UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of disturbance in permanent pastures Parish, Roberta 1987

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1987_A1 P37_2.pdf [ 7.62MB ]
JSON: 831-1.0097487.json
JSON-LD: 831-1.0097487-ld.json
RDF/XML (Pretty): 831-1.0097487-rdf.xml
RDF/JSON: 831-1.0097487-rdf.json
Turtle: 831-1.0097487-turtle.txt
N-Triples: 831-1.0097487-rdf-ntriples.txt
Original Record: 831-1.0097487-source.json
Full Text

Full Text

THE ROLE OF DISTURBANCE IN PERMANENT PASTURES by ROBERTA PARISH  A THESIS SUBMITTED  IN PARTIAL FULFILMENT OF  THE REQUIREMENTS  FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in • THE FACULTY OF GRADUATE STUDIES Department of Botany  We accept this thesis as conforming to the required standard  THE UNIVERSITY  OF BRITISH COLUMBIA  September  1987  ® Roberta Parish, 1987  In  presenting  degree  this  thesis  in partial fulfilment of  requirements  for  an  of  department  this thesis for scholarly or  by  his  or  her  I further agree that permission for  purposes  permission.  Department of The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3  DE-6(3/81)  °\  OcJr  extensive  may be granted by the head of  representatives.  It  is  understood  that  publication of this thesis for financial gain shall not be allowed without  Date  advanced  at the University of British Columbia, I agree that the Library shall make it  freely available for reference and study. copying  the  copying  my or  my written  ABSTRACT  This thesis investigates how small disturbances influence community structure in three permanent pastures. Small disturbances play an important role in providing spatial heterogeneity that permits new recruits to enter populations in closed sward communities, thereby promoting diversity and species co-existence.  The thesis has four components: the first two are based on observation and measurement of the occurrence of small disturbances, molehills and dung pats, in three pastures. Within pasture seasonal changes in disturbance regime were related to changes in patterns of species abundance by multidimensional contingency table analysis. Dactylis glomerata, Agropyron repens and Taraxacum officinale increased in highly disturbed plots, whereas Holcus lanatus and Trifolium repens decreased. Invasion of molehills and dung pats was usually by rhizomes or stolons from surrounding plants. Seedling recruitment was rare: Trifolium repens was the only species dependent on small gaps for sexual regeneration. Patterns of species replacement on and around the disturbances were non-random.  The third part of the thesis investigated the effects of selective removal of Lolium perenne, Holcus lanatus and Trifolium repens from the oldest and youngest pastures. Strong responses to the removal of these species were found only in grasses in the youngest pasture. This is consistent with the hypothesis that competition decreases over time because of niche divergence, but may also reflect biological accommodation to grazing pressure.  The fourth part of the thesis investigated changes in species composition in simulated swards in response to different regimes of mowing, fertilizer and small gap creation. Species composition was strongly influenced by mowing and fertilization but was unresponsive to small gap creation.  ii  ACKNOWLEDGEMENT  Many people gave advice and support during the years of research and preparation that led to this thesis. I would like to acknowledge the advice, guidance and patience of my supervisor, Dr. Roy Turkington, and the support of my committee members, Drs. J . Maze, W. Neill and G. Bradfield.  I am especially indebted to Dr. Malcolm Greig who provided both advice and patient counsel on statistical methods. John Emmanuel and Ken Pollson wrote computer programs to analyse species replacements.  I have been fortunate in being associated with fellow students who were gracious with their time and counsel, in particular, Loyal MehrhofT, Rich Evans and Rob Scagel. I have been doubly fortunate in having excellent technical help, notably, Elena Klein, who worked long hours in all kinds of weather at tedious and repetitive tasks without complaint.  Financial support through postgraduate fellowships from the University of British Columbia and the MacMillan Family and field support from the Natural Science and Engineering Research Council of Canada are gratefully acknowledged.  I also wish to thank Mr. Bill Chard and Ms. Mary Chard for access to their pastures and for many kindnesses during my field work.  Finally, I acknowledge the role of my children, Rochelle and Robert Adams, who have grown, while I attended Graduate School, into supportive, independent people and note that, without them, this thesis would have been finished years ago.  iii  T A B L E OF CONTENTS Abstract  ii  Acknowledgement  iii  1. Introduction 1.1. Study area  1 8  2. Patterns of disturbance and their influence on botanical composition 2.1. Introduction 2.2. Methods 2.2.1. Field methods 2.2.2. Analytical methods 2.3. Results 2.3.1. Vegetation 2.3.2. Disturbance 2.3.3. Interaction of vegetation and disturbance 2.4. Discussion 2.4.1. Vegetation 2.4.2. Disturbance 2.4.3. Interaction of vegetation and disturbance  19 19 22 22 25 29 29 37 44 48 48 50 51  3. Invasion of disturbances 3.1. Introduction 3.2. Methods 3.2.1. Field Methods 3.2.2. Analytical methods 3.3. Results 3.3.1. Species persistence: self-transition 3.3.2. Patterns of species replacement Monthly transition probabilities Transition probabilities over the survey period 62 3.3.3. Colonization of disturbances Seedling recruitment Colonization patterns 3.3.4. Changes in botanical composition 3.4. Discussion 3.4.1. Species persistence 3.4.2. Species replacement and invasion 3.4.3. Recruitment and colonization  55 55 56 56 57 58 58 60 60  4. Influence of species removal on interspecific interaction 4.1. Introduction 4.2. Methods 4.2.1. Field methods 4.2.2. Analytical methods 4.3. Results 4.3.1. 1939 pasture  90 90 93 93 96 97 97  iv  66 66 66 74 83 83 84 87  4.3.2. 1977 pasture 4.3.3. Expansion rate 4.4. Discussion  100 103 116  5. Influence of mowing, fertilization and disturbance regimes 5.1. Introduction '. 5.2. Methods 5.3. Results 5.4. Discussion  120 120 122 124 132  6. General discussion  136  7. Literature cited  146  8. Appendix 1  160  v  List of Tables Table  1-1.  Species  composition  of  Highland  Mix  with  sowing  proportions used in 1977 pasture  16  Table 1-2. Species list for the three pastures in 1985  17  Table 2-1. Mean species cover in the three pastures averaged over eight surveys from July 1982 to October 1984 Table 2-2. Mean number of disturbances, molehills and dung per 5x5m plot in the three pastures at each survey period Table Table  39  2-3. Significant interaction of species abundance with current and previous disturbance regime, molehills, and dung distributions  45  3-1. Percentage of species that persist throughout the nine month study period in the sward around molehills and dung pats  59  Table 3-2a. Average monthly transition probability of a row species being replaced by a column species for eight common species on and around molehills Table  30  61  3-2b. Average monthly transition probability of a row species being replaced by a column species for eight common species on and around dung pats  61  Table 3-3. Deviations from random of replacement patterns of common species, growing on and around molehills and dung pats in the three pastures  63  Table Table Table  3-4. Species showing significant differences in the monthly pattern of replacement when self-transitions are included  64  3-5. Species showing significant differences in the monthly pattern of replacement when self-transitions are excluded  65  3-6.  Number  of  seedlings  per  square  metre  found  on  disturbances in the three pastures Table 4-1. Species removed from eight plots in the 1939 and 1977 pastures Table 4-2. Percentage cover of removed species in four replicates in July 1983 Table 4-3. Mean difference (n = 4) in species percentage cover in the 1939 pasture from 1983 to 1984 Table 4-4. Mean difference (n = 4) in species percentage cover in the 1939 pasture from 1983 to 1985 vi  67 ..95  95 98  99  Table 4-5. Mean difference (n = 3) in species percentage cover in the 1977 pasture from 1983 to 1984  101  Table 4-6. Mean difference (n = 4) in species percentage cover in the 1977 pasture from 1983 to 1985  102  Table 4-7. Mean expansion rate for the period 1983-1985 for species found initially in at least three of four replicates  115  Table 5-1. Percentage composition of the original seed mix  124  Table 5-2. Mean and standard deviation of the percentage cover of species and bare ground on South Campus treatment plots in September 1985  125  Table  5-3. Percentage variation in abundance of common species explained by different mowing, fertilization and divot removal regimes  vii  .126  List of Figures Figure  1-1. Monthly mean temperatures at Aldergrove, B.C. summarized for the period from January 1982 to December 1984 compared with the 21-year average of monthly mean temperatures 10  Figure  1-2. Monthly mean maximum and minimum temperatures at Aldergrove, B.C. summarized for the period from January 1982 to December 1984 compared with the 21-year averages  12  Figure 1-3. Monthly mean precipitation at Aldergrove, B.C. summarized for the period from January 1982 to December 1984 compared with the 17-year average of monthly mean precipitation  14  Figure 2-1. Location of the plots on contour maps of the three pastures  23  Figure 2-2. The. percentage cover of the 10 most common species in the three pastures over the study period from summer 1982 to fall 1984  31  Figure  2-3. Reciprocal averaging ordination of the abundance of all species found in the three pastures in eight surveys  35  Figure 2-4. Maps of the three pastures showing plots with consistently high, intermediate and low levels of total disturbance  40  Figure 2-5. Maps of the three pastures showing plots with consistently high, intermediate and low numbers of molehills  42  Figure 3-1. Average number of Trifolium repens seedlings per molehill or dung pat -germinating in April and surviving until August in each pasture for two consecutive years  68  Figure  Figure  3-2. Invasion ratio of species final percentage abundance on disturbed sites divided by initial percentage abundance around the disturbance  70  3-3. Percentage abundance of species at the final survey on molehills and dung pats, in the surrounding sward, and over the entire quadrat  75  Figure 4-1. Expansion rate from 1983 - 1985 of species remaining in each of the four replicates after the removal of Lolium perenne,  Figure  Holcus lanatus and Trifolium repens  104  4-2. Expansion rate from 1983 - 1985 of species remaining after removal treatments against the percentage cover removed in each of four replicates of eight treatments (n = 32)  109  viii  Figure  Figure  5-1. Percentage abundance of nine species on fertilized and unfertilized plots subject to different mowing regimes  127  5-2. Number of species per square metre on fertilized and unfertilized plots subject to different mowing regimes  130  ix  1. INTRODUCTION  Plant communities  are characterized by heterogeneity in both space and time.  They are non-uniform systems that continually change. Changes result from the death and regeneration of individual plants and their parts.  The  view that vegetation is a dynamic system has had a long history. Aristotle  imparted the principle of motion to all natural things but held that change directed  towards  Although towards  a  goal.  Galileo believed goals,  the  This all  view motion  observation  earth I continually see  was to  be  of change  herbs, plants,  maintained circular,  was  animals  for  nearly  2000  and  hence,  not  reiterated generating  in the  years. directed  Dialogue  and decaying  was  "On  .... the  appearance of the earth undergoing perpetual change".  The  Aristotelian view of motion towards  a goal was held by one of the most  eminent early ecological theorists. Clements (1916) considered the plant community developing toward a climax vegetation state, analogous to an organism undergoing change and development.  This view was upheld and promoted by many ecologists  but not without opposition. One of the main challenges came from H.A. Gleason (1926) who argued that the botanical composition of an area resulted from the interaction  of  two  'fortuitous  and  fluctuating  factors',  immigration  and  environment.  The  view of communities as structured, orderly entities, still persists even though  Clements' organism viewpoint has been discredited with the shift in ecology, from  1  Introduction / 2 Aristotelian essentialism  to more probabilistic thinking. Structure, i.e.  order and  organization, may be imposed either from within the community, by the plants themselves,  or  externally,  through  forces  imposed  upon  the  community.  distinction was originally applied by Tansley (1935) to successional  This  communities.  He distinguished autogenic succession in which "changes are brought about by the action of the plants themselves on the habitat" from allogenic succession in which "changes  are  brought  structure,  inferred  about  from  by  external  non-random  factors".  patterns  of  Evidence  species  of  community  distribution  in  the  community, has been used to challenge Gleason's hypothesis of random association of individual species. Ecologists, however, have taken an increasingly 'Gleasonian' approach as itself,  the  evidence  accumulates  distribution  of  which  that "every in  space  species of plant is a law unto depends  upon  its  pecularities  of  migration and environmental requirements. Its dissemules migrate everywhere, and grow  wherever  they  find  favourable  believed to result from feedback composition  and  abundance  of  recruits (Harper 1977). Thus,  conditions"  from established seedlings,  although  and  (Gleason  1926).  Structure  is  individuals, controlling species  from interactions  species have  distinct life  lead to an individualistic response, pattern in their association  between these histories  that  suggests rules of  assembly rather than random events.  Disturbances are a natural component of all communities. Recently, the influence of disturbance regimes on community structure and function has become the focus of much research effort. This interest arose from two impetuses. First, a body of evidence  accumulated  against  classical  Clementsian succession  communities were continually reset to earlier successional  and showed  that  stages by disturbances.  Introduction / 3 This led to an hypothesis that community composition could be controlled by the rate and intensity  of disturbance (cf.  Connell  1978). Second, the application of  Gause's hypothesis of competitive exclusion of species occupying the same niches led to an examination of niche requirements of plant species. The argument has been advanced by Harper (1969) that "the essentially similar requirements of all green plants - solar radiation, water, carbon dioxide, and a basic set of mineral nutrients - provide little opportunity for diversification in relation to food supply. This is in strong contrast with organisms at higher trophic levels where the food net  is  focused  less conspicuously  on a few  environmental  supply factors, and  much of the diversity of fauna can be interpreted in terms of differences in food or feeding habit." Grubb (1977), who repeated and expanded this notion, contends that our lack of understanding of plant co-existence is due in part to a failure to  take  into  Furthermore,  account  the  disturbance  phenomenon  could  provide  of a  regeneration source  of  in  plant  temporally  communities. and  spatially  heterogeneous sites for vegetation regeneration that would permit niche separation and species co-existence.  A recent definition of disturbance (Sousa for  regeneration.  displacement,  Sousa  defines  1984a) focuses on gaps providing sites  disturbance  as  "a discrete,  punctuated  killing,  or damaging of one or more individuals (or colonies) that directly  or indirectly creates an opportunity for new individuals (or colonies) to become established". floods, diggings sources  Disturbance  drought,  storms,  can  result  and from  of mammals and insects. of disturbance appears  from  both  biological  physical  processes  processes,  e.g., fires,  e.g., predation, grazing,  The impact of both physical and biological  similar, removing organisms  or their parts and  Introduction / 4 creating opportunities for recruitment (Sousa 1984a).  Two additional definitions de-emphasize recruitment as a characteristic response to disturbance  and  emphasize  the  discrete  onset  of  disturbance  and  change  of  resource availability. Bazzaz (1983) defined disturbance as "a sudden change in the resource base of a unit of the detectable  change  disturbance  as  in  population  "any relatively  landscape  response".  discrete  that is expressed  White  and  event in time  Pickett that  community, or population structure and changes resources, or the physical environment". Harper (1977) distinguishes  as  a readily  (1985)  defined  disrupts ecosystem, substrate availability,  between disasters, and  catastrophic disturbances. Disasters happen sufficiently frequently in the life span of populations that they exert a selective pressure whereas catastrophies occur so infrequently that populations would lose their "genetic memory" of it by the next time it recurred (Begon et al. 1986)  and therefore exert no selective force. In  this context the eruption of Mount St Helens is termed catastrophic whereas the eruptions of Kila Laua would be disasterous.  Disturbance regimes  are characterized by various descriptors. White and Pickett  (1985) list:  distribution; frequencj  turnover  spatial  rate,  r  which  refers  to  the  and return cycle;  time  needed  to  rotation period or  disturb  a  given  area;-  predictability; size or area of disturbance; magnitude intensity which refers to the force  of  the  community; disturbances. disturbance.  disturbance  event;  and, synergism, Not  all  of  For example,  the  these  severity effect  of  on the  descriptors  it would be  the  are  impact  on  subsequent applicable  unprofitable  to  the  organism or  occurrence to  measure  ever}' the  of other type  of  magnitude  Introduction / 5 intensity of animal mounds whereas this is an important descriptor of avalanches where the force required to snap tree stems is of interest.  Natural disturbances  occur in most communities  Pickett and White 1985) (e.g.  (e.g.  intertidal  areas  1974;  Christensen (e.g.  reviews by Sousa  1984a;  including, for example, coniferous and deciduous forests  Henry and Swan  shrubland  (see  Oliver and Stephens 1985),  Dayton  grassland  1971;  Levin  1977;  (e.g. and  Runkle 1981,  Loucks Paine  et  al.  1974;  1982),  1985)  Sousa  and 1979).  Disturbance affects various levels of organization: the impact on individual growth, architecture, 1985)  may  Thompson  and dispersal result 1985),  composition,  in age  richness  (e.g.  changes (Tande  (Denslow  Watt in  1925;  Sousa  population  1979),  size  1980,  1985),  1979;  genetic  (Veblen  Canham structure  1985)  dominance  and Marks  and  and  (Jain in  1983;  community  structure  (Brokaw  1985).  While the prevalence of disturbance has been documented, there exists a lack of cogent theory concerning the processes and effects of disturbance. Two hypotheses on  the  effects  of  disturbance  on  disturbance hypothesis (Connell 1978)  communities  are  current:  the  and the "rate of competitive  intermediate displacement"  hypothesis (Huston 1979). Both are interrelated and suffer from imprecision. The intermediate disturbance hypothesis argues that species richness will be higher in communities intensity  experiencing  intermediate  level  of  disturbance  frequency,  and size, and lower in those experiencing both much lower and much  higher frequencies its  some  relationship  of disturbance. Problems arise in quantifying intermediacy and to  the  system  under  investigation.  Other  problems  arise  on  Introduction / 6 subdividing disturbance into components such as magnitude, frequency and size in which  intermediacy  (Pickett  and  in  White  each  may  1985).  The  have  different  underlying  impacts  assumptions  on of  the the  community intermediate  disturbance hypothesis are based on a successional sequence undergoing directional change toward a climax steady-state. Disturbance resets the successional sequence to  an earlier stage, increases within community heterogeneity and, over larger  geographic areas, establishes a dynamic equilibrium (Connell 1978).  Huston's (1979) "rate of competitive displacement" hypothesis may be viewed as an extension of the intermediate  disturbance hypothesis (Connell 1978), although  it owes an academic debt to both Grime (1973, see theories  of  disturbance  predator-mediated (or management,  co-existence  (Caswell  Peet et al. 1983) 1978).  In  these  or predation and herbivory) delays the  and to theories,  competitive  exclusion of one species by another by periodic reductions in population size. The prolonging of the period to exclusion promotes species co-existence and maintains diversity.  Competitive exclusion  is  accelerated  dominant species and diversity  is  increased  rate  of  nutrients similar  growth,  including  disturbance.  by increased  growth  rate  of the  by any condition that reduces  For  example,  low  levels  of  the  essential  would reduce growth rates and result in higher diversity than in a situation  with  higher  nutrient  availability  (Huston  1979),  though,  of  course, very low nutrient levels restrict the number of species able to establish leading  also  to  predator-mediated  low  diversity  co-existence  (Grime model,  1973, focuses  1979). on  Huston's  the  role  model, of  like  individual  the or  population reduction in promoting species co-existence but fails to take account of changes in resource availability initiated by disturbance. For example, disturbance  Introduction / 7 may  alter  nutrient  availability  (Bazzaz  1983;  Vitousek  1985)  or  create  sites  suitable for seedling or other propagule establishment (Grubb 1977; Sousa 1979).  There is a rich literature on the impact of disturbance on forest and intertidal communities have  ( Sousa  1984a).  mainly neglected  rangeland  (see  Studies  pastures  White  1985)  of disturbance  (but see  Jones  in  grassland  1933a,b,c,d) and have  and prairie systems.  For example,  biomass  mounds  increased  spatial  heterogeneity,  production, and enhanced  led  to  been in  in tall  prairie, Piatt (1975) studied the role of badger mounds on species Badger  communities  grass  associations.  temporal partitioning in  species co-existence by  providing a site for  colonization and establishment of a guild of fugitive species that were resticted to these  small-scale  disturbances.  Unlike  Piatt's  study,  Loucks  et  al.'s  (1985)  investigation of the response to fire, gopher mounds and small-scale erosion sites in the Wisconsin sand prairies found preferential colonization that led to changes in  species  abundance  rather  than  specialist  colonization  of  disturbances.  They  concluded that sand prairie grassland community composition is closely linked to the multiple-scale disturbance regimes.  My thesis examines the  permanent  the role of disturbance in a common grassland community,  pasture.  Pastures  are widespread  throughout temperate  climates  and contain a mix of sown and native species. They persist in many areas only because of periodic disturbance from either mowing or grazing without which they would be rapidly invaded by shrubs and trees. In addition, pastures are subject to numerous small-scale disturbances from small mammals and insects as well as scrapings and droppings from grazers.  