water and minerals are less likely to be 
a problem along railroads close to the 
edge of a road because there is com- 
monly a layer of soil mixed in with, and 
sometimes completely overlying, the 
cinders at these sites. This soil is prob- 
ably washed from the edge of the road 
by rainwater. I believe that the better 
growing conditions provided by this 
soil material, and the abundance of 
seeds deposited by vehicles, account for 
the large diversity of plant species close 
to the edges of roads. 
Disturbance is at least as important 
as stress in limiting plant diversity 
along a railroad. As soon as it emerges 
into the air, a seedling runs the risk of 
being trampled underfoot, smeared 
with oil by a passing train, or eaten by 
an herbivore. To those individuals that 
survive these insults, the coup de grace 
is delivered by the railroad company, 
which sprays weedkiller on the tracks 
every year, usually in midsummer. This 
maintenance practice means the end of 
life for the plants and the end of the 
field season for me. More than any oth- 
er disturbance phenomenon, spraying 
determines which species of plants will 
flourish on railroad tracks from year to 
year. It prevents the establishment of 
plants, including most perennials, that 
flower and set seed only after an ex- 
tended vegetative period, and it selects 
for those plants, primarily annuals, that 
grow quickly through a short vegeta- 
tive phase and then flower and set seed 
early in the summer. By the time the 
seeds of these annuals are ready to ger- 
minate at the beginning of the next 
growing season, the weedkiller from 
the previous year will have decomposed 
or been leached out of the surface layers 
of the cinders. Once a railroad is closed, 
however, disturbance becomes much 
less severe, and perennials, with their 
ability to overwinter, will eventually 
outcompete annuals, which must start 
anew from seed each year. 
On an active railroad, plants of small 
stature have a competitive advantage 
because tall individuals are likely to be 
decapitated by a passing train. Some 
tall species, such as wild parsnip, Queen 
Anne’s lace, and common mullein 
( Verbascum thapsus), do grow on rail- 
road cinders. On an infrequently used 
line, individuals of these species may 
grow to be several feet high in the inter- 
val between one train and the next. The 
first train to travel along the line after 
such an interval chops the tops off the 
tall shoots. This does not necessarily 
spell doom for the plants: I have often 
seen individuals of wild parsnip and 
Queen Anne’s lace flowering on un- 
characteristically short shoots that 
grew following decapitation of the first- 
formed, long shoot. Mullein, however, 
does not have such a mechanism of 
regeneration. The rosettes of leaves 
formed by this biennial in its first year 
grow well on railroad cinders, but indi- 
viduals that grow too close to the tracks 
will probably have their single flower- 
ing shoot cut off by passing trains long 
before they can set seed. 
Of all railroad weeds, dwarf snap- 
dragon ( Chaenorrhinum minus ) has in- 
terested me the most. In the United 
States, this species grows more abun- 
dantly along railroad tracks than any- 
where else, contributing greatly to the 
appeal it has for me. I have come to ad- 
mire the tenacity with which this tiny, 
beautiful flowering plant flourishes in 
the seemingly inhospitable environ- 
ment provided by the dry, sun-baked 
cinders of a railroad. Like most weed 
species of railroads, roadsides, waste 
Tall plants, such as wild parsnip, 
are at a distinct disadvantage on an 
active railroad. If trains do not pass 
by too frequently, however, some 
individuals can set seed before they 
are decapitated. 
areas, and fields, dwarf snapdragon is 
not native to the United States. Ecolo- 
gists believe that native weeds are the 
first to colonize newly disturbed areas 
of natural vegetation. Nonnative weed 
species follow, usually with human as- 
sistance, attached to clothing, foot- 
wear, vehicles, and equipment, for 
example. Frequently disturbed areas 
generally have more species of intro- 
duced weeds than of native weeds. 
The original home of the dwarf snap- 
dragon is believed to be the Mediterra- 
nean region of Europe; from there it has 
spread to, and become locally abundant 
in, dry fields and railroad cinder ballast 
throughout most of Europe. Before 
railroads existed, this species was prob- 
ably confined to dry, gravelly soils in 
fields and waste areas. The building of a 
continental network of railroads during 
the nineteenth century provided ave- 
nues for the snapdragon’s widespread 
dispersal throughout Europe. Dwarf 
snapdragon was first recorded in the 
United States at Camden, New Jersey, 
in 1874. Presumably, the species made 
the trans-Atlantic crossing as seeds car- 
ried along with the cargoes of ships. On 
its arrival in the United States, the plant 
encountered an existing railroad net- 
work, and since the latter part of the 
nineteenth century, dwarf snapdragon 
has spread relatively rapidly — as far as 
Kansas and Nebraska in the west, 
North Carolina in the south, and Que- 
bec and Ontario in the north. 
The reported spread of dwarf snap- 
dragon along railroad tracks prompted 
me to investigate the efficiency with 
which the seeds of this plant are dis- 
persed by passing trains and to com- 
pare this mode of dispersal with natural 
phenomena such as wind and rain. I 
quickly found that rain splash is not a 
mechanism of seed dispersal in this spe- 
cies: the two openings in the wall of 
each mature fruit, through which the 
seeds escape, close up completely when 
the fruit is wet or exposed to conditions 
of high humidity. 
For the experiments on wind disper- 
sal, I set up sticky traps around plants 
that were growing near, but not on, 
railroad tracks and left them for at least 
five days. Each trap consisted of a strip 
of filter paper painted with a thin layer 
of melted petroleum jelly and glued to a 
wooden board that lay flat on the 
ground. I arranged similar traps 
around other plants that were rooted in 
the cinders between the rails and in the 
path of approaching trains. Normally, 
most seeds that fall on the highly po- 
rous cinders quickly percolate down 
61 
