‐sized individuals (with majority of infructescences
with mature seeds) of K. arvensis were fixed in a water container in the
center of the plot for approximately 10 days. The procedure was repeated
five times during July and August. The distances of trapped diaspores were
measured in the same manner as in the case of ant dispersal, thus the
dispersal by ants and unassisted dispersal can be compared.

蚁播（Ant dispersal）指蚂蚁传播诸多适应蚁播（myrmecochory）的植物物种种子的生态过程。此前针对蚁植共生（ant–plant mutualism）对蚁播植物（myrmecochores）的益处研究，多集中于温带地区的森林生境，而本研究的实验系统选取了牧场生境。此外，本研究创新性地结合了两种实验手段：一是通过抑制蚂蚁活动以观测其对蚁播的影响，二是对比蚂蚁介导的传播与非生物辅助传播（unassisted dispersal）对母株周边种子传播距离的贡献。

我们设置了两组样地：一组未进行蚂蚁活动抑制，另一组则通过围栏（enclosures）实现蚂蚁活动抑制。在食物选择实验（cafeteria experiment）中，我们为蚂蚁提供了蚁播植物*Knautia arvensis*与非蚁播植物*Plantago lanceolata*的传播体（diaspores），并连续昼夜观测携带传播体的蚂蚁活动（此前相关研究仅关注日间蚁播过程）。我们统计了蚁播的发生频率与传播距离，并与通过粘捕法（sticky trap method）记录的非生物辅助传播数据进行对比。

研究结果显示：围栏抑制组的蚁播频率显著低于未抑制对照组，仅为其1/3.16。蚂蚁对蚁播植物的传播体表现出极强的偏好：在总计投放的各6400份两种植物的传播体中，蚂蚁对*Knautia arvensis*的搬运事件达586次，而对*Plantago lanceolata*的仅为42次。相较于非生物辅助传播，蚂蚁介导的传播可使种子传播距离更为均匀，且平均距离更长（最大传播距离可达近10倍，即994厘米）。蚂蚁介导的传播改变了*Knautia arvensis*的传播距离分布：非生物辅助传播的距离分布大致呈对称分布，而经蚂蚁传播后则变为正偏态分布。蚂蚁会将更重的传播体搬运至更远的距离。

多数传播体在搬运过程中被蚂蚁丢弃（这可降低种子的聚集度），另有约11%的传播体被搬运至蚁丘（anthills）。蚁丘属于受扰动的微生境，在竞争激烈的生境中通常更利于种子萌发。蚂蚁不仅为蚁播植物*K. arvensis*提供了不可忽视的传播服务，同时也为非蚁播植物*P. lanceolata*提供了一定程度的传播助力。

非生物辅助传播（Unassisted dispersal）的实验细节如下：为记录无生物传播媒介介导的非生物辅助传播（主要依靠重力，辅以风力），我们使用了与食物选择实验样地尺寸一致的诱捕装置：单块样地规格为3米×3米，配套粘捕陷阱。我们采用涂有昆虫胶的油毡（linoleum）来捕获自然掉落的传播体。

我们将5株生长完整、多数果序（infructescences）带有成熟种子的*Knautia arvensis*植株固定于样地中心的水培容器中，单次实验持续约10天。该实验流程于7月至8月间重复开展5次。我们采用与蚁播传播距离测量完全相同的方法，对诱捕到的传播体的距离进行测量，由此可直接对比蚂蚁介导的传播与非生物辅助传播的传播效果。