Introduction / 8 The thesis has four components: the first two are based on observations of the occurrence of small-scale disturbances and species abundance in three pastures. In the  third  and  fourth  parts,  experiments.  Field  heterogeneity  in pastures.  regimes  were  observations  related  to  First,  species were  abundance  spatial  within-pasture seasonal  changes  in  to  manipulated  study  changes  used  was  pattern  of  species  and  in  field  temporal  in disturbance  abundance.  Second,  colonization of small gaps recorded at monthly intervals was used to determine the temporal patterns of species establishment. In the third part, the effects of selective removal of dominant grasses and white clover on the abundance of the remaining species was investigated to determine the extent of the contribution of dominants to community structure and species co-existence. of the study involved the response  setting  The final component  up of an experiment to investigate  species'  to the interaction of larger-scale disturbance, mowing, with small-scale  gap formation under different nutrient regimes.  1.1. STUDY A R E A  The  study  area comprises  three  contiguous  pastures  Highway, Aldergrove, British Columbia (SW 1/4  sec.  located  at  25704 Fraser  25, Twp. 10) on a farm  owned by William and Mary Chard. The farm is situated at 49° 03' 45" N. Lat. and 122° 30' 45" W Long, in the Coastal Douglas fir biogeoclimatic zone (Krajina 1965). The elevation of the pastures varies from 110m to 122m above sea level. Two of the pastures are generally flat and slightly sloping, whereas the third is gently rolling (see contour maps in Aarssen and Turkington 1985a).  Introduction / 9 The pastures lie on a parent material of Pleistocene  glaciomarine deposits. The  soils are Luvisolic Humo-Ferric Podzols and Orthic Humic Gleysols (Canada Soil Survey Committee 1978). They are moderate- to fine-textured  clay loams which  are moderately to poorly drained. The average top soil depth (A + B horizons) is from 80-100cm. of  detailed  The pH is moderately acid, ranging from 5.2  soil  analysis  for  the  pastures  can  be  found  - 5.9.  in  Results  Aarssen  and  Turkington (1985a).  The  area  has  a  relatively  mild,  wet  winter  and  Environment Canada climatic data recorded at  a  warm,  drier  Aldergrove, 2.4km  summer.  east of the  farm, provided monthly mean temperatures for the 21 year period from 1960 to 1980  (Figure  1-1).  During  the  study  period  generally  milder  than  average:  monthly  somewhat  higher  than  normal whereas  (1982-1985),  mean  monthly  minimum mean  winters  have  been  temperatures  were  maximum  temperatures  were near normal (Figure 1-2). December 1983, however, was a particularly cold month in which precipitation was also lower than normal. Mean annual rainfall for the period 1964-1980 was 1652.3mm (S.D. 232.8), most of which fell in the winter (Figure 1-3).  May 1984  was exceptionally  wet  (265.7mm) but this  was  well below a seasonal peak of 347.4mm in November 1983.  The  area  was  continued  over  pastures.  These  respectively, originally  first the  cleared next  were  for  forty  last  farming  years  ploughed  and prior to reseeding,  sown  with  around  to  encompass  and  seeded  1900  and  all of in  a seed mixture,  comprising by  the  1939,  all had been pastures.  periodic  clearing  present  1958  and  three 1977  All pastures were  volume,  5-10%  Trifolium  Introduction / 10  Figure 1-1. Monthly mean temperatures at Aldergrove, B.C. summarized for the period from January 1982 to December 1984 compared with the 21-year average of monthly mean temperatures.  A  1982  •  1983  O  1984  •  21-year average  MONTHLY MEAN TEMPERATURE 20-,  Introduction / 12  Figure 1-2. Monthly mean maximum and minimum temperatures at Aldergrove, B.C. summarized for the period from January 1982 to December 1984 compared with the 21-year averages.  A  1982  • 1983 O 1984  • 21-year average  MONTHLY MEAN MAXIMUM TEMPERATURE  Introduction / 14  Figure 1-3. Monthly mean precipitation at Aldergrove, B.C. summarized for the period from January 1982 to December 1984 compared with the 17-year average of monthly mean precipitation.  A  1982  • 1983 O 1984 •  17-year average  Introduction / 16 repens, 15-20% Dactylis glomerata and  70-80% of a mixture  Buckerfields Highland mix. The species proportions in the exact composition  seeded is known only for the  1977  known locally  as  mix varied and the  pasture  (Table 1-1).  The  mix used in 1977 differed from that used in 1939 and 1958 by the addition of tetraploid L. perenne and the large 'Ladino' T. repens. Various other species not in the original sowing mix have become established in the pastures (Table 1-2).  Table 1-1. Species composition of 'Highland Mix' with sowing proportions used in 1977 pasture.  Species  Percent  composition  Dactylis glomerata L.  45  Trifolium pratense L.  20  Lolium perenne L .  15  Lolium multiflorum Lam.  10  Phleum pratense L .  5  Trifolium repens L .  2  Trifolium repens var. Ladino  3  Introduction /  Table  1-2. Species list for thethree pastures in 1985. 1939  1958  1977  Grasses  Agropyron repens (L.) Beauv. Agrostis alba L. Alopecuris geniculatus L. Alopecuris pratensis L. Anthoxanthum odoratum L. Dactylis glomerata L. Festuca pratensis Huds. Festuca rubra L. Glyceria declinata Breb. Holcus lanatus L. Lolium multiflorum Lam. Lolium perenne L. Lolium multiflorum x Lolium perenne Phleum pratense L. Poa compressa L. Poa trivialis L.  X  X  X  X X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  Non-grasses  Achillea millefolium L. Capsella bursa-pastoris L. Carex sp. Cerastium vulgatum L. Cirsium arvense (L.) Scop. Gnaphalium uliginosum L. Hypochoeris radicata L. Juncus sp. Medicago lupulina L. Plantago lanceolata L. Plantago major L. Ranunculus acris L. Rumex acetosella L. Rumex crispus L. Rumex obtusifolius L. Stellaria media (L.) Vill. Taraxacum officinale Weber Trifolium pratense L. Trifolium repens L.  X  X X  X  X  X  X  X  X  X  X  X  X  X  X  X  X X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  X  x - denotes species present in pasture  Introduction / 18 A strict management regime is not followed. The farm was managed for dairy production until 1983 and subsequently switched to beef production; however, no changes in management practices in the pastures ensued. The cattle may graze in  the  pastures  between  May and  November,  though  frequently,  grazing  is  delayed until after the pastures are cut for hay. In recent years, the pastures have been cut in late June or early July and then strip-grazed by 20-30 cattle once sufficient regrowth has occurred. After a pasture has been strip grazed, the cattle roam freely until the end of the grazing season, usually in November. The pastures have never been treated with chemical pesticides, herbicides or fertilizer, other  than  fertility.  barnyard  manure  which  is  spread  sporadically  to  maintain  soil  2. PATTERNS OF DISTURBANCE AND THEIR INFLUENCE ON BOTANICAL COMPOSITION 2.1. INTRODUCTION  Pastures  are  vegetation  regarded  as  biotic  plagio-climax  vegetation  (Tansley  maintained by grazing and mowing. This management  1939),  i.e.  regime retains  the pasture in a relatively stable state but there is nonetheless wide spatial and temporal  variation in  Watkin and Clements  botanical  composition  of  pastures  (e.g.,  Spedding  1971;  1978; Vickery 1981; Thorhallsdottir 1984; Snaydon 1985).  The management regime is also one of the most important ecological features of the pasture because grazing animals not only defoliate,  but also create many  small-scale disturbances within the pasture - hoof marks, urine patches and dung pats. These disturbances, and additional activities of small mammals and insects, create  openings  or gaps within the otherwise  continuous  manner of colonization of these gaps may have  sward. The rate and  an important impact on the  species composition of the pasture, species diversity and population flux.  Dung and urine have been found to affect  pastures  (e.g.  Norman and Green  1958; During and McNaught 1961). Urine provides nutrients, mainly nitrogen and potassium, to the sward which may stimulate yield and depress the abundance of clovers (During and McNaught 1961; Mundy found  that  a  single  application  of  urine  composition, but around dung patches, Festuca  rubra  all increased  had  Dactylis  in both cover  19  1961). Norman and Green (1958) a  negligible  glomerata,  effect  Trifolium  and yield. MacDiarmid  on  species  repens  and  and Watkin  Patterns of disturbance and their influence on botanical composition / 20 (1971) showed that grasses (mainly Lolium spp.) increased in relative abundance around dung patches while clovers (mainly T. repens) did not. They also found that plants underneath dung decayed rapidly and regrowth was minimal although Hutchinson  (1979) noted that  a number of species (e.g.  Poa trivialis, Agrostis  stolonifera, Lolium spp.) were able to penetrate the patch and survive. T. repens was one of the first invaders of areas left bare by decaying dung (Weeda 1967).  On average a cow produces about 9kg of urine and 28kg (fresh weight) of dung daily,  defecating  11-15  times per day  (MacLusky  1960;  Marsh  and Campling  1970). In the 150 day grazing season typical of the Aldergrove pastures, even if cows were penned at night and only 50% of dung were deposited while grazing, that would amount to about 2100kg  of dung per animal and 63,000kg for a  small dairy herd of 30 cows. The area covered by a single dung pat ranges 2 from 0.05m  2 to 0.12m  dung is estimated  (Petersen  et al.  1956)  and the total area affected by  to be about six times the actual area covered (Norman and  Green 1958; MacLusky 1960). Therefore from 7425 to 17,820m  2  (20-50% of the  three pastures) could be affected by dung in a 150 day grazing season. Harper (1977) summarized the possible effects of dung in pastures: (i) smothering and exclusion of light from plants; (ii) a local change of nutrient relations; (iii) a local change  in the  areas;  (iv)  and  pattern of grazing because  creation  of  an  area  for  animals  colonization  tend by  to new  avoid fouled or  existing  individuals.  Another source of spatial variation in many pastures is from molehills. Molehills  Patterns of disturbance and their influence on botanical composition / 21 in the Fraser Valley are produced by the coast mole (Scapinius orarius True). These  moles  eat  arthropods, annelids  and molluscans  exclusively.  A mole can  consume more than 100 worms per day. A study on moles in the Fraser Valley found that the number ranged from 2 - 10 per hectare and was correlated to the  number of earthworms  (Glendenning 1957). Moles are least active  in the  summer months. In search of food, a single mole may create 200 to 400 hills between October and March, excavating up to 500m of tunnels usually from 15 to 20cm below the soil surface (Glendenning 1957).  Tansley  and  Adamson  (1925)  noted  that  molehills  in  dune  characteristic flora differing from the surrounding vegetation. seed  in  population  molehills on  in  molehills  pastures than  (Jalloq in  the  1975) sward.  showed Rothwell  a  slacks  had  a  A study of buried  different  (1977)  buried seed  investigated  the  colonization of molehills in Welsh pastures and noted that Holcus lanatus, Agrostis capillaris  (syn.  A.  tenuis) and Anthoxanthum odoratum had significantly  lower  abundance on molehills than in the adjacent sward, whereas Ranunculus repens increased on molehills. In contrast, also in Welsh pastures, Davies (1966) found molehills to be colonized mainly by A. capillaris and implicated molehills in the degeneration of productive pasture to that dominated by weedy species. Molehills have effects in common with dung in that (i) plants are smothered and excluded from light and (ii) bare areas are created for plant colonization.  Temporal variation in plant abundance in pastures may be related to variation in the  environment  such  as  seasonal  changes,  grazing management,  or fertilizer  application (Snaydon 1985). Temporal variation may also reflect the lifespan of  Patterns of disturbance and their influence on botanical composition / 22 plants  or parts of plants  (Thorhallsdottir 1984)  in which the  genotype  of the  plant interacts with abiotic and biotic factors.  This  chapter examines  molehills, on the  the  influence  of two  distribution and relative  small-scale  disturbances,  dung and  abundance of species in a series of  adjacent pastures. The approach taken was to describe and correlate patterns of vegetation  and patterns  of  disturbance  within  pastures.  Three  objectives  were  pursued: (i) to assess temporal variation in species abundance; (ii) to measure the intensity of disturbance caused by cow dung and molehills; and, (iii) to relate the pattern of disturbance to the pattern of distribution of species abundance.  2.2.  METHODS  2.2.1. Field  methods  To measure vegetation  patterns and disturbance regimes in the three Aldergrove  pastures, each pasture was subdivided into 5 x 5m plots. In order to facilitate comparison with  data previously  plots were systematically that  were  pastures'  from  these pastures  (Aarssen  1983),  located at 10m intervals along established transect lines  approximately boundaries  collected  were  10m  apart.  used  in  Buffer all  zones  pastures  of to  obviously swampy areas. There were 48 plots in the  10m avoid 1977  -  20m  edge  from  effects  the and  pasture, 50 in the  1958 pasture and 47 in the 1939 pasture (Figure 2-1).  The plots were surveyed on eight occasions; each spring, summer and fall, from  Patterns of disturbance and  their influence on botanical composition  Figure 2-1. Location of the plots on contour Elevation in metres.  maps of the three pastures.  / 23  1 9 5 8 PASTURE  Patterns of disturbance and their influence on botanical composition / 25 July  1982  molehills .5  x  to October 1984.  At each survey, the position and diameter of all  and all dung pats were plotted.  .5m  diagonals.  quadrats per  plot  placed  Each quadrat was  at  Vegetation was 2m  intervals  surveyed using four  along  one  of  the  plot  subdivided into a grid of 25 points. The species  rooted at each point, or bare ground, was recorded. This method was preferred over that used by Aarssen (1983) in these pastures because it is less subject to bias toward larger species and error from variability in wind speed  and grass  length.  2.2.2. A n a l y t i c a l methods  Vegetation data were analysed by reciprocal averaging of the total mean rooted cover for each pasture,  using  an algorithm outlined by  Orloci (1978). Species  cover for each sampling period was compared between pastures to detect similar patterns  and  trends,  using  the  Michigan  Interactive  Data  Analysis  System  (MIDAS) (Fox and Guire 1976) version of profile analysis.  Multidimensional disturbance describes  and  the  contingency patterns  table of  analysis  species  structural relationship  was  abundance.  between  used  to  analyse  Contingency  variables  (Fienberg  patterns  table  of  analysis  1980)  and  is  especially useful when variables are not independent. Its application to ecological data is discussed  by Fienberg (1970) who recommends  its  use  if  observations  have unequal variances and if the underlying distribution is not normal.  Variables  are  cross-classified  into  discrete  categories referred  to  as  cells.  The  Patterns of disturbance and their influence on botanical composition / 26 expected value of a cell (e„) is calculated by multipying the row total the column total  ) by  and dividing by the total number of observations (N):  e.. = n.,n,JN  y  +  1+  (1-1)  +J  Equation (1-1) may be expressed as a log-linear model of the form  loge.j=logn.  This  model  can  (Fienberg 1970,  be  +  + logn j - logN  (1-2)  +  reformulated  in  terms  analogous  to  analysis  of  variance  1980)  lo .. = u + u ge  1 ( i )  +  (1-3)  u  where: u - grand mean of logarithm of expected cell frequency M  l(i)  " deviation  from grand mean  of expected counts in cell of  level i of variable 1 u. or  - deviation  from grand mean  of expected counts in cell of  level j of variable 2 The model assumes that the variables are independent and if this is not so, an interaction effect must be incorporated such that \oge.. = u + u  1 ( i )  + u  2Q)  + u  i  m  (1-4)  In this form, a greater number of variables may be incorporated but the number  Patterns of disturbance and their influence on botanical composition / 27 of  possible  models  rapidly  rises  as  the  number of  variables  increases.  For  example, in the case of two variables, only two models are possible, with three variables  there  are  8  different  models  and with  four variables  113  different  models (Fienberg 1980). It is possible that a large number of these models will fit the data and the saturated model with all possible interaction terms always produces expected values equal to the observed values. The statistic used to test goodness of fit of the models is the likelihood ratio chi-square:  (observed/expected)  (1-5)  This statistic may be partitioned into several additive parts and was used in this study to test for the presence of an interaction effect.  To  examine  whether  disturbance  regime  abundance, a four-way classification  was  has  used.  any This  relationship consisted  to  of one  species response  variable, species abundance, and three explanatory variables, disturbance regime, season and year. Species abundance was categorized as high, medium or low for species with greater  than  10% cover, and into presence  or absence  for those  with less than 10% cover. Disturbance regime, of a plot was categorized as high, medium or low. Season (spring, summer or fall) and year of the survey were included to account for differences  in abundance associated  vegetative  and  management  growth that  and vary  flowering, from  year  to  the  various  year.  The  with the timing of  aspects selected  of  climate  model  and  assumed  interactions existed between disturbance, year and season, and between abundance, year and season. To test for a relationship between disturbance and abundance,  Patterns of disturbance and their influence on botanical composition / 28 the interaction term of abundance and disturbance was added to the model. The difference between the likelihood ratios for the two models was tested against a chi-square  distribution  significantly three-way abundance,  to  improves  determine  the  interaction  fit  between  disturbance  likelihood ratio of each  of  if the  the  model.  abundance,  and year  was  addition The  disturbance  accomplished  model containing one  of  the  interaction  term  more  specific  test  and  season,  or between  by  three-way  the  for  a  subtraction of the  term from  the model  containing both terms. The two-way interaction of disturbance and abundance is a simpler  but more  specific  case  of three-way  interaction  (Fienberg  1980). The  choice of models may be termed the maximal choice in that the model contains as many terms as possible.  Effects  may be missed in this approach but any  significant improvement has no other explanation in terms of the full model used.  One potential  problem in this type  of analysis  is the  formation of categories  from discrete data. There is no best way to accomplish the groupings although the  choice  of  cut-off  points  for  categories  does  influence  the  size  of  the  interaction observed (Fienberg 1970). Cut-off points to group disturbances regimes were  kept  constant  over  the  three  pastures.  If  a  total  of  three  or  fewer  molehills or dung pats per season were found in a plot, that plot was arbitrarily considered to have a low disturbance rate,  similarly between three  and fifteen  was considered an intermediate rate and more than fifteen, a high rate. Cut-off points for each species were varied because of the probability that the carrying capacity of each field differs for some species. A number of cut-off points were tried for each species.  Patterns of disturbance and their influence on botanical composition / 29 2.3. RESULTS  2.3.1. Vegetation  The species composition of the three pastures differed in mean percent cover (Table 2-1). Percent cover of the ten most abundant species in each pasture at each of the eight surveys is shown in Figure 2-2. Profile analysis revealed no significant parallel trends in the abundance of any species in the three pastures, even though some changes in species cover appeared to be strikingly consistent; e.g. Poa compressa abundance from summer 1983 to fall pastures; Lolium perenne from fall  1982 to fall  1984 in all three  1984; and, Trifolium repens  throughout the study period in the 1939 and 1977 pastures. One trend found in Aarssen's (1983) cover data, collected from 1979 to 1981, that continued through 1982 to spring 1984, was the decline in abundance of Dactylis glomerata in the 1977 pasture. There were other major shifts in species abundance; an increase in abundance of T. repens in the 1977 and 1939 pastures, and an increase in L. perenne in the 1939 pasture, but these did not follow from any pattern observed in the previous three years. The abundance of Festuca rubra in the 1958 pasture is related to topographic features.  The short-lived, grazing-sensitive Trifolium  pratense, present in the Aarssen (1983) study of these pastures, had declined to less than  1% of the cover and consequently has been grouped with 'other  species'.  The results of reciprocal averaging ordination of species abundances from all eight sampling periods combined (Figure 2-3) showed the 1939 and 1977 pastures, the  Patterns of disturbance and their influence on botanical composition / 30  Table 2-1. Mean species cover in the three pastures averaged over eight surveys from July 1982 to October 1984. Species  Lolium perenne Holcus lanatus Dactylis glomerata Agropyron repens Phleum pratense Poa compressa Trifolium repens Taraxacum officinale Ranunculus acris Plantago lanceolata Festuca rubra Agrostis alba Anthoxanthum odoratum Cerastium vulgatum Juncus spp. Bryophytes Bare ground Dung Other  1977 17.87 13.79 9.51 5.76 2.18 13.36 17.92 9.04 1.77 1.08 0.02 1.02 0.93 0.03 0.06 2.01 2.34 0.59 0.69  1958 10.18 13.39 15.11 8.06 6.11 16.13 7.08 6.03 1.05 1.60 6.58 0.55 0.58 0.13 0.20 1.49 4.42 0.80 0.46  1939 23.05 12.98 4.98 2.76 0.91 20.78 17.37 7.09 2.33 0.21 0.28 1.34 0.51 0.08 0.00 0.66 2.03 1.31 0.97  Results /  Figure 2-2. The percentage cover of the 10 most common species in the three pastures over the study period from summer 1982 to fall 1984.  SPR - spring, SUM - summer •  1977 pasture  •  1958 pasture  •  1939 pasture  PERCENT COVER _ O  ^.  ^  Ol  O  -1  1  -• CT  1  M  M  U)  O  Oi  O  1  PERCENT COVER _ O  I I  Ol  ' —I  -» O  I  _» Ol  NJ O  J  po Ol  L  PERCENT COVER  CO o  _  O  —  Ol  —I  O  1  _A  Ol  1  ISJ  O  (sj  Ol  CO  o  I L__J  CD  as O i  r-  Q  O O  09  I' C/) I/) Q 00  CD  CD  "ft  Ci OB  va  CD  PERCENT COVER O  cn i  i  O  _.  cn  i  M  i  O  K>  cn I  co -y  o o" c  -HP  8" Q  —4-  c  i  CO CO  PERCENT COVER  PERCENT COVER  Patterns of disturbance and their influence on botanical composition / 35  Figure 2-3. Reciprocal averaging ordination of the abundance of all species found in the three pastures in eight surveys from July 1982 to October 1984. INITIAL - indicates the first survey in July 1982.  