| Soil 
              seed banks contribute to the diversity and dynamics of many plant 
              communities. In some systems they are critical for maintaining species’ 
              populations or restoring native plant communities. [Conceptual 
              model]   
              We examined the composition of the soil seed bank at Bunchgrass 
              Ridge to determine whether meadow species maintain viable seeds 
              in the soil, and by implication, whether the seed bank can contribute 
              to meadow restoration if conifers are removed.We 
            addressed the following questions: 
              Methods 
                Do meadow species maintain viable seeds in the soil? 
                Do the density and diversity of seeds decline as meadow is replaced 
                by forest?What 
                types of species dominate the seed bank?What 
                is the potential for seed banks to contribute to meadow restoration 
                if trees are removed? 
               
                | Collecting 
                  soil samples |   
                |  |  Field 
              sampling. Samples 
              of the soil seed bank were taken in May 2004 from subplots also 
              sampled for forest age 
              structure and ground vegetation. A total of 209 subplots was 
              sampled.  Sampling 
              was stratified, using a simplification of the chronosequence 
              to represent three distinct stages of encroachment: open meadow, 
              young forest, and old forest (see table, right). 
                
              In May 2004, three soil cores (6 cm in diameter and 10 cm deep) 
              were extracted from each subplot and combined (~850 cm3/subplot). 
              In Jul and Aug 2004, above-ground vegetation was sampled in the 
              same subplots. Analyses. 
              Germinants were tallied into one of three groups defined by habitat 
              affinity: ruderal, meadow, and forest understory species (see Plant 
              species list). For each group, ANOVA or Kruskal-Wallis tests 
              were used to assess differences in species richness (number of taxa 
          per subplot) or germinant density among the three stages of encroachment. | 
               
                | Greenhouse 
                  methods |   
                | 
                     
                      | Germination 
                        flats |   
                      |  |   
                      | 
                           
                            | Soil 
                              samples were mixed with sterile potting soil and 
                              placed in germination flats in the University of 
                              Washington greenhouse. Seedling emergence was monitored 
                              for seven months. |  |  |  
 
 
               
                | Light 
                  availability, tree density, and basal area among open meadow, 
                  young forest, and old forest subplots. |   
                | Structural 
                  variable | Open meadow
 | Young forest
 | Old forest
 |   
                | Transmitted 
                  light (%) | 46 | 22 | 17 |   
                | Tree 
                  density (no./ha) | 96 | 1,552 | 1,172 |   
                | Basal 
                  area (m2/ha) | 6 | 36 | 116 |  |