Density of species

(d) Complete additive w

Density of species 1

Fig. 10.12. Four types of competition experiments: (a) partial additive, (b) replacement series, (c) additive series, and (d) complete additive (redrawn from Silvertown and Lovett Doust, 1993).

1993) (Table 10.4). Obviously having more density combinations will give a better understanding of competitive effects, but with it comes an increased requirement of time and money. Numerous reviews have been written examining various aspects of competition experiments (Cousens, 1991; Goldberg and Barton, 1992; Gurevitch et al., 1992; Goldberg and Scheiner, 1993; Hamilton, 1994; Freckleton and Watkinson, 2000; Joliffe, 2000; Connolly et al., 2001). We will briefly examine each of the four types of experiments. The main difference among the designs is that they vary in the total density of individuals and in the proportion of each species present.

The partial additive design is the simplest type of competition experiment. Here, a 'target' species is kept at a constant density and grown in competition with a second species at a range of densities. This design is useful only when looking at the effect of increasing density on some component of plant growth. It is often used in agricultural experiments looking at weed density effects on crop yield (Buchanan et al., 1980).

In the replacement series design, the total density of individuals is kept constant, while the relative densities of both species vary. Although this method was formerly in common use, researchers stopped using it because it was impossible to distinguish the effects of total density from relative density (Silvertown and Lovett Doust, 1993). Remember, in this design total density is kept constant; only the relative proportion of individuals changes.

The additive series and complete additive designs incorporate both proportion and density into the design. These experiments use a range of total densities and the relative proportion of each species also varies. The

Table 10.4. Advantages and disadvantages of four basic experimental designs used to examine the effect of density on plant competition (based on information in Silvertown and Lovett Doust, 1993).

Advantages

Disadvantages

Partial additive

Replacement series

Additive series and complete additive

Good for applied situations where only concern is the effect of increased density of one species on another; for example, in agricultural systems; little expense, time Little expense, time

Density of both species varies independently therefore more useful; better understanding of competitive effects because all combinations of density used

Limited application

Cannot distinguish the separate effects that each species has on the other; uses relative density not absolute density Expense, and time consuming, but less so in additive series difference between the additive series and complete additive is that the latter uses all combinations of densities. With a complete additive design, it is possible to construct a response surface (Fig. 10.13) (Firbank and Watkinson, 1985).

Landhausser et al. (1996) used an additive series design to test whether three alternative species could be used to reduce com petition by blue-joint grass (Calamagrostis canadensis) but allow white spruce (Picea glauca) to regenerate. They looked specifically at the effect of density and relative planting date of red and white clover (Trifolium pratense and T. repens) and fire-weed (Epilobium angustifolium). They found that all three species reduced the growth and survival of blue-joint grass and

Fig. 10.13. Results of a competition experiment with a complete additive design. The effect of competition between wheat (Triticum) and corncockle (Agrostemma githago) was measured as the yield per plant. The effect of density on the yield of a) wheat, and b) corncockle is shown as a response surface diagram (Firbank and Watkinson 1985; with permission of Blackwell Science).

Fig. 10.13. Results of a competition experiment with a complete additive design. The effect of competition between wheat (Triticum) and corncockle (Agrostemma githago) was measured as the yield per plant. The effect of density on the yield of a) wheat, and b) corncockle is shown as a response surface diagram (Firbank and Watkinson 1985; with permission of Blackwell Science).

that the effect increased as the time between the planting of replacement species and blue-stem grass increased. The three alternative species reduced the growth of white spruce, but not to the extent that blue-joint grass did.

Using field experiments to examine competition

The above experiments are carried out in controlled situations where the environment is controlled and plants are grown in pots. The difficulty with pot experiments is that we are never sure that what we perceive to be competition will actually occur in real field situations. Greenhouse-type experiments can be used to generate predictions that can then be tested using field experiments (Freckleton and Watkinson, 2000)

Using both field and pot experiments, Berendse (1981, 1982, 1983) examined competition between narrow-leaved plantain (Plantago lanceolata) and sweet vernalgrass (Anthoxanthum odoratum). In the pot experiments, the entire soil column was colonized by roots of both species during the course of the competition experiment. However, since plantain usually has deep roots and vernalgrass shallow roots, a field experiment was done to allow for a more natural rooting profile. They used a replacement series with a constant density of 36 plants per 50 x 50 cm plot. Five plantain:ver-nalgrass ratios (36:0, 27:9, 18:18, 9:27, 0:36) were used. Two treatments were imposed: different rooting depths possible (50 cm depth), and different rooting depths prevented (20 cm depth). Plantain had a much higher shoot biomass when rooting depth was not constrained. Therefore, interpretation of experiments must be done remembering the constraints placed by the experimental design. In this case the pots produced an artefact that prevented the researchers from understanding root competition.

There are two types of field experiments used to examine competition. Either the density of neighbour plants around a target individual can be changed (neighbour manipulation) or else resources can be altered (increased or decreased) so that the intensity of competition is changed (remember, plants only compete when resources are limited).

Neighbour manipulation

In this approach, the density of neighbours around the target plant is either increased or decreased (Aarssen and Epp, 1990). Neighbours can be removed by a combination of clipping to ground level once or repeatedly, and/or applying herbicides. Sometimes roots are removed through hand digging. All these methods have the disadvantage of creating a disturbance around the target plant. If roots are left intact, then they may add nitrogen as they decompose, but digging loosens the soil and may increase or decrease root growth of the target. Trenching around the base is sometimes used to look at the effect of root competition vs. shoot competition. Increasing neighbour density is done by either planting seeds or individuals around the target plant. Again this creates a disturbance around the target plant, and the transplant may suffer from transplant shock.

Altering resource levels

Increasing resource levels can be done simply by adding nutrients to increase soil fertility. Wilson and Tilman used this technique in their previously described experiment. Alternatively, the levels of some nutrients can be lowered. For example, soil nitrogen can be decreased temporarily by adding a carbon source such as sugar and/or sawdust to the soil. This stimulates the growth of soil microbes, which accumulate the nitrogen in their bodies making it temporarily unavailable to plants (Morgan, 1994).

Alpert and Maron (2000) used such a methodology to test whether decreasing nitrogen (by using sawdust) would decrease the invasion of non-native species in a Californian grassland. They used patches of grassland where the native nitrogen-fixing bush lupin (Lupinus arboreus) had been killed with a herbicide (glyphosate). Patches of dead lupin tend to increase the invasion of non-native species because they release nitrogen when they die. Alpert and Maron (2000) used three treatments:

• control - dead lupin patch, no change in soil;

• tilled + sawdust - soil was tilled to incorporate sawdust into the soil; sawdust could not be left on surface as this created a mulch that affected other characteristics such as soil temperature and moisture;

• tilled only - to test for the effect of tillage alone.

They measured species presence and biomass for 2 years. They found that carbon addition (reduction in nitrogen) had no affect on the total above-ground biomass of native species (primarily forbs), but did decrease that of non-native species. This occurred because nitrogen reduction decreased the biomass of grasses (which were mostly non-native) while increasing the richness of broadleaf species (which were mainly native) (Table 10.5).

0 0

Post a comment