AXIS 1 (49% variation)  AXIS 1 (49% variation)  Patterns of disturbance and their influence on botanical composition / 37 oldest and youngest pastures respectively, grouped near each other on the first axis,  distant  from  the  gradient based on the  mid-aged  1958  pasture.  The  first  axis  represents  a  average abundance of each species and the second axis  reflects the change in abundance of species located on the extremes of underlying physical  gradients  (Hill  1973).  On  the  second  axis,  all  pastures  displayed  approximately the same range of variation, although at slightly different positions along the axis. When points were joined sequentially to form a time-line between surveys, species composition in all pastures followed the same direction along the second axis. On the third axis, species composition appeared to oscillate in the 1977 and 1939 pastures.  2.3.2. Disturbance  Molehill measure.  and dung counts The  three  in each  pastures  plot  show  were  different  summed  to  disturbance  form  a disturbance  regimes  (Table  2-2).  Means for each pasture were tested by using SNK on square root transformed data. The mean number of disturbances in the 1939 and 1958 pastures was not significantly different but dung formed the major component in the 1939 pasture and  molehills in the  1958 pasture. The 1977 pasture had a significantly lower  number of disturbances  than  the  other  two  pastures  were  a larger number of molehills than in the  1958  had similar  average  amounts  of  dung.  1939  The  (p<0.05)  although  pasture.  The 1977 and  amount  of  dung  there  and the  number of molehills varied both seasonally and yearly in each field.  To  determine  if  the  distributions  of  molehills,  dung  and  disturbances  were  Patterns of disturbance and their influence on botanical composition / 38 random,  these  were  tested  against  the  Poisson  distribution,  or  against  the  negative binomial distribution if the variance- mean ratio indicated contagion (i.e. > 1). The overall disturbance pattern fitted either the Poisson (11 out of 24) or the  negative  binomial (11  of  24)  distributions  and the  remaining two  were  contagious distributions that did not fit any of the simple unimodal distributions (Table 2-2). Dung deposition usually fitted a Poisson distribution (19 out of 24), whereas Complex  molehill  distribution fitted  contagious  bimodal  a  negative  distributions  binomial in  were  not  16  tested  of  because  24  cases. of  the  computational difficulties surrounding them.  The number of disturbances in a plot was examined to determine whether certain areas  of  the  pastures  had  consistent  patterns  of  disturbance.  Categories  of  disturbance regimes previously defined for contingency table analysis were used to group plots. All three pastures had plots that had consistently high, intermediate and low disturbance rates (Figure 2-4). There were also a number of plots that varied greatly in the amount of disturbance from one sampling period to the next. In the 1977 and 1958 pastures, the consistently low disturbance plots were associated mole  with a lack of molehills. The 1939  activity  consistently  but  the  higher  dung  low in disturbance. Molehill  levels  pasture also had areas of low in  this  pasture  distribution has been  made  no  plot  similarly mapped  (Figure 2-5). Because dung deposition was almost always random, the amount in each plot varied at each survey although there was a slight tendency for more to be located at the narrow entrance to the 1939 pasture.  Table  2-2.  period.  Mean number o f  The  distribution  data or  (*)  disturbances,  fitted a f i t t e d no  Poisson simple  molehills  distribution  a n d dung p e r except  5x5m p l o t  those  in  indicated  the by  three  (*)  pastures  fitted  a  at  each  negative  survey binomial  distribution.  Molehill  Disturbance 1939  1958  1977  1939  1958  6 .00  1 . 98  7.23  1 .87*  9.26  13.24*  7.75*  4.62*  Dung 1977  1939  1958  1977  3 .77*  4. 13  0 . 14*  3 .46  3 .96*  4.64  4.. 18  3 .79  3 .73*  5.66  3 ..58  1 ..19  3 .58  5.43*  3..04  0 .0  Spring  1983  6.68*  9.80*  4 .92*  1 .02*  Summer  1983  7.26  7.22  3.58  1 .83*  * 1 .84 9.06* * 6 . 22 4. 18*  13 .73  13.56  1 .98*  11.46*  6 .33*  1 1 .75  6 ,.46  7 .23  18 .63 * 2.21 11.45*  * 17.92 15.50*  12.19*  4.40*  12.64*  6 .04*  14.23*  2 . 86  6..15*  5.30*  4.42*  2.21*  5.30*  4 .42*  0.0  0 .,0  0..0  12.02*  9.73  4.34*  7.92*  4 .96*  7.11  4.. 10  4 .77 .  9 .40  10.37  7.92  2.78  7 . 33  4 .60°  6 .62  3 ., 0 5  3. 3 2  Summer Fall  Fall  1982  1982  1983  Spring  1984  Summer  1984  Fall  1984  Mean  Means  of  Disturbance,  a  Molehill,  or  a  b  Dung t h a t  do  not  a  share  b  a common  code  (a,b,c)  a  are  b  significantly  b  different  (p<0.05).  .co  Patterns of disturbance and their influence on botanical composition /  Figure 2-4. Maps of the three pastures showing plots with consistently high, intermediate and low levels of total disturbance. Plots left blank varied in the amount of disturbance from one survey period to the next.  high level  i  intermediate level low level  Patterns of disturbance and their influence on botanical composition / 42  Figure 2-5. Maps of the three pastures showing plots with consistently high, intermediate and low numbers of molehills. Plots left blank varied in the number of molehills from one survey period to the next.  Si  high number intermediate number low number  Patterns of disturbance and their influence on botanical composition  / 44  2.3.3. Interaction of vegetation and disturbance  The  interaction  of species  abundance  and  the disturbance  regime  was  first  determined for two possible response sequences: (i) an immediate response to the current disturbance regime and (ii) a delayed of  the previous  season.  These  two  response to the disturbance regime  sequences  were  components of disturbance, molehill distribution interaction disturbance shown  with  abundance  patterns  in Table  some  increased,  was  some  and dung  decomposed  species  decreased  of the nine  showed  and  an  some  into the  distribution. Component  also determined. Significant  and the abundance  2-3. Each  then  interactions between  most  common  individualistic  interacted  species are  response  with  season  pattern: or  year,  increasing or decreasing over certain time periods.  Lolium perenne had no change in abundance regime  of the current  survey.  There  increase in abundance in the next and  1977  pasture  pastures.  was  dependent  the  species  non-significant  survey  response  on  year,  was,  however,  a  the disturbance  significant  (p<0.05)  in highly disturbed plots in the 1939  of abundance that  with  to disturbance  is, management  in the 1939  practices  and  climatic  L. perenne increased in two pastures, it should not be assumed  regimes. Although that  The  associated  always  negative  increases; interaction  in  the  between  1958  pasture,  abundance  there  and  was  a  disturbance  (0.5>p<0.1).  Holcus lanatus had approximately 15%)  and  the same  cut-off  the same  points  were  abundance used  in all pastures  in all analyses.  (about  H. lanatus  Patterns of disturbance and their influence on botanical composition Table  / 45  2-3. Significant interaction of species abundance with current and previous (1) disturbance regime, (2) molehills, and (3) dung distributions: (+) indicates increased abundance and (-) indicates decreased abundance at high disturbance rates. * - p<0.05; ** - p<0.01; *** - p<0.001; s - seasonally dependent interaction; y - yearly dependent interaction.  Species  Pasture  (1)  Lolium perenne  1977 1958 1939  Holcus lanatus  1977 1958 1939  Trifolium repens  1977 1958 1939  Dactylis glomerata  Current season disturbance  1977 1958 1939  (2)  Previous season disturbance  (3)  +  *** **  **  *** *  + s  +  +  ***  ***  s/y  +  +  Taraxacum officinale  1977 1958 1939  +  +  s  + '  1977 1958 1939  +  Phleum pratense  1977 1958 1939  Ranunculus acns  1977 1958 1939  + :  1977 1958 1939  +  *** **  +  Agropyron repens  Poa compressa  (2)  +  +  *  + *  +  s/y  +  + '  (3).  Patterns of disturbance and abundance season.  decreased  In  the  (p<0.05): a was  a  the  1977  decrease  to disturbance  pasture, however, the  response  in abundance  in the  occurred  in the  was  summer  spring.  pastures  (p<0.05). This may  contribution  to the  also a significant decrease 1958  be  and  fall  regime  1977  in these  interaction was  dung  deposited  negative in the spring and  season  but  there  dung made  pastures. There  p<0.05, respectively). In the to  on  (p<0.05) and  in plots disturbed in the previous season  lanatus showed a seasonal response  current  H. lanatus appeared to  because molehills rather than  disturbance  pastures (p<0.001 and  (p<0.05). The  in the  dependent  primarily with increasing number of molehills in the  major  and  in response  slight increase in disturbed areas  decrease 1958  in all pastures  their influence on botanical composition / 46  in the  in the  1958  was 1977  pasture,  previous  H.  season  summer but positive by  the fall.  Trifolium repens decreased in response to current disturbance in the two pastures in  which  and  1958  significant interactions  pastures respectively). The  the 1977  the  1958  significant 1977  and  (p< 0.001  overall decrease  and  p<0.05 in the  is suggested  to be  1977  due  to  T. repens increased in abundance with increasing dung deposition  molehills because in  were found  pasture (p<0.05) although the increase was pasture  (p< 0.001).  in the  survey  1958  pastures  The  in the  overall  fallowing  respectively),  and  decrease  season this  restricted to summer in in  (p<0.01 decrease  abundance  was  still  and  p<0.05  in the  was  associated with  molehills.  Three showed  species,  Dactylis glomerata, Agropyron repens and  significantly  increased  abundance  in highly  Taraxacum officinale  disturbed plots  in all three  Patterns of disturbance and their influence on botanical composition / 47 pastures.  In  suggesting  addition,  that  it  all  was  three that  species  component  increased of  significantly  disturbance  which  with most  molehills affected  abundance. There were some exceptions to the overall pattern (Table 2-3).  T.  officinale, in  its  particular, had  a  strong  seasonal  and yearly  component  to  interaction with disturbances.  The response of Poa compressa to disturbance was highly variable. In the pasture,  P.  compressa showed  a  seasonal  interaction  with  1958  current disturbance  (p<0.05), increasing in the summer and fall but decreasing in the spring. The interaction with the previous disturbances was also seasonal but only increased in the  fall,  summer  that  is,  increase  the into  decrease in spring persisted the  fall,  carryover into spring. In the  but  1939  the  fall  into the  increase  in  summer  and the  abundance  did not  pasture, P. compressa abundance interacted  with disturbance (p<0.05) and was most abundant at medium disturbance rates. There was  a seasonal  component to the interaction with molehills in which P.  compressa only increased in the summer (p<0.01).  Phlenm pratense was also inconsistent in its response to disturbance. In the pasture,  P. pratense increased  in highly  disturbed plots  1977  (p<0.05). P. pratense  showed an overall decrease with disturbance in the 1958 pasture (p<0.05) and dung in the 1939 pasture (p<0.05) and showed seasonal and yearly variations in its interaction with molehills. P. pratense was absent from molehill sites formed in the spring but present on them in the summer and fall. Because P. pratense generally had peak abundance in the spring (Figure 2-2) and declined in relative abundance  over  the  growing season,  the  overall effect  of disturbance  was  to  Patterns of disturbance and their influence on botanical composition / 48 decrease  abundance.  deposited  (p<0.05):  P.  pratense responded  presence  seasonally  in spring was  positively  to  the  dung previously  related  to dung of the  previous fall and, to lesser extent, summer abundance to spring dung but fall abundance was negatively related to dung deposited in the summer.  No overall significant response  to disturbance was recorded for Ranunculus acris.  This species showed decreased  abundance (p<0.05) at high molehill densities in  the  1939  pasture.  disturbance,  In  although  the  decreases  disturbances. R. acris was present  in  medium  1977  pasture, in  there  abundance  absent from plots  density  dung  sites.  was were  no  response  associated  to current  with  previous  with many molehills but it  These  interactions  led  to  was  no overall  significant response to disturbance.  2.4. DISCUSSION  2.4.1. Vegetation  Species abundance in the three pastures is characterized by spatial and temporal heterogeneity.  Fluctuations in abundance from year to year have been documented  in other grassland communities (e.g. Rabotnov 1966, 1985)  although  remarkably  the  little  chalk  over  a  grasslands  studied  12  year  -  15  1974; Kleter 1968; Snaydon  b3' Grubb  et al.  (1982) changed  period. Undoubtedly,  some temporal  variation is associated with seasonal growth patterns such as the spring increase in Phleum pratense abundance and the fall increase of Trifolium repens (Figure 2-2).  Whether  gains  represent  vegetative  expansion  or  recruitment  into  the  Patterns of disturbance and their influence on botanical composition / 49 population, or declines result from death of entire plants or from their parts, is unknown. This presents a problem, however, only for demographers interested in the fate of individuals.  Despite  the  lack  of  significant  parallel changes  in  species  abundance  among  pastures, reciprocal averaging ordination indicated trends in botanical composition followed the same direction in all pastures. The absence of trends associated with pasture age differs from the findings of Aarssen (1983) in these pastures. He found a strong directional trend in community composition and high variability in species composition in the 1939  pastures  1977 pasture. Since Aarssen's survey, the  1977 and  have become more similar in species composition. It is  possible  that topographical variation evident in the 1958 pasture differentiates its botanical composition  from  the  other  two  relatively  flat  pastures.  The method of data  collection may have contributed to differences in results. Rooted cover is a two dimensional method and as such is less variable than Aarssen's method which allows for a three - dimensional estimate. Information from the vertical dimension is sacrificed in rooted cover estimates for less bias and greater repeatability.  Aarssen (1983) attributed differences the and  in the variability in species composition in  1977 compared to 1939 pasture to the time available for biotic interactions, hence,  for  selective  forces  to  accumulate  and  generate  biological  accommodation. Biological accommodation between individual plants led to increased stability  in  the  accommodation  old pasture. takes  place  Results within  from the  this  first  study, five  however,  years  after  suggest  that  seeding  and  establishment. This accommodation may be in response to both biotic and abiotic  Patterns of disturbance and their influence on botanical composition / 50 factors and involve both genetic change and  Bjorkman  1980;  Bradshaw  (Snaydon 1978)  1965).  Subsequently,  and acclimation (Berry community  composition  stabilizes and oscillates in response to such factors as management, grazing and fertilization  (e.g.  Jones  1933a,b,c,d),  and  climate  (Garwood and Tyson  1979;  Kleter 1968).  2.4.2. Disturbance  Because the  distribution of dung in the pastures  Poisson distribution, all the dung.  Molehills,  distribution.  This  consequences  to  models.  The  distributions,  on  the  plants have other  hand,  distribution can plants  growing near  distribution each  arise  may  having  a  different  random, usually Fitting a  an equal chance of being affected by generally from  molehills  result  is  from  a  fitted number  are quite a  mean,  a of  that  binomial  models  different  mixture  such  negative  of  and  under these  several  the  the  means  Poisson are  a  2 continuous variate with a X  distribution (Bliss and Fisher 1953). The negative  binomial distribution is also related to the Poisson such that, in some cases, data fit  a  Poisson  at  low  population densities  and a negative  binomial at higher  densities. This is not simply the case with the dung and molehill densities seen in  Table  2-2  because  some  higher  densities  fit  the  Poisson  and some lower  densities fit the negative binomial. A negative binomial distribution may also arise from non-random contagion. If a distribution increases  is  the  unimodal chance  and  the  "presence  of  one  individual  in  a  division  of other individuals falling into that division, a  negative  Patterns of disturbance and their influence on botanical composition / 51 binomial will  fit  best" (Student  1919,  in Bliss  and Fisher  1953). A  negative  binomial can result then from a mixed or compound Poisson distribution without contagion, or from contagion which changes the probability of the following events (Bliss  and Fisher  then  the  chances  1953).  If the  distribution results  of being killed are  from  approximately  the  a compound Poisson, same  throughout  the  pasture but if it results from contagion, plants growing near molehills are more likely to be killed than those growing further away.  2.4.3. Interaction of vegetation a n d disturbance  Each  species responded  differently  to  the disturbance pattern.  roughly divided into three groups: increasers, decreasers  Species may be  and indifferent, although  there are exceptions within groups. The response of some species (e.g. Taraxacum officinale, Poa compressa) to  disturbance  years,  of the  or  seasons,  or parts  was  growth  to  increase  cycle  but  and to  only  in certain  decrease  in  others.  External factors such as climate (e.g., a wet spring or summer), or management (e.g., intense grazing in summer or fall) influence species response. response  have  been described but the mechanisms  generating  Patterns of  the response  are  still to be investigated.  The  pattern  of response  had no  grass  (Dactylis glomerata), a  tillers  (Agropyron repens), and a  {Taraxacum officinale) were  correlation to  rhizomatous  all  clonal  found  to  grass dicot  morphology.  A  which usually that  increase  spreads with  by  robust tillering exists  as  single  apomictic  disturbance.  seed  Similarly,  Holcus lanatus, a closely tillering, clumped grass which only occasionally  spreads  Patterns of disturbance and their influence on botanical composition / 52 by  stolons and, Trifolium repens a  clonal, stoloniferous  and capable  of rapid  expansion and independent existence of ramets were both found to decrease with disturbance.  The  nature  of  the  disturbance  also  varies.  Dung  is  not  only  an  agent of  mortality but also acts to fertilize, releasing nutrients to surrounding plants. It also affords protection, because (Norman and Green  cows rarely graze within 30cm of fouled areas  1958). Many  seeds can survive passage through the cow  and subsequently germinate (Hutchinson 1979). Dung may serve to disperse seed over long distances from plants both within and outside a pasture.  Molehills have a less complex interaction: during formation plants are uprooted and  smothered.  molehill  building  In  addition to  also  brings  freeing  up  a  areas  seed  for  colonization,  source  different  the  from  process  that  of  of the  surrounding sward (Jalloq 1975). Many  of these seeds have  come  so that not only are different species  from  represented  a temporally distant but  also  different  source  genotypes,  containing  an  been buried and  "evolutionary memory"  (Harper 1977) of the genetic make-up of the pasture.  The foregoing discussion of the static  patterns,  i.e.  increased  results  abundance  has  inferred a dynamic response  was  achieved  from  through colonization or  invasion of open space. This is not necessarily the only mechanism. For example, Taraxacum officinale can invade only by seed but its seedings rarely survive on molehills  (R. Parish,  unpublished data),  however,  its  long  tap  survives burial so that it persists while other species are removed.  root frequently  Patterns of disturbance and their influence on botanical composition / 53 Static pattern need not have a dynamic corollary. There may exist a group of correlated patterns of which only two have been related. For example, Holcus lanatus proliferates in areas where water is abundant and is often found where water tables are high (Beddows  1961; Watt and Haggar 1980; R. Parish, pers.  obs.). In winter, if the water table in these areas is sufficiently high to exclude moles  during  their  very  active  period,  October  to  March,  high  abundance  associated with lack of disturbance may not represent exclusion from disturbance sites but a correlation with a third factor, high water levels.  RothwelPs (1977)  data, however, argues against this interpretation. He found that when H. lanatus was  present in the sward it was  significantly  lower in abundance on adjacent  molehills which suggests that it was unable to colonize these disturbances.  There are a number of limitations to studies using correlation of patterns. The coincidence  of patterns  is  not  sufficent  to  imply causation.  Patterns  may be  linked by other correlated factors leading to dangers in inferring mechanisms or processes based on changes in patterns. Some of these problems could have been circumvented by an experimental or manipulative approach. One approach would have been to disturb repeatedly certain areas of the pasture. Experimental work such as were  that of Sousa  (1979) in the  intertidal showed  eliminated by high disturbance rates  that  and ephemerals  perennial species  came  to dominate.  Another approach would be to remove moles and thereby eliminate a source of disturbance in order to compare which species dominated in their presence or absence.  If it had been feasible  to create  artificial  disturbances throughout the  pastures  and manipulate density,  the analysis of subsequent vegetation  response  would have been unlinked to correlations with groups of other factors. In the  Patterns of disturbance and their influence on botanical composition / 54 following chapter, I examine invasion of molehills and dung pats in an effort to resolve some of these issues.  3. I N V A S I O N  OF DISTURBANCES  3.1. I N T R O D U C T I O N  Pastures are subject to a variety of disturbances of various magnitudes such as drought,  erosion,  disturbances diversity  in  mounds,  grazing,  scrapings  and  create spatial heterogeneity and thus influence the  community.  dynamic mosaic opportunistic  animal  Disturbance  of small to large,  invaders  (Piatt  1975;  is  currently  1985)  viewed  then  result  establishment  in  an  increase  in  species  as  providing a  can be utilized by  or be recolonized in a  more or less predicable pattern of succession (Begon et al. may  Such  species and genetic  short-lived patches that Loucks et al.  droppings.  diversity  1986). Disturbance by  permitting  of opportunistic species that cannot compete in the  closed  the  sward  and increase genetic diversity by providing 'safe sites' for seed germination and establishment  of  both  colonizing  and  established  species.  Grubb  (1977)  suggested that the formation of gaps may be essential to the maintenance  has of  species richness in communities.  