I used a one-dimensional hierarchical design to determine the effect of loc
al versus long-distance neighbors on the growth and reproduction of the ann
ual weed, Cardamine pensylvanica. I measured the response of target plants
to the presence or absence of neighbors lit the first- and second-nearest n
eighbor positions and to the overall spatial pattern (clumped, even, or ran
dom) of neighbors. In addition, I investigated how variation in interplant
spacing and nutrient levels combined with neighbor effects to influence the
growth and reproduction of C. pensylvanica. How plants were affected by th
eir neighbors depended upon nutrient levels, plant spacing, and the number
of first- and second-position neighbors. Under low nutrient conditions and
the closest spacing (0.5 cm), plant growth was so restricted that individua
l neighbor effects were overshadowed by the overall density effect. At the
intermediate spacing (1.0 cm), both the first and the second neighbors infl
uenced fruit number, but at the farthest spacing (1.5 cm) only the first ne
ighbor was significant. Under high nutrient conditions, at all three spacin
gs, first neighbors significantly affected biomass and fruit number, but th
ere were differences among the spacing treatments in how target plants were
affected by neighbors. Furthermore. under high nutrients there was a signi
ficant global pattern x first neighbor interaction. C. pensylvanica grown i
n the clumped global pattern with no neighbors at the first position produc
ed significantly more fruit than did plants under any of the other global x
neighbor combinations. Although long-distance interactions were only impor
tant under special circumstances. their effect was dramatic (an almost two-
fold increase in fruit production).
This study represents the first experimental test of the assumptions underl
ying the use of cellular automata models to model plant population dynamics
. The results support the basic assumption of cellular automata models, i.e
.. that nearest neighbor interactions explain most of the variation in grow
th of plants. However, they also show that under certain conditions long-di
stance interactions also can have a great influence on the local neighbor e
ffects. Before cellular automata models are extensively applied to model pl
ant populations. further empirical work is needed to determine how competit
ive relationships among plants change with environmental conditions.