Plants are generally considered to be immobile organisms but the growth form of many  pasture  species  allows  movement  through  space  horizontally. This mobility permits the individual to sample environment  and  allows  colonization  of  gaps  within  the  both  vertically  and  a spatially variable sward.  The  relative  contribution of new individuals compared to the expansion of old individuals will alter spatial heterogeneity and affect community composition. If species differ in their ability to colonize disturbances, the community will reflect those differences.  55  Invasion of disturbances / 56 The  pasture  environment  provides  different  challenges  to  component  species.  Typically, pasture species are perennials which have the ability to persist and expand  within  the  sward.  Superimposed  upon this  are  numerous  gaps  which  provide an environment where the ability of a species to colonize and multiply rapidly is a more important element of success.  In this study, the colonization of individual gaps created by molehills and dung pats was monitored along with associated changes in the botanical composition of the surrounding sward. The objectives invasion  of  disturbance;  (ii)  to  were (i) to determine patterns of species  determine  replacement; and, (iii) to compare the  patterns  of  species  persistence  and  abilities of species to invade both open  space and closed sward.  3.2. METHODS  3.2.1. Field Methods  The  plots  described in Chapter  Ten plots  were  randomly chosen in each pasture for each of two survey periods (August  1983  to May 1984,  and August 1984  2 were  used in this  study.  to May 1985). Within each plot, one recently  formed molehill and one recently deposited dung pat were selected. A .5 x .5m quadrat, subdivided into 100 systematically  arranged points was centered on the  molehill or dung pat and the plant species rooted directly under the (crosswires  100 points  on the quadrat) were recorded. Depending on the availability of the  pasture, surveys could begin in August to October and were repeated monthly  Invasion of disturbances / 57 until the molehill or dung pat was completely covered with vegetation  or until  May when the grass was too long to record accurately by this method.  The method required repeated Because  there  were cattle  measurements  and mowing  in accurately  equipment  in the  relocatable  quadrats.  pastures,  permanent  stakes could not be used to mark quadrats. Instead, acrylic paint was applied to the ground at the four corners of the quadrat and this allowed relocation of the quadrat, in subsequent surveys to be accurate within a few millimeters.  3.2.2. A n a l y t i c a l methods  Transition  probabilities  were  calculated,  using  the  CROSSTABS  procedure  of  SPSS.*, from the monthly record of species at each point in the quadrat. Data were pooled over the  10 molehills and 10 dung pats per pasture and analysed  separately by pasture and by year.  A  model developed  species  replacement  describes  by de Jong and Greig over  the  study  (1984) was  period deviated  used  determine if  from random. The model  the behaviour of First-order Markov chains where transitions in which  the identity of a species at any one point remained the same were  to  excluded  from  the  analysis.  The  removal  of  (self-transitions)  self-transitions  obviated  distinguishing between individuals of the same species occuping the same space at different  times.  To test the  overall pattern of species replacement,  a program  was written by J. Emmanuel (University of British Columbia, pers. comm.) based on a maximium likelihood fitting algorithm developed by de Jong et al. (1983). A  Invasion of disturbances / 58 second  program  developed  written  by  K.  Pollson  (pers.  comm.),  based  on  algorithms  by de Jong and Greig (1985), tested for randomness in pattern of  replacement eliminated  of  each  species  pair.  Rare  from  the  analysis  because  species,  a  less  than  minimum expected  2% cover, value  of  a  given  were  3  was  specified for the calculation of chi-square values.  3.3. RESULTS  3.3.1. Species persistence: self-transition  Species  vary  Persistence,  in  their  ability  to  hold  onto,  or  persist  in,  space.  defined as the percentage of a given species, present in the initial  survey, that occupied the same point in the sward throughout each nine month survey period, ranged from 0 to 100% (Table 3-1). The variability of a species' ability  to  persist  entire  range  was  within  large; for example,  the  same  field  and  pasture, disturbance type and year was was  calculated to determine if average  species  persistence  differed  in  different  Ranunculus year.  acris, encompassed  Species  persistence  within  the a  ranked and Friedman rank correlation ranks were years,  the same.  among  The pattern of  pastures,  and  whether  around molehills or dung pats (p<0.001). The resultant groupings from multiple range tests of the rankings suggested that persistence 1984  than  1983,  i.e.,  fewer  species in  1984.  When the  was generally higher in  plant parts died and were percentage  replaced by different  of species persisting  were  pooled over  disturbance, pasture or year, only differences in rankings between years emerged significant  (p<0.001)  and  ability  of  species  to  persist  showed  no  significant  Table 3-1. Percentage of species that persist throughout molehills and dung pats. Included is the percentage of the 1977 p a s t u r e molehill dung Species  Dactylis glomerata Holcus lanatus Lol ium perenne Phleum pratense Poa compressa Agropyron repens Trifolium repens Taraxacum officinale Ranunculus acris Moss Molehi1 1 Dung Rank  t  1958 p a s t u r e molehill dung  1983 5 1.8 42 . 5 31 . 7 64 .3 34 .0 2.7 29 . 1 33 . 3 45.. 1 15..0 21 .6 0..0  1984 40..3 66 . 1 47 . 7 1 1. 1 29..0 7 .7 48 . 1 48..9 60,.0 5.,4 27 .5 0..0  1983 68..6 44 .4 41 .5 . 60..0 47 .2 . 2 .4 45,,2 45. 8 0..0 1 1 .4 . 4..5 16..4  1984 58,. 1 70..9 54,.8 50..0 37., 1 6 .7 . 49, 6 55..6 57,, 1 1 1 6. 0..0 5.8  1983 35 . 7 9 .3 3 .9 31 .3 9. 3 9 .8 21 .4 15 .6 0 .0  4  7  5  8. 5  Friedman Test S t a t i s t i c = 49.31, Multiple range test of ranks 12=11=10=8.5=7 / 1 12=11 2 12/3  p<0.0001  the nine month study period in the sward around molehill and dung pat that remains uncolonized. 1939 molehill  1983 70 .9 32 .2 33 .9 72 .0 53 .5 41 .0 52 .6 81 . 3 60 •0 45 .8 15 .4 26 . 5  1984 78 .0 29 .8 20 .0 55 .6 36 .8 21 .6 55 .0 51 .9 50 .0 15 .9 23 .6 22 .5  1983 33 .3 , 16 .8 33,.6  30 .9 0 .0  1984 48 .8 26 .2 32 .3 52 . 3 22 .0 4 .8 61 .9 48,. 3 30,.0 12 ..8 44 .8 0..0  2  6  11  10  3  -  pasture dung  1984 76 . 7 40 .4 36 .8  -  -  27..7 0,.0 40,.7 37..7 100 7.,7 26 . 8 4..0  41 .7 , 30 .O 46,.6 55 . 2 42 ,.9 13., 3 27 .2 , 0,,0 8 ,5 .  1983 31 .6 5. 1 25,.6 0 .0 14,.4 4 .3 19,.8 34, 8 0..0 0..0 5..7 1 1 ,4 .  1984 83.0 51.3 60. 5 60.0 53.7 30.8 55.2 54.5 71.4 15.4 4.8 4.3  1  12  Invasion of disturbances / 60 differences between pastures nor disturbance types.  3.3.2. Patterns o f species  replacement Monthly transition probabilities  Tables of species transitions  were  generated  (using the  CROSSTABS program)  from one monthly survey to the following one, and from the initial to the final survey (see Appendix 1). Each observed value was expressed as a percentage of the  total row count so that the  tables  sum to  100% along the  rows.  Each  percentage value can be considered to be the probability of the replacement of the row species by the column species. Of the 22 species found on or around disturbances,  not  all were  sufficiently  replacement. The values generated probability examples).  of  transition  for  abundant to  be consistently  involved in  were used to calculate the average monthly  each  species  (see  Table  3-2  The diagonals are the probability that the same  for  representative  species remains in  the same location from one survey period to the next (self-transition) off-diagonals  are  the  probability of  its  replacement  by  probability of persisting from one survey to the next was  another  and the  species. The  very high for most  species, but this probability is dependent on the time interval between surveys. For  example, in 1983, Dactylis glomerata, in the sward around molehills in the  1977 pasture, had a 87.9% (Table 3-2) probability of persisting from one month to the  next  but a  51.8%  (Table 3-1)  probability of persisting over  the nine  month survey period. This analysis also estimated monthly colonization rates on disturbances, for example, because  dung in 1983  in the  1977  pasture had a  Table 3-2. (a)Average monthly transition probability of a row species being replaced by a column species for eight common species on and around molehills, and the average monthly probability of molehill colonization by column species in the 1977 pasture during 1983-1984 survey period. Molehi1 1 Molehi11 Dactylis glomerata Holcus lanatus Lolium perenne Phleum pratense Poa compressa Agropyron repens Trifolium repens Taraxacum officinale  70. 6  3 .7 3. 0 5. 3 5. 6 3 .3 5. 9 6. 3 3. 7  Dactyl is glomerata 1. 5 87 .9  0 .5 1.8 0 .0 2 .2 3,.4 1. .1 1.6.  Holcus 1anatus 1.0 . 0 .8 83.. 1  1 .8 0 .0 0. 8 2 .0 . 1. . 1 0..0  Loli um Phieum Poa Agropyron Trifol ium Taraxacum perenne pratense compressa repens repens officinale 2.6 6,.9 1, 3 6.. 1 3.. 1 3..6 2..0 2,.3 0.4 0..6 0..4 0..8 4 .4 1..0 2 .5 3.. 2 0.3 0..3 76..9 7.. 5 1..4 0.8 2..6 0..4 0.5 89. 9 0..0 2 .5 , 0 .0 0 .5 2 .. 1 1 .4 3..9 82..4 2..0 0. 4 67. 2 0.5 8.. 3 7 ,4 , 2 .5 . 0. 0 4 .6 2 .7 . 0. 8 79. 1 0.9 0. 8 6 ,6 1 .6 . 84.8 2 .5 . 2.,2 0. 0.  (b)Average monthly transition probability of a row species being replaced by a column species for eight common species on and around dung pats, and the average monthly probability of dung colonization by column  species in  Dung Dactylis glomerata Holcus lanatus Lolium perenne Phleum pratense Poa compressa Agropyron repens Trifolium repens Taraxacum officinale  the 1977 pasture during 1983-1984 survey period. Dung Dactyl is Hoicus Loli um Phieum Poa glomerat a 1anatus perenne pratense compressa 1. 1 3 .2 75 .4 2 .0 4 .7 0..4 1. 2 1.9 1.5 1.5 90..8 0..0 1.5 85 .3 1.7. 4. 2 0..3 0.,5 1 .9 1.0 . 1. 5 82..8 5..8 0., 1 4 .9 0..0 0..0 92. 7 0..0 0 •0 1.. 1 1 .9 3. 6 85. 6 0 .8 0. 2 2 .5 0..6 3 .8 6..3 5..6 0. 0 3..8 1 .8 2 .5 2,.2 0.. 3 0. 1 2 .. 4 1.4 0..5 0..5 0 .0 0. 0  Agropyron Trifolium T araxacum repens repens of f ic i naU 2 .4 7. G 2.0 0.4 0..0 1 .5 0.7 3. 5 0..5 0. 9 2..9 1 .0 0.,0 2..4 0.0 0..9 2..0 0.9 2.5 58. 8 9 .4 86 .0 0.8 0..3 4 .8 . 89.5 0.,0  Invasion of disturbances / 62 75.4%  probability of remaining as dung, about 24.6% of dung was  colonized each  month. Transition probabilities over the survey period  Values  from  the  transition  probability  matrices  (Appendix  1)  were  used  to  estimate the probability of replacement of one species by another over the survey period from  fall to spring in a given year. Tests for randomness of the pattern  of replacement both  showed that on both disturbance types in all three pastures, in  years, species replacement  strong  pattern  was  to the replacement  the replacement  pattern  of each  not random  process  pair  (p<0.0001) and there was  a  (Table 3-3). Test for randomness in  of species, e.g. the replacement  of  Lolium  perenne by Trifolium repens, or the replacement of Holcus lanatus by Agropyron repens, showed that in almost all cases the pattern was non-random.  To each  determine survey  pattern and  if the pattern was  of replacement  other species at  constant over time, contingency table analysis of the monthly  of replacement  was  used. Results from  excluding self-transitions  Colonization  of one species by  are  patterns on molehills  the study period. In the sward  shown  in Tables  and dung around  analyses including 3-4  pats changed  and over  self-transition  3-5 time  respectively. throughout  disturbances, some species, such  as Poa  compressa, were replaced by different species at different times of the year whereas others, such as  Dactylis glomerata, had the same pattern of replacement  each  some  month.  In  fact,  species,  D.  glomerata, Taraxacum officinale and  Ranunculus acris, in particular, had such strong patterns of persistence that they  Invasion of disturbances / 63 Table  3-3. Deviations from random of replacement patterns of common species, growing on and around molehills and dung pats in the three pastures.  Year 1977 Pasture 1983 1984  1958 Pasture 1983 1984  1939 1983 1984  Disturbance  *2  Significance  Molehill Dung Molehill Dung  214.5 366.3 211.5 203.8  p<0.0001 p<0.0001 p<0.0001 p<0.0001  Molehill Dung Molehill Dung  285.9 346.2 179.7 237.4  p<0.0001 p<0.0001 p<0.0001 p<0.0001  Molehill Dung Molehill Dung  207.9 390.7 271.1 471.5  p<0.0001 p<0.0001 p<0.0001 p<0.0001  Pasture  Table  3-4. Species showing differences in the monthly included: * - p<0.05; ** - p<0.01; *** - p<0.001; insufficient replacements for analysis.  1958  1977 p a s t u r e molehill dung Species  Dactylis glomerata Holcus lanatus Lolium perenne Phleum pratense Poa compressa Agropyron repens Trifolium repens Taraxacum officinale Ranunculus acris Moss Molehi11 Dung  1983  1984  1983  1984 ***  1983  of  replacement  when  pasture  molehill **  **  pattern  1939  dung  1984  1983  1984  *** ***  1983 **  * ***  *** ***  1984  *** ***  are  pasture dung  molehi11  * ** ***  self-trans 11ions  **  1983  1984  *** ***  ** *** *** *** ***  ***  *  ***  ***  ***  ** *** ***  ***  ***  ***  Table 3-5. Species showing significant differences in the monthly pattern of replacement when sel f-transi11ons are excluded: * - p<0.05; ** - p<0.01; *** - p<0.001; - insufficient replacements for analysis.  1977 p a s t u r e molehill dung Species  Dactylis glomerata Holcus lanatus Lol ium perenne Poa compressa Agropyron repens Trifolium repens Moss Molehi11 Dung  1983 ** ***  1984  1983  1958 molehi11  1984  1983  1984  ** **  ***  pasture dung  1983  1939 p a s t u r e molehi11  1984  1983  1984  1983  ** *** ***  **  ** *** ***  **  dung  1984  ***  ***  ***  ***  cn  Invasion of disturbances / 66 had  an insufficient number of replacements by other species to allow analysis.  3.3.3. Colonization  In  o f disturbances  general, about 12 species invaded disturbed sites and the number of species  was not significantly different between dung and molehills, pastures nor year of survey. Seedling recruitment  Very  few  seedlings were recorded on disturbed sites throughout the  even fewer became established  year and  (Table 3-6). Trifolium repens seedlings were found  only in late April and May on molehills and decayed dung pats. Seedling have not been  observed  in the  sward in the  Aldergrove pastures.  Seedlings  of T.  repens were marked and their survival followed over the summer through mowing and  later  grazing.  August ranged from  Survivorship from 2% to  45%.  In  germination 1984,  in  spring to  both a greater  the  following  number of seeds  germinated and seedlings survived (Figure 3-1). Colonization patterns  To  compare the ability of different species to invade disturbed sites, a ratio was  calculated of the percentage this  final percentage  on the  disturbed site divided by the initial  abundance around it (Figure 3-2). Percentage  ratio to standardize  differences  in size between  abundance was used in area  disturbed  and that  Invasion of disturbances / 67  Table  3-6. Number of seedlings per square metre found and dung pats) in the three pastures.  Species Lolium perenne Poa compressa Holcus lanatus Alopecuris pratensis Agrostis alba Taraxacum officinale Ranunculus acris Rumex acetosella Plantago lanceolata Cerastium vulgatum Medicago lupulina  on disturbances (molehills  1939 pasture  1958 pasture  1983  1983  1984  1984  3.80 1.32 10.0 0.68  3.8  1.92 1.60  1.32 2.68 2.0 1.32 0.68 2.0  1977 pasture 1983  1984  7.92 0.92 0.92  1.16 0.56 0.56 1.16 0.56  1.08 0.98 0.48  Invasion of disturbances / 68  Figure 3-1. Average number of Trifolium repens seedlings per molehill or dung pat germinating i n A p r i l and surviving until A u g u s t i n each pasture for two consecutive years. initial number per disturbance i n A p r i l ^  final number per disturbance i n A u g u s t  69  Invasion of disturbances / 70  Figure 3-2. Invasion ratio of species final percentage abundance on disturbed sites divided by initial percentage abundance around the disturbance. If the ratio is >1, species are more abundant on the disturbed site than around it; i f the ratio is <1, species are less abundant than in the surrounding sward.  NVA90N RATIO  M A S O N RAT©  NVA90N RATIO  NVA90N RATIO  M A S O N RATIO  Invasion of disturbances / 74 remaining  around  abundant  on  the  disturbance. If the  disturbed  areas  than  in  ratio the  repens, Poa compressa); if the ratio was disturbed  sites  than  in  the  surrounding  was  >1,  surrounding  < 1,  the  the  species was  more  {e.g. Agropyron  sward  species was  less abundant  on  {e.g. Dactylis glomerata, Holcus  sward  lanatus).  3.3.4. Changes in botanical composition  The  pastures  are characterized by  species decreased in  this survey  changes in species abundance over time: some  whereas others increased. The  was  governed by  its behaviour  change in abundance of a species in the surrounding  ability to colonize disturbed areas. Figure  3-3  type,  in the  percentage  abundance  of a  species  area, in the sward around the disturbance the  sampling  final survey or tillers  quadrat. has  The  change  was  final  pasture survey  and on  disturbance  the  disturbed  in both, i.e. the area covered  in percentage  abundance  from  the  by  initial to  been calculated from the change in number of rooted individuals  for both  the  area around  (disturbed site plus surrounding areas  and  shows, by  sward plus its  always  100%  the disturbance  sward). Obviously  increase over  initial  and  the entire area  sampled  percentage change on disturbed  abundance. For  some species, such  as  Trifolium -repens and Agropyron repens, if there were a decline in abundance  in  the sward, an  overall increase could still occur because of proportionally higher  invasion onto the disturbed site.  Invasion of disturbances / 75  Figure 3-3. Percentage abundance of species at the final survey on molehills and dung pats, in the surrounding sward, and over the entire 0.25m^ quadrat. The percentage change in composition from initial to final survey has been calculated for the surrounding sward and the entire quadrat area. The percentage change in species abundance on the disturbed area is either zero (no invasion) or 100%.  disturbance surrounding sward quadrat  average  Invasion of disturbances / 83 3.4. DISCUSSION  3.4.1. Species  persistence  In general, the probability that a species would be found at the same point over the  nine  month  heterogeneity  study  period was  was high. Persistence  low  (Table  3.1),  suggesting  serves as an estimate  that temporal  of genet survivorship  for species such as Ranunculus acris and of ramet survivorship for clonal species such as  Taraxacum officinale which grow as independent individuals. For other  species (i.e., tillering grasses, Trifolium repens and Poa compressa), the relationship between persistence release because  of of  space  and survivorship is may  difficulties  result  from  obscured. This  death  of  parts,  distinguishing between different  was  not  not  only  because  individuals, but  individual  plants  also  of the  same species occupying the same space in successive time periods.  The greatest death risk to established plants appears to occur during the period of most active growth (Sagar 1970; Sarukhan and Harper 1973). The timing of this period varies with the species under consideration: most grasses grow faster in the spring, April to May, at a temperature optimum between 13 and 18°C, whereas  Trifolium repens grows  faster  later  in the  season  at  a temperature  optimum around 24°C (Spedding 1971). The estimates of persistence  reflect the  intense growth/death phase for both T. repens and grasses, because  the survey  usually ran from August, when T. repens was growing rapidly, to April, when the grasses were growing vigorously.  Invasion of disturbances / 84 3.4.2. Species replacement a n d invasion  Potentially, any species could be replaced by any of the species in the pasture but  only  the  most  abundant  processes. Replacements  were  consistently  involved  in  the  replacement  were not random, i.e., space capture by a species was  not proportional to the abundance of that species in the pastures. The tests for non-randomness were based on a first order Markovian model. This assumes that the system has no memory, i.e., that the current state is determined only by the  previous  one.  Over the  study  period, the  sample  size was  inadequate  to  include sufficient multiple transitions to test for higher order effects in which the probability of a species replacement would depend on states prior to the previous one.  Higher  temporary  order  seasonal  processes  would  advantages  and  not  be  improbable  pre-empted  space  if  some  that  was  species  had  subsequently  relinquished to the prior occupant.  Thorhallsdottir (1984), using a Markovian model to describe species replacements in a pasture in Wales, found that between-season replacements were non-random but between-year pasture.  She  replacements  concluded  that  were  porportional to  temporal heterogeneity  species was  abundance  in  the  very fine-grained but  species became thoroughly mixed over longer periods. In addition, she suggested that the majority of non-random interactions  took place during periods of rapid  growth and expansion.  The  considerable variation among the ability of species to invade disturbed areas  appeared  to  be  linked to  morphological  form.  This finding differs  from  the  Invasion of disturbances / 85 conclusion of the previous chapter. Pattern detected at one scale of observation 2 2 (0.25m ), was lost at another scale (25m ). In general, rhizomatous {Agropyron repens and Poa compressa) and stoloniferous disturbed  sites  in  proportionally  higher  {Trifolium repens) species invaded  abundance  than  they  had  in  the  surrounding sward. Tillering grasses {Dactylis glomerata, Holcus lanatus and Lolium perenne) were less abundant on disturbed sites than in the  surrounding sward.  There were exceptions to this: Phleum pratense, a tillering grass, invaded dung although it was a poor invader of molehills.  The inability of Phleum pratense to invade disturbed sites in the  1939  pasture,  but its ability to invade in the 1977 pasture appears to be based on a genetic difference in the P. pratense seeded in the pastures. Prior to approximately the 1960's in the pastures  Fraser  but this  Valley,  practice  was  an old land race of P. pratense was discontinued.  sown in  The more robust variety  'Climax'  used in the mix after the early 1960's (Richardson Seed Co., pers. comm.) has the ability to invade disturbed areas.  The  apparent  artefact  ability  of  because no new  Taraxacum officinale to individuals established,  invade  disturbed  sites  rather, mature plants  is  an  emerged  from under the disturbance. Emergence was often as long as three months after the disturbance had occurred. This agrees with Roth well's (1977) findings that T. officinale could survive burial. The high abundance of T. officinale associated with disturbance is due to root regeneration on sites on which species less tolerant of burial are removed.  Invasion of disturbances / 86 The invasion of disturbed areas is a non-competitive process whereas performance in the sward involves both non-competitive  and competitive interactions.  of  order  the  closed  horizontally, density  nature  another  of  the  must  however,  in  relinquish space.  independent forces  replacements,  sward  such as  suggests  frost  that  for  an  Release  of  individual to  Because expand  space may be due to  or trampling. The non-randomness of  replacement  was  governed  by  species  interactions not merely proximity. Performance on disturbed sites has been related to  species mobility and investment  in horizontal growth. There is no a priori  reason to expect these characters to be linked to performance in the sward. But some species show similar performances increased  in both situations,  e.g., Poa compressa  in both whereas Holcus lanatus and Lolium perenne, which  usually  failed to establish well on disturbed areas, also decreased in the sward.  In contrast, some species had divergent responses. Phleum pratense increased in the  sward  enrichment  but  invaded  facilitated  dung  lateral  only,  growth,  suggesting  that  a  consistent  finding  a  response with  to  nutrient  the  fertilizer  treatment in Chapter 5. Its lack of response on molehills may have been due to lower  moisture  availability  (M.  Pitt,  pers.  comm.).  Dactylis glomerata  also  increased around dung but it was still a poor invader of disturbed sites.  Agropyron repens was  the only species showing  an ability to invade disturbed  sites but generally decreasing in the surrounding closed sward. This suggests that disturbance may provide a refuge  serving as the primary agent in determining  the pattern of abundance of A. repens in the pastures.  Invasion of disturbances / 87 3.4.3. Recruitment a n d colonization  The  role of dung and molehills in providing an entry into the community for  new species, or even for sexually produced progeny of established negligible.  Colonization  by  seed  occurred  consistently  only  in  species, was two  species,  Ranunculus acris and Trifolium repens. R. acris germinated both on disturbed sites and  in the  sward but T. repens was  restricted to disturbed areas.  Similarly,  Turkington et al. (1979) observed T. repens seedlings on molehills but not in the sward,  and seed sown  whereas  into the  on cleared plots,  sward failed to establish  (0.89% germination)  seed germinated (39% germination). It appears  that  sexual propagation of T. repens in pastures is obligately linked to disturbances.  In  contrast to these findings, Piatt (1975) described a guild of fugitive species  specialized to exploit badger mounds in tall-grass prairie. Other North American grassland studies have confirmed a role for disturbance in providing suitable sites for  germination  and  survival of  annuals  (Allen and  Knight  1984),  or  taxa  differing from the surrounding area (Polley and Collins 1984; Loucks et al. 1985). Differences  between  these  studies  in  native  prairies  and  in  the Aldergrove  pastures could be used to support an argument for evolution of specialists based ort native systems being of sufficient longevity to foster such niche differentiation. Additional  support for this  interpretation can be derived from  Belsky's (1986)  observations in the old and relatively pristine Serengeti ecosystem that  colonization  of  disturbed  sites  was  dominated  by  seedlings  (Whyte and  1974)  not by  vegetative expansion. Disturbances often present different micro-environments which provide  opportunities  for  differential  regeneration  of species  in the community  Invasion of disturbances / 88 (Grubb  1977).  Badger  Collins  1984)  retain  physical  characteristics  mounds higher of  (Piatt  soil  1975)  and buffalo  moisture  molehills  were  than not  wallows  (Polley  surrounding areas.  measured  in  this  and  Although  study,  it  is  assumed that they have a characteristic pattern of higher soil temperatures and lower  moisture typical of mounds in northern latitudes  (cf.  The presence of species that can exploit this heterogeneity  Spittlehouse  1985).  probably depends on  the frequency of disturbance in addition to the time dimension. Time could be either of evolutionary or ecological length. Evolutionary time would permit the development of mechanisms to exploit heterogeneity an  ecological  mechanisms  time to  scale reach,  would  allow  for  influx  example,  of  and differentiate species, and species  previously  with  limited dispersal  glaciated  or  formerly  non-agricultural sites.  One factor tends to obscure comparisons between this study and most on native grasslands. Molehills and dung pats average about 0.07m  2  in area whereas most 2  of the previous studies were of larger disturbances (0.2 - 2.0m ). Size of opening can strongly influence patterns of germination and survival (Davis and Cantlon 1969;  Miles 1974;  relationship  Goldberg and Werner  between  germination  and  1983). Miles (1974) found an inverse survival 2  proportionally more seeds germinated in 25cm  in  gaps  of  different  sizes;  gaps but survivorship was highest  2 in  2500cm  gaps.  Rapp and Rabinowitz (1985) using artificial  disturbances of  2 about 27cm  found no difference in seedling survival in disturbed versus control  areas. In addition, a plethora of evidence exists to support differentiation among species in pre-germination requirements and conditions for survival and growth. Although  Grubb  (1977)  has  argued  that  these  differences  in  regeneration  Invasion of disturbances / 89 requirements heterogeneity  maintain  species  richness  in  communities,  it  appears  that  provided by molehills and dung pats is relatively unimportant to  species richness in these pastures.  The finding that space made available by disturbances (molehills and dung pats) is  generally  filled by the  expansion  Sousa's (1984b) observations.  of existing  individuals is  in accord with  Sousa observed that small patches were colonized  rapidly by encroachment of adults along the edges and attributed the high rate of closure  to the  closure, i.e.,  larger perimeter to  area ratio of small disturbances. Rapid  within a year, of small disturbances may restrict colonization by  species without persistent seed or long distance dispersal mechanisms. Because of the restriction of recruitment of new individuals these disturbances act primarily as agents of mortality removing individuals but providing limited opportunities for new individuals to enter the population. Individuals of a species are eliminated at random but replacement is by individuals (species) with specific characteristics.  4. INFLUENCE OF SPECIES REMOVAL ON INTERSPECIFIC INTERACTION  4.1. INTRODUCTION  Over the  past twenty-five years,  perturbation  experiments  have been used  increasingly to determine relationships between species and analyse the structure of communities, (e.g., Sagar and Harper 1961; Pinder 1975; Gross 1980; Fowler 1981; Hils and Vankat 1982). One type of perturbation experiment involves selective removal of one or more species from the community and measurement of the magnitude of the response by the remaining associated species.  Early removal experiments measured the response of target species to the mass removal of associated groups of species. For example, Sagar and Harper (1961) found that vegetative expansion and seedling establishment of Plantago spp., and Putwain and Harper (1970) for Rumex spp., was restricted by grasses rather than by other dicotyledenous species. Later studies aggregated the target species to examine community structure. For example, Pinder (1975) removed dominant grasses in old-fields and detected a diffuse response of increasing biomass in the remaining subordinate species. Also in old-fields, Allen and Forman (1976) found removal of tall species caused the greatest changes in the response of remaining species. The response of species to removal was non-reciprocal; for example, few species could invade patches from which Potentilla simplex had been removed but P. simplex was able to invade areas made vacant by the removal of other species. 90  Influence of species removal on interspecific interaction / 91 Removal experiments were used to determine pairwise interspecific interactions by Fowler (1981) in grasslands marsh-dune  communities.  and by Silander and Antonovics (1982) in coastal  Species  interaction  in  both  these  communities  were  rarely reciprocal and could be specific or diffuse. In addition, interactions varied seasonally  and  changed  when  the  same  species  pairs  interacted  in  different  environments (Silander and Antonovics 1982).  The ability of removal experiments  to determine competitive  interaction between  species has been challenged by Bender et al. (1984). They argued that except in a two-species community, the presence of a net gain in one species following the removal  of  another  is  not  a  sufficient  or necessary  condition to  demonstrate  competition. The gain reaction may actually result from interaction with a third species.  Similarly,  the  absence  demonstrate  absence  response  removal may  Bender  to et  al.  of  (1984)  of  response  competition. be  suggest  weak the  to  Because  removal of  a  or non-existent. separate  removal  may  network  not of  To avoid this of  each  necessarily interactions, ambiguity,  species  in  the  community to create a complete interaction matrix. This approach is conceded to be  impractical in  most  multispecies  communities,  not  to  mention prohibitively  expensive in time and cost.  Another consideration raised by Bender et al. (1984) is that of overlooked species. In defining a community such as a pasture in terms of botanical composition, some obvious and some inconspicuous members of the community are overlooked. Pastures, pasture  by is  definition,  enormous.  always  Even  include  if the  grazing  effects  of the  species,  whose  overlooked  effect  species on  on  the  target  Influence of species removal on interspecific interaction / 92 species are very large, the interaction is negligible if the effects of the target species  on  reasonable  the to  ignored  species  assume that  are  very  in pastures,  small  (Bender  although  the  et  al.  1984).  cumulative  effect  It  is  of all  herbaceous species on grazers is very large, the effect of each component species on the grazer is very small. Thus, the interaction between plant species can be validly studied in isolation from the consumer effect.  Ecological theory argues that coexisting plants diverge in their use of resources and thereby  reduce  or avoid competition;  Harper  (1967) called this  "ecological  combining ability". Likewise, Fowler (1981), working in an 30-year-old grassland in North Carolina stated that weak  and  approximately  "this is a community characterized by relatively  equal  component species " (p. 851).  competitive  relationships  Aarssen (1983), however,  among  all  of  its  suggests an alternative  possibility: he argues for strong competitive interactions between individual species that become more equal over time. Furthermore, he argues that if competition is an organizing force in determining community structure, there should be strong, unequal interaction in recent pastures and strong, equal interactions in older ones. The predictions of both viewpoints found  in  recently  established  should be detectable if they through  niche  experiments  divergence.  to differentiate  are the  communities.  same - strong interactions In  older  communities,  are strong but not if they  Unfortunately, between the  it  is  beyond  mechanisms  are weak the  that  scope lead  to  will be  interactions or reduced of removal equality in  interactions among species over time.  The  physical  proximity  and similar  sowing  and management  histories  of  the  Influence of species removal on interspecific interaction / 93 Aldergrove  pastures  make  them  suitable  to  test  for  changes  in  species  interactions over time. This chapter describes a removal experiment carried out using three of the common species: Lolium perenne, one of the originally seeded grasses,  Trifolium repens a seeded legume, and the weedy grass Holcus lanatus.  The design allowed an examination of the species interactions with each other and with associated species and the effects of their combined removal. If one of the  three  control the  dominant species is  removed, the  remaining dominant species could  resultant interactions. Moreover, removal of species combinations as  well as single species allows any substructuring of the community to be observed.  4.2. METHODS  4.2.1. Field methods  Four sites were selected in the 1939 and 1977 pastures. The sites were chosen to conform to the following criteria:  (i)  absence  of any mole activity; (ii) no  visually obvious discontinuities in vegetation; (iii) slopes of less than 5%; and (iv) absence of any evidence of waterlogging. The 1958 pasture was excluded from the  study because  of the difficulty in finding sufficient  sites which met" these  criteria.  The sites were staked into 2.8 x 6m blocks and further subdivided into eight one metre square plots. Around each plot was a 10cm boundary from which all vegetation  was  removed, and an uncleared zone  20cm wide was  left  between  Influence of species removal on interspecific interaction / 94 plots.  Plots within each site were assigned  randomly to one of eight removal  treatments shown in Table 4-1.  The  four sites in each  pasture  frequency in each plot was  were initially surveyed  in July  1983.  Rooted  measured by placing a 50 point, 0.5x0.5m square  grid, four times, to give a total 200 points per lm  2  plot.  Although the sites selected for the experiment were visually uniform, considerable variation  existed  consequently,  in  rooted  frequency  of  the  three  in the amount of bare ground exposed  removal  species  and  after removal (Table  4-2).  Removal of the grasses and clover began in late July and continued throughout August and September.  All plots  were cleared  of regrowth in September and  October. A screwdriver or large iron nail was used to dig out the grasses and lift the Trifolium repens stolon and roots. Care was taken to avoid disturbing the remaining vegetation possible,  plots  and loosened  were kept  soil was pressed back into place. As far as  dung-free  during the  duration of the  experiment by  removing any depositions.  In July  1984,  one year after the initial survey, the plots were re-surveyed in  the same manner as in 1983.  Subsequently, the plots were cleared of regrowth  of all three  species through  fall  survey  made  1985.  was  in  July  1984 In  and again an  effort  in spring to  1985.  minimize  The final  seasonal  and  management-induced variations, each survey was made as soon as possible  after  haymaking, although a timelag was necessary to allow regrowth of species to an identifiable size.  Influence of species removal on interspecific interaction / 95  Table 4-1. Species removed from eight plots in the 1939 and 1977 pastures.  Treatment Species removed C T H HT L LT LH LHT  control plot, no species removed Trifolium repens removed Holcus lanatus removed Holcus lanatus and Trifolium repens removed Lolium perenne removed Lolium perenne and Trifolium repens removed Lolium perenne and Holcus lanatus removed Lolium perenne, Holcus lanatus and Trifolium repens removed  Table 4-2. Percentage cover of removed species in four replicates in July 1983.  Percentage cover removed  Treatment 1939 pasture T H HT L LT LH LHT  J_ 6.5 9 16 26 54 47 48.5  _2 12 14 17 25 41 41 47.  _3 9.5 23 43 24.5 32.5 54.5 70  _4 11 29 31 18 38.5 54.5 70.5  mean 9.8 18.8 26.8 23.4 41.5 49.3 59.1  6.5 18.5 38 28 39 32.5 43.5  19 27.5 59 34 45.5 54.5 75  12.4 16.3 38.5 21.5 37.0 33.9 50.4  1977 pasture T H HT L LT LH LHT  15.5 11 30 9.5 35 26 37  8.5 8 26 14.5 28.5 22.5 46  Influence of species removal on interspecific interaction / 96  4.2.2. A n a l y t i c a l m e t h o d s  To determine the  effects of Lolium  perenne, Holcus  lanatus and  Trifolium repens  removal on the abundance of remaining species, the difference in rooted frequency between the initial 1983  and final 1985 survey was used. The model chosen is  univariate, that is, effects of all removals were tested on one respondent species at a time. Some species, for example Rumex acetosella and Alopecuris geniculatus, occurred too infrequently for statistical  analysis.  Analysis of variance computer  program 2V of BMDP (Dixon 1983), analysis with three fully crossed factors and no groupings, was used. The identification of three factors, removal of L. perenne, H.  lanatus and T. repens from each relevant treatment whether as single species  or  combination examined the effects of single removals as well as the two- and  three-way interactions.  Because  the  removed species and the  respondent  species have  different initial  abundances, an expansion rate, (R..) was calculated to standardize the response of U  species i to removal of species j. R.. is measured by using the change in rooted y frequency  of  species  i  from  1983  to  1985,  divided  by  the  original  1983  frequency, i.e. R..  ij  = (N.. - N.)/N. ij i i  (4-1)  where N.  - original abundance of species i (in 1983)  N..  - abundance of species i (in 1985) when removed y  If  species j is zero  the original abundance of a species in a plot was zero, the expansion rate  Influence of species removal on interspecific interaction / 97 was not calculated for that treatment plot to avoid division by zero. Expansion rate  (R„) was  examined in response  to the abundance of the removed species  and as a response to amount of area available for colonization.  4.3. RESULTS  4.3.1. 1939 pasture  There was little of  considerable regeneration of  Holcus  and September, adjacent  and  lanatus T.  areas.  repens  invaded the  repens  In July  Trifolium  Lolium  1985,  perenne  in July  after initial removal but 1984; however,  in August  removal plots by stolon extension  regeneration of removed species was  from  essentially  zero. In the 1939 pasture, differences between controls and treatment plots were not  significant  removal  (i.e.  of large  colonization.  p>0.05) amounts  in most of cover  Bare ground increased  follow  that  H.  lanatus  (Tables  and the  H.  4-3  and 4-4)  despite the  opening up of bare ground for  overall in the  significantly less in those from which necessarily  cases  lanatus  experimental  sites but  was  had been removed. It does not  removal sites  were more  readily colonized.  Rather, species had an extra growing season to invade bare ground created in 1983 because  H.  lanatus  from seed in 1984,  did not regenerate  after removal and did not colonize  so that little further treatment was required and no new  bare spots were made in the second season.  Neither  Lolium  perenne  nor to the removal of  nor  Holcus  Trifolium  responded to the removal of the other  lanatus repens.  T.  repens  showed a significant increase  Table  4-3. Mean d i f f e r e n c e (n=4) i n species percentage cover i n the 1939 p a s t u r e f r o m 1983 t o 1984. Year effect indicates differences in abundance averaged over a l l t r e a t m e n t s . Removal effect indicates differences i n a b u n d a n c e a t t r i b u t e d t o d i f f e r e n t t r e a t m e n t s . * - p<0.05; ** - p<0.01; *** - p<0.001.  Spec i es Agropyron Dactylis Holcus  repens glomerata lanatus  Lolium perenne Poa compressa Phleum pratense Anthoxanthum odoratum  control 2. 3 -10. . 2 -26. . 1 -7. 4 -1 . .9 0.. 3 1 .8 .  T -2..0 0..4 -26 .. 4  HT -0 .5 6 .7  L -3 .3 0 .0 -28 .0  LT 4 .8 -0 .5 2.8  LH 1 .5 0 .9  LHT  -o  .2 6 .9  -  -  -4 .6 -5 .3  2 .8 13 .8  -5 .3  18 .5  9.7  23 .7  0..31 0..34  1 .8 . 2..2  0..3 1 .3 .  3 .6 1 .8  -1 .5 3..8  0 .0 4 .0  1 .3 2 .4  -0 .4 1 .8 .  0.. 22 0.,05  -1 . 3 23 ,.7 1 .8 . -5.. 3  0 .0  -2..5 30, 8 7..8  0 .0  0.5 39 .9  5..0  -  3 .0 -5 .8  0 .8 -11 . 2  5..3 1 .0 ,  0. 0 39. 2  -3. 8  Ranunculus Taraxacum  0.. 3 -1 .,8  2..9 -2. 9  -  -  2 .9 -2 .4  -  0..0  -  -  -  year effect 0.. 59 0.. 59 0..01**  5..3 19..3  Agrostis alba Trifolium repens acris of f i c i n a l e  H 4..8 4..8  -  -  removal effect  0..73 0.,30  H  0..02* 0.,02*  LTH  , LT  co  Table  1939 p a s t u r e from 1983 t o 1985. Year 4-4. Mean d i f f e r e n c e (n=4) i n s p e c i e s p e r c e n t a g e c o v e r i n the all treatments. Removal effect indicates effect indicates differences in abundance averaged over - p<0.001. differences i n abundance a t t r i b u t e d t o d i f f e r e n t t r e a t m e n t s . * - p<0.05; ** - p<0.01; ***  year Species Agropyron  repens  Dactylis  glomerata  Holcus  lanatus  Loli um perenne Poa compressa Phleum pratense Ant hoxant hum odoratum Agrostis Trifolium  alba repens  Ranunculus Taraxacum  acris officinale  control 12..3  T -1 .5  5 .8  15 .5  -  -  0. .06  -  -  -  10.,3 0. .0 1 .5  30 .5  18 .5  47 ,.8  3 .3 1 .5  O .0  -  -2.,8 32. 5  3 .0 8..0 6,.0  5 .0 -6 .3  1 1 .0 . -3.,8  9.5 -5 . 3  0. ,38 0. .07 0. ,04* 0. . 19 0. , 25 0. ,02* 0. ,04* 0. ,01**  -  -4 .8 24 .0 4 .5 7 .5  -3,.5 22 .3  -7..5 -1 .3  0. .8 2.. 3  0 .0 4..8 -3. 8  1 .8 1 .0  4 .0 -4..0  LHT 1 .8  -  2 .5 -17 .5  -  LH 14 .5  HT 3 .0 10 .8  -6..0 -18. .0 -18, . 3 6 .3  0. .0 23. 8 2,.3 -4 .. 5  LT 13 .0 5 .5 5 .0  H 16 .5 4 .3  -4..5 23,.5 6 .0 -9,.8  -1 .0  L 5. 8 3,,5 -19, ,5  9 .3 4.5 -  0 48 7 -13  .0 .8 .5 .5  -  10.. 3 2,.8  effect 0. .23 0. .24  removal effect  HT*  HT*  vo  VO  Influence of species removal on interspecific interaction / 100 (p = 0.017) in abundance in the two year period but did not expand preferentially in either the L. perenne or H. lanatus removal plots.  Both Poa compressa and Taraxacum officinale responded to the joint removal of Trifolium repens and abundance  increased  Holcus under  lanatus (p<0.05);  this  treatment,  the  accelerated. Abundance of T. officinale in the the  experimental  period  (p = 0.006,  see  but  whereas,  decline  of  T.  P. compressa officinale was  1939 field declined gradually over  also  Table  4-4).  Ranunculus acris  abundance increased (p = 0.036) although it was not significantly increased by any treatment.  4.3.2. 1977 pasture  In July 1984,  there was little regeneration of the removed species in three of  the four replicates  in the  1977  pasture,  although Trifolium repens invaded the  removal plots in August and September. The fourth replicate was recleared and, along with the other recleared replicates, species by July  1985.  showed  little regeneration of removed  Several grasses responded to the  removal of Trifolium  repens and the combined removal of T. repens and Holcus lanatus (Tables  4-5  and 4-6). Dactylis glomerata increased with the removal of T. repens{p<0.01), H. lanatus (p<0.05)  and with  the  LHT removal  (p<0.01),  and Phleum  pratense  increased with the combined removal of T. repens and H. lanatus (p<0.05). Both Poa compressa and Anthoxanthum odoratum increased in overall abundance during the  experimental  significantly  with  period T.  (p = 0.020  repens removal  and  p = 0.004  (p<0.01  and  respectively): p< 0.001  both  increased  respectively)  and,  Table  4-5. Mean d i f f e r e n c e (n=3) i n species percentage cover i n the 1977 p a s t u r e from 1983 t o 1984. Year effect indicates differences in abundance averaged over a l l treatments. Removal effect Indicates differences i n a b u n d a n c e a t t r i b u t e d t o d i f f e r e n t t r e a t m e n t s . * - p<0.05; ** - p<0.01; *** - p<0.001.  Species Agropyron repens Dactylis glomerata Holcus lanatus Lolium perenne Poa compressa Phleum Agrostis Trifolium Ranunculus Taraxacum  pratense alba repens acris officinale  control  -16, .3 -2 .3 -5. .0 -10. . 3 -3 .. 3 2.7 7.7 27. , 3 0. 7 -8 .. 3  r H 1 .6 -0. .7 4. ,7 0 .0 0..3 -21 ..3 -13 .3 2 .7 10. 0 -1 ..3 2 .7 1 .3 0..7 - 28 .7 -1 .7 -0. 3 -7 .3 2. .3  HT  6 .7 10 .0 -  -9 . 3 5.3 3 .3 2 .7  -  3 .3 -0 .7  L  -1 .0 0.7 5 .0  LT  12.0 5.3 13.0 -  3.. 3 7.0 1 .3 . 2.7 0.7 0,.7 24. .3 1 .7 1 .7 . -8. .7 -1 .33  LH  LHT  -o .7 8.3  7.7 2.3  -  0.7 0 .0 1 .7 26 .3 1 .3 -8 .0  -  39.7 -1.7 2.3  -  2.0 -8.33  year effect  0.65 0. 24 0.13 0.07 0.07 0. 53 0.40 0.89 0.30 0.03*  removal effect  r*,LHT  r**  Table  4-6. Mean d i f f e r e n c e (n=4) i n s p e c i e s p e r c e n t a g e c o v e r i n t h e 1977 p a s t u r e f r o m 1983 t o 1985. Y e a r effect indicates differences in abundance averaged over a l l t r e a t m e n t s . Removal effect indicates differences i n a b u n d a n c e a t t r i b u t e d t o d i f f e r e n t t r e a t m e n t s . * - p<0.05; ** - p<0.01; *** - p<0.001.  S p e c 1es Agropyron Dactylis Holcus Lolium  repens glomerata lanatus perenne  Poa compressa Phleum pratense Anthoxanthum odoratum Agrostis alba Trifolium repens Ranunculus acris Taraxacum o f f i c i n a l e  control  r  H  -10. .8 3 .8 .  -21 .5  -15 .5  13..0 -25..0 -8 .. 3 4.3 2 .. 5 3.,5 17..3 3.,5 -10. 5  14 .5  10 . 3  10 .8 -14 .0 -20 .0 17 . 3 15 .8 1 .0 0 .3 14 .3 5 .3 0 .0 0 .3 31 .0 3 .0 0. 3 -7 .75 -8 . 25  HT -9.0 26.0 -  1 -10. .8 5..3 18. . 3  LT -9 .8 18 . 3  5.5 0.8 -  6.3 -4 . 25  0..07  w*,r** LHT**  -16 .0 17 .0  _  4..8 1 .5 .  17 . 3 2 .8  5..8 -3. .0 24.,0 3..5  10 . 3 0 .3  -12. .0  16 .5  LHT  21 .8  -4.5 22.5 5.5  removal effect  -8 .5  year effect 0.. 10  LH  -  1 .75 - 9 . 75  _  8 .3 1 .3 8. 5 1 .0  42 .5 0 .8 13 .8 0 .5  39 .5 3. 75 -7 .0  7 .5  -  -11.25  0..06 0.. 12 * 0. .02 0.. 14 0,.004* 0..45  H* T** HT* T***,HT*  0..0004* 0. 094 o..09  to ro  Influence of species removal on interspecific interaction / 103 additionally, A. odoratum responded to the combined removal of T. repens and H. lanatus (p<0.05).  Of  the removed  species,  only  Lolium perenne  showed  a significant  (p<0.05)  response to the removal of another species: L. perenne declined in abundance over the experimental period but its decline was less where Holcus lanatus had beeen removed. This relationship was not reciprocal: the removal of L. perenne did not affect the abundance of H. lanatus.  Among  the dicots,  the trend of declining  Taraxacum  officinale  and increasing  Trifolium repens and Ranunculus acris abundances was the same as in the 1939 field (see also Table 4-6). No dicot responded significantly to any of the removal treatments.  4.3.3. E x p a n s i o n rate  Expansion rate (R„) was calculated using equation 4-1 and plotted as a function  of the amount of Lolium perenne, Holcus lanatus and Trifolium repens removed (Figure 4-1). The response of R.. to the amount of each species removed used only single species removals to avoid the complication of differing or synergistic responses to two or more species removals. Figure 4-2 illustrates the relationship between expansion rate and the area available for colonization and uses cover removed from all four replications of all removal treatments as the independent variable.  Expansion  transformation.  rate  of  The coefficient  a  species  was  normalized  using  a  Box-Cox  of determination of the regression of normalized  Influence of species removal on interspecific interaction / 104  Figure 4-1. Expansion rate from 1983 - 1985 of species remaining in each of the four replicates after the removal of Lolium perenne, Holcus lanatus and Trifolium repens. Expansion rate at 0% cover removed is the average change in cover in the control plots.  • Lolium perenne removed A  Holcus lanatus removed  • Trifolium repens removed  1939 PASTURE  1939 PASTURE Lolium perenne  5  10  16  20  Trifolium repens  25  30  PERCENTAGE COVER REMOVED  PERCENTAGE COVER REMOVED  1977 PASTURE  1977 PASTURE Trifolium repena  0.5-1  Lolium perenne  0  5  10  15  2  0  2  5  3  PERCENTAGE COVER REMOVED  0  3  5  PERCENTAGE COVER REMOVED  o  1939 PASTURE Taraxacum officinale  PERCENTAGE COVER REMOVED 1977 PASTURE Taraxacum officinale  8  K>  15  -i 20  r 26  PERCENTAGE COVER REMOVED  i  30  38  Influence of species removal on interspecific interaction / 109  Figure 4-2. Expansion rate from 1983 - 1985 of species treatments against the percentage cover removed in each eight treatments (n=32). Coefficient of determination and regression were calculated from normalized values of the  remaining after removal of four replicates of significance of the expansion rate.  EXPANSION RATE  EXPANSION RATE ro  ^  o  _  NI  _J_ 0-41>>  CR  s-- ^  >  s  cn  K>  t>  >  on  5  3 ft  _  <o o i .  gtt s°  gs B  3  o ** C zD o  trjg  2  1_  CM-  cn -  fig  CO  _1  cn  1 1  s  B g  1  to  EXPANSION RATE  EXPANSION RATE O  0 -  *  cn  J  I  a  I  «J  £  01  OB  1  '  1  » o >  o  en -  c •  K>  O 0  1 >  CO _  > 5  s00 ,s< m 8-  >  AS  en _ cn  3cn  on  1  8co o  8 8!  m  g-  cn. 01  >c  •o o  > fe  101  s: o  1  5!  cn -  CO 0  P cn  0  —*  I8  I  cn  s-  8-  o  is B g  CO <A CD 0>  s 3  > > >  —  1  101  '  N  r» 01  ' 1  EXPANSION RATE 6  bp —I  <S is  o *  eb k>  1  1  1  o  I  EXPANSION RATE o io  Xk  I  o I  o  O  cn I  cn I  t>  8CO  co  5  CO  3D  CO 30  m  m  1  g  A |  3 a  a> o  EXPANSION RATE  EXPANSION RATE  o  cn _l_  to _JL_  fo cn _l  u I  6  co cn 1_  cn  cn  > t> > > x>  8H co co co  3  5 33  m  A 3=  I 2  III  8  "I i i s CO  EXPANSION RATE  EXPANSION RATE cn  _I_  Ol-  cn -  >  a<*-  i  _l  _L_  2 °i  888-  > >  a-  5  s$-  o  8-  a-  8as: 8-  9s 3 3  i  EXPANSION RATE 6  o  _1  Ol 1 B»  cn 1  EXPANSION RATE cn i  1  6  O cn  cn  1  >  o- >>  >>  o  1  3  3-  >  z o  >  o  >  >  >  5  >  8° \J at  m" -Ml  -  I 9. 8 § 5-  sc?  211  3a 9 8  a  a-  8  EXPANSION RATE 0  6  6  6  J. te e» * i 1 1 ;  0  k* 1  O  1  M I  EXPANSION RATE 0  * I  0 to I  0  6  to L_  0  6 I  _ J  6  0  I  I  I  I  0 I  o-  > >  5 > >  7-s 8 3  EXPANSION RATE  EXPANSION RATE i  O  _ _ l  M I  >  CO L_  >  o»  _i  a  «J  OB  as  1  1  '  '  CP _ l  >  5 t> m  P  a  i  £11  9  8 3  -si  CO  I  '  I  I  •  Influence of species removal on interspecific interaction / 114 scores on percentage cover removed is included in Figure 4-2.  Poa compressa which showed a significant response to Trifolium repens removal in the  1977  4-4),  pasture and to HT treatment in the  showed  a concommitant increased  lanatus removal  in  both  pastures.  1939  expansion  Moreover,  pasture (Tables 4-6 and  rate  with  expansion  increased Holcus  rate  increased  with  increased T. repens removal in the 1939 pasture but fell off sharply when large amounts of T. repens were removed in the 1977 pasture (Figure 4-1).  Expansion rates of Dactylis glomerata and Lolium perenne showed no consistent pattern of response to removal (Figure 4-1); however, both D. glomerata and L. perenne were able to take advantage of open space and respond with increased expansion rate (Figure 4-2). Ranunculus acris appeared to have a high expansion rate; this is anomalous because  it changed from zero presence in some plots.  Both the rate of R. acris expansion and the rate of T. officinale decline were independent of treatment.  Expansion rates of species were averaged over replications of each single species removal  and over the change  in abundance in the  controls for  1983  -  1985  period (Table 4-7). Values are only for those species originally present in three or more of the four replicates. Expansion rates in controls were very similar to those  in removal treatments.  consistently  Some  species,  such  high expansion rates while others,  as  Trifolium repens showed  such as Dactylis glomerata and  Agropyron repens, showed rapid spread only on occasion. Analysis of variance on Box-Cox transformed expansion rates and a non-parametric test (Kruskal-Wallis)  Influence of species removal on interspecific interaction / 115 Table 4-7. Mean expansion rate for the period 1983-1985 for species found initially in at least three of four replicates. R is the response in control plots, Rj/j in Holcus lanatus removal plots, R^ in Lolium perenne removal plots and Rj in Trifolium repens removal plots. IC  (  Species  Agropyron repens  E x p a n s i o n rate  1939  1977  1.767 -0.410  -0.715 -0.745  X  0.783  -0.797 -0.730  -0.567 -0.040  0.100 1.117  0.450 1.148  0.755 0.055  Rfc  -0.562  0.507  it  -0.330  0.325  R«  -0.355  0.772  -0.347  -0.535  -0.015 -0.110  -0.385 -0.227  0.485 0.805  -0.227 1.220  0.307 0.457  1.560 0.980  1.412 2.767  1.070 1.022  1.815  1.322  -0.275  -0.625  -0.207 -0.437  -0.535 -0.685 -0.722  R  ic  it ih  R  R  Rfl  Dactylis glomerata  z'c  R  it ih  R  R  u  R  Holcus lanatus  R  Lolium perenne  R  R  R  Poa compressa  R  R  Trifolium  ih ic it ih  R  R  repens  ic it  il  W  R  ih il  R  R  Taraxacum  R  officinale  R  ih il  R  R  ic it  -0.150  x- species present in less than three of four replicates in 1983.  Influence of species removal on interspecific interaction / 116 on untransformed scores showed the rates were not significantly associated with treatment, previous  and consequently, section  cannot  significant  be  explained  responses to removal discussed by  differential  expansion  rates  in the of  the  responding species.  4.4. DISCUSSION  Grass species had a greater number of significant responses to removal in the 1977  pasture than in the older 1939  removal  experiments  interspecific  pasture. Other authors have found that  in old fields and pastures  interactions.  For  example,  demonstrate  Fowler  (1981)  weak  found  no  or  diffuse  evidence  of  differences in magnitude of response of one species to the removal of others and concluded  that  relationships  among  species  were  relatively  weak  and  diffuse.  Veresoglou (1983) found that removal of subdominants, both in an old pasture and under more controlled glasshouse conditions, had little impact on remaining species.  That members of the same species showed a significant response to removal in the newer but not in the older pasture supports the hypothesis that competition between neighbouring species abates over time, (Harper Support,  either through niche divergence  1964), or by increased equality of resource acquisition (Aarssen 1983). however,  representing  a  depends  on  chronosequence.  acceptance  of  Like  chronosequences  all  the  1977  and  1939  based  on  pastures spatial  arrangements of differing ages, this can be challenged. There is no doubt that the genetic composition of the sown species differs from 1939  to 1977. Lolium  Influence of species removal on interspecific interaction / 117 perenne is the diploid variety in the pasture.  Nevertheless,  phenotypic growing  differences,  for  two  L.  perenne collected  had  years  1939 pasture and tetraploid in the  in  complete a  from  overlap  of  common garden  Similar results were observed for  both  pastures,  despite  morphological characters (R.  Turkington,  unpubl.  1977 initial after data).  Trifolium repens and Holcus lanatus. Abiotic  factors are unlikely to be substantially different in the two pastures because of close proximity and similar topography. Although soil differences exist between the two pastures (Aarssen and Turkington 1985a), many of these difference may be due to soil development over time, i.e. the multiplicity of processes triggered by plant growth and decay. This is not to say that the the present condition of the history  influences  further  1977 pasture will mirror  1939 pasture 38 years in the future because past development.  Rather,  it  suggests  that  relationships  between species will follow a similar course of development.  If  strong  compete?  competitive The  interactions  increased  occur  abundance  of  in the  the  newer  grasses,  pasture,  what  species  Dactylis glomerata, Poa  compressa and Anthoxanthum odoratum, upon removal of Trifolium repens, may have resulted from  their release  from  suppression by T. repens, although this  seems unlikely. An alternative and testable hypothesis is that the removal of T. repens provides a source of nitrogen for which other species compete. Nitrogen products left behind by decaying T. repens roots and nodules can facilitate growth and expansion. If the grasses are able to take advantage of this nitrogen source, competition is not between T. repens and the grasses but among grasses for the nitrogen made available by T. repens decay.  Influence of species removal on interspecific interaction / 118 That the increased abundance of grasses on Holcus lanatus removal is because of release from competitive competition was and  Aarssen  suppression is supported by other studies.  found in replacement  1984)  and pairwise  series studies (Aarssen  competition  1985b), in which H. lanatus demonstrated  studies  Evidence of  1983;  Turkington  (Aarssen and Turkington  a strong competitive  advantage  over  Lolium perenne which, however, decreased with pasture age. In Dactylis glomerata -  H.  lanatus interactions,  demonstrated pasture  age  a  similar  D.  trend  glomerata of  had  increasing  (Aarssen and Turkington  a  competitive  equality  of  advantage,  yield  with  but  increasing  1985b). These trends were postulated  to  result from two different mechanisms: niche differentiation between D. glomerata and H. lanatus, and balancing of the competitive  abilities of L. perenne and H.  lanatus. Thus, H. lanatus may act to control the expansion of other grasses in the initial stages of community formation.  In  removal  experiments,  it  is  possible  that  species  presence of open space. Dactylis glomerata in the  respond  directly  to  the  1977 pasture showed a trend  of increased expansion rate in response to increased available space and increased significantly  in the LHT combined removal which always  open area. A lack of response  left large amounts of  to L. perenne removal, however,  indicates  that  space alone is not the only consideration. The presentation or geometry of that space may be involved. Jayasingam (1985) found that the removal of densely rhizomatous, clumped plants had a greater impact on the abundance of remaining species  than  species, we species,  left  removal of loosely observed that larger,  more  rhizomatous  species.  In removing the  clearing Holcus lanatus, a densely discrete  gaps  than  the  other  various  tillering, clumped removals.  These  Influence of species removal on interspecific interaction / 119 observations  do not explain the differences  in response  of the same species to  the same treatment in two different pastures.  The applicability and utility of removal experiments to study community structure and determine competitive interactions has already been challenged (Bender et al. 1984). Additional limitations of the approach became evident during the study I conducted.  Removal experiments  in pastures  albeit selective grazing, and therefore, , Indeed,  at  distinguish  the  end  removed  of from  the  inadvertently  mimic grazing,  may not represent a novel perturbation.  grazing  grazed  may  season,  areas.  it  Grazing,  was  visually  however,  impossible  leaves  roots  to and  rhizomes that later regenerate.  Grazing provides an intermittent but directional pressure to which plants in the older  pasture  have  been  subject  for  a  longer  time  period.  I  have  argued  previously (Chapter 2) that biological accommodation seems to take place rapidly, i.e. within five years from initial establishment, and will show in Chapter 5 that species composition responds rapidly to management practices. In this chapter, the experiment may be uncovering fine-scale  accommodation that continues for many  subsequent years. Some portion of the measured response may be attributed to dimunition  of  competition  grazing pressure.  but  some  also  reflects  convergence  in  response  to  5. INFLUENCE OF MOWING, FERTILIZATION AND DISTURBANCE REGIMES  5.1. INTRODUCTION  Pastures are maintained by grazing or mowing. These activities remove biomass, eliminate  individual  competitive  plants  and  influence  the  distribution,  abundance  and  balance between species. Darwin (1859) recognized that grazing and  mowing prevented vigorous species from eliminating the less vigorous in a lawn experiment. It has long been recognized that grazing management of pastures has a profound influence on botanical composition (e.g., Jones Milton  1940).  frequency  Evidence from both grazing trials  and  intensity  indicate  that  species'  1933a,b,c; Baker 1937;  and manipulations response  to  of cutting  changes  in  the  management regime can be quite rapid, taking place in a time frame of within a few months to a year or more (Jones 1933d; Gervais 1960). In addition, the experimental exclosure  of grazers such as rabbits (Tansley and Adamson 1925;  Watt 1957) and sheep (Welch and Rawes 1964) have demonstrated a decrease in the  number of plant species and, the  addition of grazers  in biological control  experiments (Dodd 1940 in Harper 1969; Huffaker and Kennet 1959) has shown an increase in species diversity. The results of these latter additive experiments are unfortunately  confounded  by disruption of the  existing  vegetation  prior to  experimental manipulation (Harper 1969).  Grime mowing  (1973)  has  prevent  proposed competitive  that  forms  species  of  from  120  management attaining  such  maximum  as  grazing or  size,  thereby  Influence of mowing, fertilization and disturbance regimes / 121 reducing their competitive  edge and increasing species diversity.  Caswell (1978)  viewed the grazer as a predator, and argued that grazed species are maintained through  a  exclusion. grazing  form In the  densities,  of  co-existence  which  absence of grazing, competitive  exclusion  grazers  and exert  predation pressure grazing densities,  predator-mediated  would be  non-selective  delays  competitive  dominates; a  at high  uniformly high  resulting in a decrease in species co-existence; and, at low grazers would be selective and exert  an intermediate,  patchy  pressure resulting in an increase in co-existence (Caswell 1978).  Low nutrient levels constrain the rate  of dry matter  (Grime  usually  1973,  1977)  while  fertilization  (Spedding  1971). In pure stands,  mortality  (Sukatschew  competitive  outcome  found  no change  grass  mixtures.  mixtures higher  and  1928, (Stern  in the  in Harper and Donald  fertilizer levels.  1977)  increases  size  vegetation  and growth  growth rate increases  and in mixtures, can alter the  1962a,b).  elatius remained  eliminated  the  rate  the rate of  Mahmoud  and Grime (1976)  hierarchy of dominance when fertilizer was  Arrhenatherum  completely  accelerated  production of  the  dominant  subdominant  The addition of fertilizer  to  grass,  added to  grass  in  all  Festuca ovina, at  nutrient-deficient  vegetation  usually causes large-statured species to expand and reduces the number of species in a given area (Grime 1979).  Ecological and agronomic studies confirm the importance of grazing and nutrient levels  in determining  plant  abundance  and influencing  species co-existence but  little attention has been given to their interaction with the small gaps that are inherent in animal grazing e.g.,  rabbit scrapings  and cattle hoof skids.  These  Influence of mowing, fertilization and disturbance regimes / 122 small gaps may function to provide micro-sites for seed germination and seedling establishment facilitate  (Silvertown  1981), serve as a refuge for non-aggressive species, or  vegetative expansion  of existing  plants. Both Grubb (1977) and Grime  (1979) have stressed the importance of gaps in facilitating the ingress of, and maintaining, species in the community.  To  properly  understand  the  influence  of  grazing  and  fertilization  on  pasture  composition requires controlled field experiments. This facility was not available at the Aldergrove pastures so that to estimate this, a pasture-like area was seeded at  the  University  experiment simultaneous  was  of  British  designed  to  Columbia  measure  the  South effects  Campus on  Field  pasture  Station.  species  of  An the  manipulation of levels of mowing, fertilization and removal of small  divots to create small gaps and simulate discrete disturbances. Although mowing differs from grazing in many ways (e.g. it is homogeneous, lacks the biting and pulling action of the grazer, and is non-selective) it is a best approximation of grazing in a field station facility.  5.2. METHODS  Preparation for the  experiment  commenced  in May 1982  in a section of the 2  South Campus Field Station. Seed was sown at a rate of 4g/m ; the seed mix (Table 5-1) past  few  approximated the proportions of Buckerfields Highland Mix over the years.  This  Aldergrove in 1977. 20-20-20 fertilizer  is  the  same  mix  used  It was fertilized in June 1982 and weeded  as  in  the  youngest  pasture  at  with half normal dilution of  required. By May  1983,  the  sown species  Influence of mowing, fertilization and disturbance regimes / 123 dominated the plots and no further weeding was necessary.  The experimental procedure began in May 1983. The seeded area was subdivided 2 into 8 blocks. Four of the blocks received 150g/m  of 4-12-8 dry fertilizer every  3 weeks and the remaining four received no fertilizer. Each pair of fertilized unfertilized blocks were further subdivided into 4 mowing treatments: every week, once every 3 weeks, once every 6 weeks and no mowing. Within each of the 2 eight combinations of mowing and fertilizer treatments were four lm plots. 2 From these lm circular  divots  plots, different 'disturbance regimes' were imposed by removing 10cm  in  diameter.  Originally  I  had  planned  four  small  gap  intensities: no divots, 1 divot every week, 2 divots every 2 weeks and 4 divots every 2 weeks but this proved to be too intense. Subsequent to July 1983, the regime was  modified to: no divots,  1 divot every  2 weeks, 2 divots every 4  weeks and 4 divots every 4 weeks. The divots were made by randomly locating the appropriate number of spots and pounding a metal plug into the ground to remove the upper l-2cm of soil and rooted material. The treatments May to October of each year from  ran from  1983 to 1985 and are planned to continue  beyond this date to assess longer-term effects. All plots were surveyed in September 1985, using four .5 x .5m quadrats per 2 lm  plot. Each quadrat was subdivided into a grid of 25 points and the rooted  cover at each point was recorded (100 points per plot). Lolium perenne and L. multiflorum were not distinguished because of difficulties in differentiating them in the vegetative state.  Influence of mowing, fertilization and disturbance regimes / 124 Table 5-1. Percentage composition of the original seed mix.  Species  Percentage  Dactylis glomerata Lolium perenne Lolium multiflorum Phleum pratense Trifolium pratense Trifolium repens  46 12 8 4 15 15  5.3. RESULTS  In  September  present  1985,  in the  plots  all of  the  seeded  (Table 5-2).  species  plus  several  volunteers  were  The origin of Festuca pratensis is uncertain  although it can occur in commercial seed sources of Lolium spp. Of the seeded species,  only  Dactylis glomerata was  abundance  was  similar  ryegrasses  (Lolium spp.)  to  the  and  found  input  clovers  in  frequency  all of  treatments 46%.  (Trifolium repens and  and  The  its  mean  abundance  of  T. pratense) was  considerably lower than had originally been seeded.  Analysis of the abundance data was univariate by species using a BMDP (Dixon 1983) species  fully crossed (>2%  factorial analysis  cover)  plus  bare  ground were  significant effect on the percentage fertilization  had a  significant  of variance. Only the nine most common  effect  analysed.  Mowing had a highly  cover of all species except Phleum pratense; on six  species; but, divot removal had a  significant effect only on Dactylis glomerata (p = 0.0509) (Table 5-3).  Influence of mowing, fertilization Table  and disturbance regimes / 125  5-2. Mean and standard deviation of the percentage cover of species and bare ground on South Campus treatment plots in September 1985.  Species Dactylis  glomerata  Lolium spp. Agrostis sp. Hypochoeris radicata Festuca pratensis Trifolium repens Rumex acetosella Trifolium pratense Phleum pratense Poa sp. Holcus lanatus Taraxacum officinale Cirsium arvense Plantago major Medicago lupulina Cerastium vulgatum Plantago lanceolata Alopecuris pratensis Agropyron repens  Moss Bare ground  Because  divot  combined  removal  to show  rarely  Mean  S.D.  45.5 8.89 6.73 5.04 4.91 4.48 3.44 2.89 2.34 0.43 0.23 0.21 0.20 0.13 0.13 0.07 0.05 0.02 0.02 0.03 14.3  16.4 7.01 12.1 5.83 4.60 6.20 5.21 4.58 2.51 1.63 1.12 0.77 0.62 0.61 0.69 0.36 0.21 0.12 0.12 0.17 11.9  had  a  significant  species response to fertilization  effect,  these  and mowing  treatments  (Figure 5-1). The  percentage cover of Dactylis  glomerata  species responded  positively  to some frequency  Agrostis  abundant only in the frequently mown blocks, whereas  spp.  sp., was and  Festuca  frequencies. Trifolium  pratensis  more  repens was reduced  taller, more upright T. pratense. radicata  were  was reduced  Weedy  were  by frequent mowing but other  of cutting. The fine weedy grass,  abundant  at  intermediate  Lolium mowing  less by more frequent mowing than the dicots, Rumex  acetosella  were most abundant at intermediate mowing frequencies.  and  Hypochoeris  Influence of mowing, fertilization  and disturbance regimes / 126  Table 5-3. Percentage variation in abundance of common species and bare ground explained by different mowing (m), fertilization (f) and divot removal (d) regimes. *** - p<0.001; * * p<0.01; * - p<0.05.  Species  #Dactylis glomerata  m  f  ***  **  31  37  8.2  m-f  d  m-d  f-d  mfd  1.2  0.4  3.1  * 1.7  ***  *  **  #Lolium spp.  60 •  7.1  5.5  1.2  1.9  0.9  * 3.9  #Phleum pratense  3.0  44  1.6  3.4  4.5  1.4  3.2  ***  **  30  32  18  ***  **  * *  23  31  16  *  #Trifolium repens #Trifolium pratense  * 1.3  1.5  1.9  1.9  0.3  4.6  0.2  4.1  **  ***  Festuca pratensis  20  0.4  Hypochoeris radicata  4.9  12  1.0  7.2  0.4  1.8  0.2  1.5  ***  ***  Agrostis sp.  2.7  53  1.2  6.2  **  **  ***  24  27  11  1.3  2.6  0.6  2.1  2.0  9.8  2.5  3.5  0.3  1.8  2.0  2.4  0.9  2.1  **  Rumex acetosella  28 ***  bare ground  66  ** 2.1  6.2  Species labelled with an # were constituents of the original seed mix.  Influence of mowing, fertilization  and disturbance regimes / 127  Figure 5-1. Percentage abundance of nine species on fertilized and unfertilized plots subject to different mowing regimes.  CS B CD • SI C3  Agresf I* tp. HyfwxJiooii i«»urtu Fnrhicn pffltwwto Irlfatum r«p«f» (bjmax oeatoaale VtfoAufn prfftanM  Fertilized plots 70 - i  605040302010-  Unfertilized plots 60-i O C  o  T5 C =3  O  <D  50 40  30-  o CD Q_  10  Influence of mowing, fertilization and disturbance regimes / 129 Fertilization increased the abundance of all the responding grasses and decreased the abundance of clovers, a result typical of grass-clover mixtures (Herriott and Wells 1960; Stern and Donald 1962a,b; Dennis and Woledge 1985).  Many species showed a strong interaction effect between fertilization and mowing and some species responded to an interaction with divot removal. Mowing and low intensity gap formation increased the abundance of Festuca pratensis whereas lack of fertilizer and medium intensity gaps increased the abundance of Trifolium  repens.  The amount of bare ground decreased increased was  mowing frequency.  a greater  At the  with the  addition of fertilizer and with  highest mowing frequency,  proportion of bare ground in the  most  intense  however,  there  divot removal  treatment leading to an interaction effect.  2  The  number of species per  species per plot. least  frequently  lm  plot  ranged  from  The lowest number of species was mowed  plots  and the  frequently mowed plots (Figure 5-2). of species per plot was  greatest  2  -  was  used  in  number with  with fertilized,  unfertilized,  most  The distribution of the variance of number  unknown, and therefore,  an  averaging eight  associated  analysis  of variance was not  strictly applicable to test for treatment effects. Nonetheless, variance  14,  exploratory  manner,  because  factorial analysis of  the  large  number of  treatment cells constrained the use of alternate non-parametric techniques. Results were similar for both untransformed and square-root transformed values: highly significant  mowing and fertilizer effects  (p< 0.001) and strong interaction effects  Influence of mowing, fertilization and disturbance regimes / 130  Figure 5-2. Number of species per square metre on fertilized and unfertilized plots subject to different mowing regimes. fertilized unfertilized  tex  Influence of mowing, fertilization and disturbance regimes / 132 (p<0.05) but no effect from divot removal on the number of species in the plot.  5.4. DISCUSSION  Mowing and fertilizing were found to strongly influence plot  composition.  The response  regimes  has been associated  species  such  as  of  species  abundance  with tolerance  species abundance and  to  mowing  and fertilizer  of shading and many large-statured  Dactylis glomerata, responding  primarily  to  nitrogen,  shade  smaller-statured species (Stern and Donald 1962a,b; Snaydon 1971; Rhodes and Stern 1978). Shading is further exacerbated by infrequent cutting. Fertilization of grasses frequently  suppresses low growing clovers, because grasses are able to  grow taller in response to nitrogen application. Mowing can also provide nitrogen to  grasses,  increasing  because  the  amount  cutting  clovers  of nitrogen  causes  transferred  growth (Bland 1967). Unlike fertilization alone,  leaf  and  to  grasses  nodule  drop,  thereby  and favouring their  the repeated cutting of grasses  and clover prevents shading, thereby ameliorating suppression of clover.  The  cultivars used in this  experiment  Most cultivars sown in pastures  are typical of those sown  in pastures.  are selected for herbage production, i.e., rapid,  upright growth and to a lesser extent, seed production (Snaydon  1984). These  cultivars require a nutrient-rich environment to maintain them as the dominant species (Grime 1979). Grime  relates the  past thirty years in European meadows  reduction in species density and pastures  to increased  over the  "competitive  dominance brought about by stimulating the yield of more robust and productive species and genotypes through application of high rates of mineral fertilizer" (p.  Influence of mowing, fertilization and disturbance regimes / 133 125).  In the absence of continued fertilizer input, however, botanical composition  of pastures  changes to a mixture of sown, naturalized and weedy species. In  addition, grazing selects against many of the characters for which cultivars are bred so that naturalized pasture populations are more likely to contain more low growing and laterally  spreading forms  than the  original seed source. Snaydon  (1984) also points out that the traits of cultivars may well carry the penalty of poorer persistence  so that established  pastures  will contain a greater  range of  variation in plant populations as well as higher species density.  In pastures, grazing removes biomass that would shade and suppress the growth of less vigorous or small-statured species, thereby delaying competitive  exclusion  of these species. Competition for light, water and nutrients interacts with grazer - predator frequency, intensity and  species  potential  co-existence.  species density  and selectivity  Grime's (1973)  to determine botanical composition  proposed  model,  in  which maximum  is found at intermediate levels of nutrient availability,  predicts that maximum species density occurs in relatively infertile pastures. In infertile pastures, the interaction of the five processes (Grime 1979) that control species density  are  such that  (i)  dominance  of  large  species is  reduced; (ii)  nutrient "stress" and (iii) disturbance are increased which promotes (iv) ingress of suitable species and (v) niche differentiation. Accordingly, the continued decline in abundance of Dactylis glomerata in the Aldergrove pastures, could be attributed to failure  to  maintain high fertility  levels coupled with  reduction because of  its  sensitivity to grazing.  The experiment in this chapter confirms that mowing and fertilization  influences  Influence of mowing, fertilization and disturbance regimes / 134 species composition and co-existence but negates a role for small disturbances. The created gaps appear to neither facilitate nor impede species abundance and spread,  and  to  have  no  effect  on  species  richness.  These  gaps  provide no  additional opportunities for new species to establish in the plots. These results agree  with  the  observations  of  Rapp  in  tall  and  Rabinowitz  (1985)  who  created  found  that  seedling  2 disturbances  of  survival was probably  about  27cm  grass  prairie  and  no greater on disturbances than in undisturbed controls. This  related  to  a  number  of  factors.  First,  constituent  species  of  is the  community may have either no need for gaps or require a larger area to trigger germination (Grime 1979; see Hilton et al. 1984 for variation within a species). A second factor is the influence of surrounding plants on survival and growth of seedlings. (Gross  In general, more space leads to higher seedling survival and growth 1980;  desiccation (Harper  Snaydon  may  limit  and  Howe  establishment  1986) of  although  some  intolerance  species  in  very  of  heat  large  and  openings  1977). A third consideration is that even if gaps were favourable for  seedling establishment, the amount of divot removal was inadequate and did not significantly  increase  the  amount of bare  areas  available.  Rather, the  results  indicated that plant spacing in response to cutting and fertilizer treatment had a greater effect on creating bare ground than actual divot removal.  The  results  of  this  experiment  invoke  a  comment  on  certain  experimental  practices to determine competitive outcome on interactions between species. It is apparent greater examples  that or  the  lesser  outcome extent,  exist such as  of by  the  species the  interactions  cutting  and  can be  fertilization  manipulated, to regime.  a  Numerous  increased production of Lolium perenne relative to  Influence of mowing, fertilization and disturbance regimes / 135 Dactylis glomerata with increased frequency of cutting or earliness of defoliation (England  1968)  tenuis) by Replacement  L.  or the reduction in suppression of Agrostis capillaris perenne with  series,  additive,  severe and  defoliation  Nelder  (Harris  designs  and  used  to  (syn.  Thomas test  A.  1972).  competitive  interactions, all require management regimes which often include fertilization and clipping  for biomass  outcome  of  species  estimation. interactions  Such and  generalized only with extreme caution.  management the  results  of  can  strongly  such  influence  experiments  the  can be  6. GENERAL DISCUSSION The sun rises and the sun sets and hurries back to where it rises. The wind blows to the south and turns to the north round and round it goes, ever returning on its course. All streams flow to the sea, yet the sea is never full. To the place the streams come from, there they return again.  Ecclesiastes, Chapter 1, vs 5-7  The chapters pasture  of this thesis focus on aspects of the  community.  maintenance  Disturbance  by  grazing  or  role of disturbance in a  mowing  of pastures in non-grassland climates.  is  intrinsic  to  the  In addition, on a local scale  "grazing animals frequently sit, lie, scratch and paw on the pasture in addition to walking, running and jumping on it" (Spedding 1971)  and they also deposit  dung and urine. Add to these activities, mounds of small animals, insect grazing and  mechanical  disturbance  from  haymaking and harrowing and the  view of  pastures as dynamic and disrupted is confirmed.  It is not surprising that one feature of the Aldergrove pastures is the variability associated with plant abundance because fluctuations and pastures Vickery  1981;  are well documented Thorhallsdottir 1984;  (e.g.  Rabotnov  Snaydon  in abundance in grasslands 1966,  1974;  1985). They are  reversible,  and associated with variations in meteorological  conditions  on  a  seasonal,  yearly  or  short-term  136  (e.g.  1-3,  Spedding  1971;  characteristically  and (or) hydrological 5-10  years)  basis  General discussion / 137 (Rabotnov 1974). Biotic interactions, such as the nitrogen-mediated feedback loop that is posited  for Lolium/Trifolium fluctuations  (Blaser  and Brady  1950), also  lead to changes in relative abundance. Moreover, management decisions on timing and  extent  of  mowing,  grazing  and fertilizing,  effect  species  composition  and  abundance.  Yet,  despite the flux of species composition,  the  botanical composition  of the  pastures is remarkably stable. Over the eight years that records have been kept for  these pastures,  only  composition and abundance and seeding.  the  1977  and this  No species present  eliminated from these pastures  field  showed  ceased 4-5  in the  initial  although the  directional change years  subsequent  seed mix has  in species  to ploughing  been  completely  abundance of Trifolium pratense is  severely reduced. Rather, it appears that the initial species composition determined the  subsequent  Ingress  of  composition although proportions have changed  weedy  species  occurs  rapidly  -  witness  the  (cf.  Egler 1954).  abundance  lanatus, Taraxacum officinale, Agropyron repens and Ranunculus  of Holcus  acris. The  composition of the pastures appears, nevertheless, to be constrained by the major biomass/abundance  contributors which vary in relative proportions from year to  year. Less abundant species are interspersed in various patterns which apparently have little impact on the total dynamics. One can imagine visiting the pastures in future years and seeing the same species maintaining various proportions and distributions of abundance within some set of limits for which neither boundary conditions nor rules of assembly are yet defined or understood.  Maintenance of the limits and rules appears to be dependent on the continuation  General discussion / 138 of the management practices in the pastures. Holling (1973) notes that in 'real world  examples'  changes  the  conditions.  systems  driving  For  may  variables  example,  persist and  permanent  indefinitely  flips or  the  long  until  system term  a  novel  outside  removal  introduction  the  of  boundary  herbivores  or  repeated harvesting for maximum yield are likely to strain the resilience of the system  and  push  it  beyond  current  boundaries.  But,  within  the  current  management practices, the pastures appear to exhibit considerable variation within stable limits.  The  three  within  pastures  are  not  entirely  closed  a matrix of semi-rural copses, roads  communities,  existing  and rights-of-way  as  that  they  do  constrains  species and gene influx. Both plant and animals are able, however, to disperse to a limited extent between the pastures. Observations suggest that the continued presence  of Trifolium pratense in the older pastures  from adjacent right-of-ways  is derived from recruitment  whereas the occasional local outbreak of Gnathalium  uliginosum is from a persistent seed bank. These species are minor constituents of the community, but major constituents, into  the  pastures,  not  only  such as Holcus lanatus, have dispersed  from adjacent  areas  but  also,  via a seed bank  accumlated over time (see Grime 1979).  Small disturbances  such as molehills  and dung pats have been shown to play  little or no role in providing access for weedy species to establish neither in the Aldergrove however,  pastures  nor  on  South  Campus. In  order  a space of some size must be available.  for  seedlings  to  grow,  The size required will be  dependent on species (e.g. Denslow 1980) or genotype (Hilton et al. 1984). Many  General discussion / 139 species have a requirement for light, or high or fluctuating temperatures in order to  germinate.  growing  in  Hilton closed  et al.  (1984)  grassland  found  swards  that  germinated  Poa trivialis seed from plants under  low  red:far-red ratios  similar to dense shade conditions whereas those from open, arable land required higher red:far red ratios. Other studies have emergence  shown that the  is lower when vegetative cover is high (e.g.  rate of seedling  Fenner 1978; Silvertown  1981; Gross 1984) and growth and survival is typically reduced (Ross & Harper 1972).  On average, between 1.2% (in the 1939 pasture) and 3.4% (in the 1958 pasture) was bare from various causes during the study period. Grazing and activities  removed  considerable  biomass  and left  numerous  associated  small gaps because  plants, some more readily than others (e.g. Dactylis glomerata) were more readily uprooted. Very small gaps may be sufficiently large for many pastures species to regenerate.  Indeed, root death after loss of shoot mass may even provide, after  complex cycling through soil fauna, an enriched nutrient environment for seedling establishment (Newbery 1979). Several authors (e.g. King 1971; Snaydon & Howe 1986)  have  deterrent  of  shown  that  seedling  nutrient  limitation by  establishment  although  the  established extent  to  plants which  is  a major  this  affects  survival is species • dependent.  In the  pastures,  molehills  Trifolium repens, and species  required  these  to  and dung pats a  gaps  lesser to  extent,  maintain  provided sites Ranunculus their  for  establishment  acris seedlings.  presence  in  the  of  Neither  community.  Ranunculus acris was found to germinate and survive in a variety of gap sizes  General discussion / 140 whereas T. repens spread by stolons throughout the pastures. Cahn and Harper (1976) suggested that essential  to  frequent  inputs  of  new  avoid local dominance by a few  long-lived clones.  Occasional seedling  genotypes either  establishment  of  clonal  competitively  (3% of the  species  is  dominant or  ramet population  per year) is sufficient to maintain considerable genet variation within a population (Soane  and  establishment Parish,  Watkinson of  this  unpublished)  1979).  magnitude but  the  In  good  was  calculated  frequency  seed  of  years, to  'good'  e.g.  occur in seed  1984, the  years  seedling  pastures  has  yet  to  (R. be  established.  The variability in seed germination recorded for Trifolium repens, cautions against generalizations  that  molehills  and  dung  pats  never  regeneration of pasture species. In the fall of 1982,  have  a  role  in  sexual  a large flush of monocot  seedlings was observed in the three pastures both on and off molehills. It is a pity  that  no crystal ball  survivorship subsequent  that  first  or seer provided the  year  because no other  foresight  to  monocot flush  measure was  seedling  observed in  years. It appears that seed germination is highly episodic in these  pastures.  Gaps such as molehills  and dung pats did provide opportunities for vegetative  expansion. These gaps appeared to function a in manner similar to small gaps in forests (Runkle 1982)  allowing competitive release of neighbouring individuals.  Pasture species that are able to take advantage  of these openings are usually  those capable of rapid lateral spread either by rhizomes (e.g. Agropyron repens, Poa compressa) or stolons (e.g. Trifolium repens).  General discussion / 141 Formation of molehills and dung pats killed most plants that were buried under them although those capable of regeneration from fleshly roots (e.g. Taraxacum officinale) survived temporary burial. In general, the probability of death for a genet depends on its size. The contribution of small disturbance to the removal of  genets of  such  species  as  Trifolium  repens and Poa compressa, in which  ramets may be spread over large areas was probably minimal, but genet death could  result  for  perenne. McNeilly  closely  tillering grasses  & Roose  such  (1984) found that  as  Holcus lanatus and Lolium  the  number of genotypes of L.  perenne per unit area of an old pasture was considerbly lower than in younger pastures. The loss of genotypes during sward establishment  has been related to  the loss of smallest individuals, often resulting from small differences in timing of seedling emergence (Ross & Harper 1972). Subsequent losses have been attributed to  competitive  sward  dominance of  management  interactions  have  yet  practices to  be  a few  superior  (McNeilly  &  demonstrated  genotypes which are Roose  in  1984).  pastures  Strong  (Strong  generally accepted statements that they do exist (Donald 1963;  adapted  to  competitive  1983)  despite  Snaydon 1978).  Small disturbances serve to randomly remove individuals yet provide no consistent opportunity for new recruits to enter. Depletion of genotypes may, in part, result from random removal without replacement, rather than selective processes.  Species respond differently to disturbance of various sizes. Trifolium repens could utilize  molehills  and  dung  pats  colonization. Dactylis glomerata was  for  both  frequently  sexual  propagation  and  found to occupy the  vegetative same point  from fall to spring. In addition, D. glomerata generally increased in the sward around dung but not around molehills, proliferated when fertilizer was added to  General artificial  swards, and responsed  observations release,  that D.  suggest  either  by  positively  glomerata  pre-empting  was  nitrates  T.  when  released  by  removed. These  advantage  decaying  of nutrient  nodules  or by  had, however, little ability to  molehills and dung pats.  response of Lolium  The  was  able to take  capturing nutrients leached from dung. D. glomerata invade  repens  discussion / 142  perenne  which Holcus  from  lanatus  increases in several species although  of  other  agree with space  those  declined less in  had been removed. Removal of H.  to  species nor could  variable. In general, it did not invade  of declining abundance, L. perenne  small disturbances. In a year areas  was very  it readily  of Thorhallsdottir  against invasion by other  H.  lanatus  invade  lanatus led  did not increase with  removal  molehills and dung. These findings  (1984) who  found  species but showed  H.  lanatus  little  effectively held  tendency  to pre-empt  gaps in the sward.  The  response pattern of individual species to molehills and dung pats recorded at  0.25  m  2  scale  association  of observation  was  not evident  at 25  m  2  . The  of species and disturbance, in general, was not predictable from one  scale of observation to the other. There were exceptions: Agropyron invaded  small  disturbances  disturbance .and Holcus in  disturbed  significantly beyond  areas.  and  lanatus, For  other  scale. From  interaction  between  its abundance  increased  in  repens readily areas  of  high  unable to exploit disturbances, was not abundant species,  influencing distribution  the local  significant  direction of  invasion  pattern  was  not  of abundance within the pasture  these  discrepancies, I must  pattern  of species  conclude  abundance  and  a  factor  at anything that despite disturbance,  General discussion / 143 disturbances  per se have  little  organizing or causal  explanation of the pattern within the pastures interactions were associated linked to those that consume  earthworms,  force.  The most  is that because most  obvious  significant  with molehills, plant species were tracking resources  moles were also exploiting. species  positively  Because  associated  with  moles predominantly increased  density  of  molehills may be those exploiting areas of increased soil fertility.  Aarssen (1983) and Aarssen and Turkington (1985a) have proposed a model of community processes derived from observations  and experiments  on species from  these pastures. They suggested that the community evolved over time as less fit genotypes were eliminated by competition and grazing. The resultant community is  characterized by strongly  competitive filtered  stalemate  through  in  competitive  which  interactions  none  with  individuals that  can  achieve  existing  co-exist because of a  an advantage.  individiuals  so  that  Recruits are only  those  of  equivalent abilities in that time and location persist.  Evidence to support or refute this model was not collected in this study in the same pastures.  Yet, a discussion  of some of the ideas seems appropriate. In  order to have selection, there must be genetic variation, which is generated by mutation and/or recombination. In higher plant populations, usually  achieved  by  sexually  produced  seeds.  This  study  genetic variation is of  the  Aldergrove  pastures provided little information on rates of recruitment for most species but because of the  focus  Trifolium repens has  on small disturbance, some idea of the  'life  emerged.  highly  Seed  germination is  Probably it is restricted to gaps of 500cm  2  apparently  history' of episodic.  or greater; none were observed in  General discussion / 144 any of the high.  smaller gaps.  However,  density-independent  Seedling density  early  mortality  factors  as  varied and mortality was  resulted  from,  in  the  generally  main,  such  grazing by slugs and small mammals, burial by  re-opening of mole tunnels, or up-rooting by haymaking. Initial interactions with other individual plants  were most certainly intraspecific  but occurred after  the  plant became established on the disturbance. Without specific demographic studies following the fates of these survivors as other species spread into the gaps, it is impossible suggest,  to refute Aarssen's hypothesis of competitive however,  competition-free  the  following  scheme:  once  combining ability. I will  established  in  the  relatively  environment of the molehill, T. repens begins to spread laterally  into the sward. In areas where  conditions  are favourable, internodes  are short  and individual plants concentrate biomass in these areas but where condition are unfavourable,  internodes  (Solangaarachchi adequate  are  long  and individuals do not  establish  abundantly  1985). What constitutes a 'favourable' site? Axiomatically, it is  resource  levels. This can result  from numerous  factors;  (1)  resources  may not be limiting, i.e. abundant beyond the capability of the plant to deplete in  the  time  frame  available;  (2) resources  may  be  available  because  of  partitioning by niche differentiation (ecological combining ability); (3) resources may be available because ability to exploit is equivalent (competitive combining ability). The relative importance of these factors was not tested but the role of selection appears to be much less than previously postulated:  recruits are rare, random  events predominate, yet the pastures remain stable..  If pastures are characterized by random event, it does not follow that they are also characterized by random processes. Events may trigger an orderly series of  General discussion / 145 processes.  The  replacement  of one  complex, interweaving of scale  species  by  another  and pattern observed  is not  random.  in the pasture  The  suggests  combination of process and forces whose interaction await further study.  a  7. LITERATURE CITED Aarssen, L.W. 1983. Interactions and coexistence of species in pasture community evolution. Ph.D. thesis, University of British Columbia.  Aarssen,  L.W. and R. Turkington. 1985(a). Vegetation dynamics and neighbour associations in pasture community evolution. J. Ecol. 73:585-603.  Aarssen,  L.W. and R. Turkington. 1985(b). Competitive relations among species from pastures of different ages. Can. J. Bot. 63:2319-2325.  Allen,  E.B. and R.T.T. Forman. 1976. Plant species removals and old-field community structure and stability. J. Ecol. 57:1233-1243.  Allen,  E.B. and D.H. Knight. 1984. The effects secondary succession in sagebrush Southwestern Naturalist 29:407-422.  Baker,  H. 1937. Alluvial meadows: a comparative meadows. J . Ecol. 25:408-420.  Bazzaz,  Beddows,  of introduced annuals on grassland, Wyoming.  study  of grazed  and mown  F.A. 1983. Characteristics of populations in relation to disturbance in natural and man-modified ecosystems. In Disturbance and Ecosystems, H.A. Mooney & M. Godron (eds.), pp. 259-277. Springer-Verlag, Berlin.  A.R. 1961. Biological flora J. Ecol. 49:421-430.  of the British  Uses.  Holcus  lanatus L.  Begon,  M.,  J.L. Harper and C.R. Townsend. 1986. Ecology: Individuals, Populations and Communities. Sinauer Assoc., Inc. Sunderland, Mass.  Belsky,  A.J. 1986. Revegetation of artificial disturbances in grasslands of the Serengeti National Park, Tanzania. I. Colonization of grazed and ungrazed plots. J. Ecol. 74:419-437.  Bender,  E.A., T.J. Case and M.E. Gilpin. 1984. Perturbation experiments in community ecology: theory and practice. Ecology 65:1-13.  146  Literature cited / 147 Berry,  J . and O. Bjorkman. 1980. Photosynthetic response and adaptation to temperature in higher plants. Ann. Rev. Plant Physiol. 31:491-543.  Bland, B.F. 1967. The effect of cutting frequency and root segregation on the yield from perennial ryegrass - white clover associations. Agric. Sci. Camb. 69:391-397. Blaser, R.E. and N.C. Brady. 1950. Agron. J . 42:128-135. Bliss,  Nutrient competition in plant  C.I. and Fisher, I. 1953. Fitting the negative biological data. Biometrics 9:176-200.  associations.  binomial distribution to  Bradshaw, A.D. 1965. Evolutionary significance of phenotypic plasticity in plants. Adv. Genet. 13:115-155. Brokaw, N.V.L. 1985. Treefalls, regrowth, and community structure in tropical forests. In The Ecology of Natural Disturbances and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp. 53-69. Academic Press, Inc. Orlando. Cahn, M.G. and J . L . Harper. 1976. The biology of the leaf mark polymorphism in Trifolium repens L. Distribution of phenotypes at a local scale. Heredity 37:309-325. Canada Soil Survey Committee. 1978. The Canadian system of soil classification. Can. Dept. Agric. Publ. 1646. Supply and Services, Ottawa. Canham,  C D . and P.L. Marks 1985. The response of woody plants to disturbance: patterns of establishment and growth. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp 197-216. Academic Press, Inc. Orlando.  Caswell, H. 1978. Predator-mediated coexistence, a nonequilibrium model. Amer. Nat. 112:127-154.  Literature cited / 148 Christensen,  Clements,  Connell,  N.L. 1985. Shrubland fire regimes and their evolutionary consequences. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp. 86-100. Academic Press, Inc. Orlando.  F . E . 1916. Plant Succession. An Analysis Vegetation. Carnegie Inst., Washington. J.H. 1978. Diversity in Science 199:1302-1310.  tropical  rain  of  the  forests  Development  and  coral  of  reefs.  Darwin, C. 1859. The Origin of Species. Penguin Books, New York, 1968 edition. Davies,  H.  1966. Molehills Soc. 21:148-149.  and  pasture  reversions.  J.  British  Grassld.  Davis, R.M. and J . E . Cantlon. 1969. Effect of size area open to colonization on species composition in early old-field succession. Bull. Torrey Bot. Club 96:660-673. Dayton,  P.K. 1971. Competition, disturbance and community organization: the provision and subsequent utilization of space in a rocky intertidal community. Ecol. Monogr. 41:351-389.  Dennis, W.D. and J . Woledge. 1985. The effect of nitrogenous fertilizer on the photosynthesis and growth of white clover/perennial ryegrass swards. Ann. Bot. 55:171-178. Denslow,  Denslow,  Dixon,  J.S. 1980. Gap partitioning Biotropica 12; Suppl., 47-55.  among  tropical  rainforest  trees.  J.S. 1985. Disturbance-mediated coexistence of species. In The Ecology of Natural Disturbances and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp.307-323. Academic press, Inc., Orlando.  W.J. 1983. BMDP Berkeley.  Statistical  Software.  University  of  California  Press,  Literature cited / 149 Donald,  C M . 1963. Competition among crop and pasture 15:1-118.  plants.  Adv. Agron.  During, C. and K . J . McNaught. 1961. Effects of cow urine on growth of pasture species and uptake of nutrients. N.Z.J. Agric. Res. 4:591-605. Egler,  F . E . 1954. Vegetation science concepts 1. Initial floristic composition, a factor in old-field vegetation development. Vegetatio 4:412-417.  England,  Fenner,  Fienberg,  F.  1968. Competition in mixtures of herbage grasses. J. Appl. Ecol. 5:227-242.  M. 1978. A comparison of species to establish J. Ecol. 66:953-963. S.E. 1970. The analysis Ecology 51:419-433.  the abilities of colonizers and closed-turf from seed in artificial swards.  of  multidimensional  contingency  tables.  Fienberg, S.E. 1980. The Analysis of Cross-classified Categorical Data. 2nd ed. MIT Press, Cambridge, Mass. Fowler,  N.  1981. Competition and coexistence in a north Carolina grassland. II. Effects of the experimental removal of species. J. Ecol. 69:843-854.  Fox, D.J. and K . E . Guire. 1976. Documentation for MIDAS. Statistical Research Laboratory, University of Michigan, Ann Arbor. Garwood, E.A. and K.C. Tyson. 1979. Productivity and botanical composition of a grazed ryegrass/white clover sward over 24 years as affected by soil conditions and weather. In Changes in Sward Composition and Productivity. A . H . Charles and R.J. Haggar (eds.), Brit. Grassld. Soc. Hurley, pp. 41-46.  Gervais, P. 1960. Effects of cutting treatments on Ladino clover grown alone and in mixtures with grasses. I. Productivity and botanical composition of forage. Can. J . Plant Sci. 40:317-327.  Literature cited / 150 Gleason, H.A. 1926. The individualistic concept Torrey Bot. Club 53:7-26.  of the  plant association.  Bull.  Glendenning, R. 1957. Biology and control of the coast mole Scapanus ovarius True, in British Columbia. Can. J . Anim. Sci. 39:34-44. Goldberg,  D.E. and P.A. Werner. 1983. The effects of size of opening in vegetation and litter cover on seedling establishment of goldenrod {Solidago spp.). Oecologia (Berlin) 60:149-155.  Grime, J.P. 1973. Control of species density in herbaceous vegetation. J . Envir. Man. 1:151-167. Grime,  J.P.  1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Amer. Nat. 777:1169-1194.  Grime, J.P. 1979. Plant Strategies Sons, Chichester.  and Vegetation Processes.  John Wiley and  Gross, K . L . 1980. Colonization by Verbascum thapsus (Mullein) of an old-field in Michigan: experiments on the effects of vegetation. J . Ecol. 68:919-927. Gross, K . L . 1984. Effects of seed size and growth form on seedling establishment of six monocarpic perennial plants. J . Ecol. 72:369-387. Grubb, P.J. 1977. The maintenance of species-richness in plant communities: the importance of the regeneration niche. Biol. Rev. 52:107-145. Grubb, P.J., D. Kelly and J . Mitchley. 1982. The control of relative abundance in communities of herbaceous plants. In The Plant Community as a Working Mechanism. E.I. Newman (ed.) pp. 79-97. Special Publication No. 1 of British Ecological Society. Blackwell Scientific Publications, Oxford.  Harper, J . L . 1964. The nature and consequence of interference amongst plants. In Genetics today, XI International Congress of Genetics 2:465-482.  Literature cited / 151 Harper, J . L . 1967. A Darwinian approach to plant ecology. J . Ecol. 55:247-270. Harper, J . L . 1969. The role of predation in vegetational diversity. In Diversity and Stability in Ecological Systems, Brookhaven Symp. in Biol. 22, 48-62.  Harper, J . L . 1977. Population Biology of Plants. Academic Press Inc., London. Harris, W. and V . J . Thomas. 1972. Competition among pasture plants. II. Effects of frequency and height of cutting on competition between Agrostis tenuis and two ryegrass cultivars. N.Z. J . Agric. Res. 15:19-32. Henry, J.D. and J.M.A. Swan. 1974. Reconstructing forest history from live and dead plant material - an approach to the study of forest succession in southwest New Hampshire. Ecology 55:772-783. Herriott,  Hill,  J.B.D. and D. Wells. 1960. Clover nitrogen J . Br. Grassld. Soc. 15:63-69.  M.O. 1973. Reciprocal averaging: J . Ecol. 61:237-249.  an  and sward productivity.  eigenvector  method  of  ordination.  Hils, M . H . and J.L. Vankat. 1982. Species removals from a first-year old-field plant community. Ecology 63:705-711. Hilton,  Holling,  J.R., R.J. Frond-Williams and J. Dixon. 1984. A relationship between phytochrome photoequilibrium and germination of seeds of Poa trivialis L . from contrasting habitats. New Phytol. 97:375-379. CS.  1973. Resilience and Ecol. Syst. 4:1-23.  stability  of  ecological  systems.  Ann. Rev.  Huffaker, C.B. and C.E. Kennet. 1959. A ten-year study of vegetational changes associated with biological control of Klamath Weed. J. Range Mgmt. 12:69-82.  Huston, M . 1979. A general model of species diversity. Amer. Nat. 113:81-101.  Literature cited / 152 Hutchinson, C.S. 1979. Some effects of cow dung on pasture composition. M.Sc. thesis, University College of North Wales, Bangor, Wales. Jain,  S.K.  1983. Genetic characteristics of populations In Disturbance and Ecosystems. H.A. Mooney and M . Gordon (eds.) pp. 240-258. Springer-Verlag, Berlin.  Jalloq, M.C. 1975. The invasion of molehills by weeds as a possible factor in the degeneration of reseeded pasture. I. The buried viable seed population of molehills from four reseeded pastures in West Wales. J . Appl. Ecol. 12:643-657.  Jayasingam, T. 1985. Comparative demography of four Composites. Ph.D. thesis, University College of North Wales, Bangor, Wales. Jones,  M.G. 1933a. Grassland management and its influence on the I. Factors influencing the growth of pasture Emp. J . exp. Agric. 1:43-47.  sward. plants.  Jones,  M.G. 1933b. Grassland management II. The management of a Emp. J . exp. Agric. 1:122-128.  and its influence on the sward. clovery sward and its effects.  Jones,  M.G. 1933c. Grassland management III. The management of a Emp. J . exp. Agric. 1:223-234.  and its 'grassy'  Jones,  M.G. 1933d. Grassland management and its influence on the sward. IV. The management of poor pastures. V. Edaphic and biotic influences on pastures. Emp. J . exp. Agric. 1:361-367.  influence on the sward. sward and its effects.  de Jong, P. and M. Greig. 1984. First-order Markov chains with a zero diagonal transition matrix. Biometrics 40:101-107. de Jong, P. and M. Greig. 1985. Models and methods for pairing data. Can. J . Stat. 13:233-241.  de  Jong,  P.,  M . Greig and D. Madan. 1983. J. Amer. Stat. Assoc. 78:332-336.  Testing  for  random pairing.  Literature cited / 153 King, J. 1971. Competition between established and newly sown grass species. J. Gr. Grassld. Soc. 26:221-229.  Kleter, H.J. 1968. Influence of weather and nitrogen fertilization on white clover percentage of permanent grassland. Neth. J . Agric. Sci. 16:43-52. Krajina,  V.J. 1965. Biogeoclimatic zones and classification Ecol. Western N. America 1:1-17.  of British Columbia.  Levin, S.A. and R.T. Paine. 1974. Disturbance, patch formation and community structure. Proc. Nat. Acad. Sci. (USA) 71:2744-2747. Loucks,  O.L., M.L. Plumb-Mentjes and D. Rogers. 1985. Gap processes and large-scale disturbances in sand prairies. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp. 72-83. Academic Press, Inc. Orlando.  MacDiarmid, B.N. and B.R. Watkin. 1971. The cattle dung patch. I. Effect of dung patches on yield and botanical composition of surrounding and underlying pature. J . Br. Grassld. Soc. 26:239-245. MacLusky,  D.S. 1960. Some estimates of the areas of pasture fouled by excretion of dairy cows. J . Brit. Grassld. Soc. 15:181-188.  the  Mahmoud, A. and J.P. Grime. 1976. An analysis of competitive perennial grasses. New Phytol. 77:431-435.  ability in three  Marsh,  dung.  Herbage  perennial  ryegrass  R.  and R.C. Campling. 1970. Abstracts 40:123-130.  Fouling of  patures  McNeilly,  T.  Miles,  1974. Effects of experimental interference with stand structure establishment of seedlings in Callenetum. J . Ecol. 64:745-755.  J.  and M.L. Roose. 1984. The distribution of genotypes in swards. New Phytol. 98:503-514.  by  on  Milton, W.E.J. 1940. The effect of manuring, grazing and cutting on the yield, botanical and chemical composition of natural hill pastures. J. Ecol. 28:326-356.  Literature cited / 154 Mundy,  E.J. 1961. The effect of urine and its components on the botanical composition and production of a grass/clover sward. J . Brit. Grassld. Soc. 16:100-105.  Newbery,  D. McC. 1979. The effects of decomposing roots on the grassland plants. J . Appl. Ecol. 16:613-622.  growth of  Norman, M.J.T. and J.O. Green. 1958. The local influence of cattle dung and urine upon the yield and botanical composition of permanent pasture. J . Br. Grassld. Soc. 13:39-45. Oliver, C D . and E.P. Stephens. 1977. Reconstruction of a mixed-species forest in central New England. Ecology 58:562-572. Orloci, L . 1978. Multivariate Analysis in Vegetation Research, 2nd ed. Junk, The Hague. Peet, R.K., D.C. Glenn-Lewin and J . Walker Wolf. 1983. Predictions of man's impact on plant species diversity. A challenge for vegetation science. In Man's Impact on Vegetation. W. Holzner, M.J.A. Werger and I. Ikusima (eds.) Junk, The Hague. Petersen,  R.G., H.L. Lucas and W.W. Woodhouse, Jr. 1956. The distribution of excreta by freely grazing cattle and its effect on pature fertility: I. Excretal distribution. Agron. J. 48:440-444.  Pickett,  S.T.A. and P.S. White. 1985. The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, Inc. Orlando  Pinder,  J . E . III. 1975. Effects of species community. Ecology 56:747-751.  Piatt,  removal  on  an  old-field  plant  W.J. 1975. The colonization and formation of equilibrium plant species associations on badger disturbances in a tall-grass prairie. Ecol. Monogr. 45:285-305.  Polley, H.W. and S.L. Collins. 1984. Relationship of vegetation and environment in buffalo wallows. Amer. Midi. Nat. 112:178-186.  Literature cited / 155 Putwain,  P.D. and J.L. Harper. 1970. Studies in the dynamics of plant populations. III. Influence of associated species on populations of Rumex  acetosa  L.  and  R.  acetosella  L.  in  grassland.  J. Ecol. 58:251-264.  Rabotnov,  T.A. 1966. Pecularities of the structure communities. Vegetatio 13:109-116.  of  Rabotnov,  T.A. 1974. Differences between fluctuations Handbook of Vegetation Science Part 8. R. Knapp (ed.) pp. 19-24. Junk, The Hague.  polydominant  meadow  and successions. In Vegetation dynamics.  Rapp, J.K. and D. Rabinowitz. 1985. Colonization and establishment of Missouri prairie plants on artificial soil disturbances. I. Dynamics of forb and graminoid seedlings and shoots. Amer. J . Bot. 72:1618-1628. Rhodes, I. and W.R. Stern. 1978. Competition for light. In 'Plant Relations in Pastures'. J.R. Wilson (ed.), pp. 175-189. CSIRO, Melbourne. Ross, M.A. and J . L . Harper. 1972. Occupation of biological space during seedling establishment. J . Ecol. 60:77-88. Rothwell,  P.I.  1977. The plant ecology of molehills. College of North Wales. Bangor, Wales.  M.Sc. thesis,  University  Runkle, J.R. 1981. Gap regeneration in some old-growth forests of the United States. Ecology 62:1041-1051.  eastern  Runkle, J.R. 1982. Patterns of disturbance in some old growth mesic forests of eastern North America. Ecology 63:1533-1546. Sagar, G.R. 1970. Factors controlling the size of plant populations. Proc. X. Brit. Weed Control Conference pp. 965-979. Sagar,  G.R.  and  J.L. Harper.  populations  of  1961.  Plantago  Weed Res. 1:163-176.  Controlled lanceolata,  P.  interference major  and  with  natural  P.  media.  Literature cited / 156 Sarukhan, J . and J.L. Harper. 1973. Studies on plant demography: Ranunculus repens L . , R. bulbosus L., and R. acris L. I. Population flux and survivorship. J . Ecol. 61:675-716. Silander, J.A. and J . Antonovics. 1982. Analysis of interspecific interactions in a coastal plant community a perturbation approach. Nature 298:557-560. Silvertown,  Snaydon,  J.W. 1981. Micro-spatial heterogeneity and seedling species-rich grassland. New Phytol. 88:117-128.  demography  in  R.W. 1971. An analysis of competition between plants of Trifolium repens L. populations collected from contrasting soils. J. Appl. Ecol. 8:687-697.  Snaydon, R.W. 1978. Genetic change in pasture populations. In Plant Relations in Pastures. J.R. Wilson (ed.) pp. 253-269. CSIRO, Melbourne. Snaydon,  R.W. 1984. Plant demography in an agricultural context. In Perspectives on Plant Population Ecology. R. Dirzo and J. Sarukhan (eds.). pp. 389-407, Sinauer Assoc. Inc. Sunderland, Mass.  Snaydon, R.W. 1985. Aspects of the ecological genetics of pasture species. In Structure and Functioning of Plant Populations 2. Phenotypic and Genotypic Variation in Plant Populations. J. Haech and J.W. Woldendorp (eds.). pp. 131-152. North-Holland Pub. Co., Amsterdam.  Snaydon,  Soane,  R.W. and C D . Howe 1986. Root established ryegrass and J . Appl. Ecol. 23:667-674.  I.D.  and shoot invading  and A.R. Watkinson. 1979. Clonal variation Ranunculus repens. New Phytol. 82:557-573.  competition between grass seedlings.  in  populations  of  Solangaarachchi, S.M. 1985. The nature and control of branching pattern in white clover (Trifolium repens L.) Ph.D. thesis, University College of North Wales, Bangor, Wales.  Literature cited / 157 Sousa,  W.P.  1979. Disturbance in marine non-equilibrium maintenance Ecology 60:1225-1239.  intertidal boulder of species  fields: the diversity.  Sousa, W.P. 1984a. The role of disturbance in natural communities. Ann. Rev. Ecol. Syst. 15:353-391. Sousa,  W.P. 1984b. Intertidal mosaics: patch size, propagule availability, and spatially variable patterns of succession. Ecology 65:1918-1935.  Spedding, C.R.W. 1971. Grassland Ecology. Clarendon Press, Oxford, pp. 221. Spittlehouse, D. 1985. Components of the seedling environment. In Interior Spruce Seedling Performance: State of the Art. Northern Silviculture Committee, Prince George, B.C.  SPSS.* (1986) SPSS.* User's Guide, 2nd. edition. SPSS Inc., Chicago. Stern, W.R. and C.H. Donald 1962a. Light relationships in grass-clover swards. Aust. J . Agr. Res. 13:599-614. Stern, W.R. and C.H. Donald 1962b. The influence of leaf area and radiation on the growth of clover in swards. Aust. J. Agr. Res. 13:615-623. Strong,  Tande,  D.R. Jr. 1983. Natural variability and the manifold ecological communities. Amer. Nat. 122:636-660.  mechanisms  of  G.F. 1979. Fire history and vegetation pattern of coniferous forests in Jasper National Park, Alberta. Can. J . Bot. 57:1912-1931.  Tansley,  A.G. 1935. The use and abuse Ecology 16:284-307.  Tansley,  A . G . 1939. The British Islands University Press, Cambridge.  of  vegetational  and  Their  concepts  Vegetation.  and terms.  Cambridge  Literature cited / 158 Tansley, A.G. and R.S. Adamson. 1925. Studies of the vegetation of the English Chalk. 111. The chalk grasslands of the Hampshire-Sussex Border. J . Ecol. 13:177-223.  Thompson, J.N. 1985. Within patch dynamics of life histories, populations, and interactions: selection over time in small spaces. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp. 253-264. Academic Press, Inc. Orlando.  Thorhallsdottir, T.E. 1984. The dynamics of a grassland community with special reference to five grasses and white clover. Ph.D. thesis, University College of North Wales, Bangor, Wales. Turkington, R. and L.W. Aarssen. 1984. Local scale differentiation as a result of competition. In Perspectives in Plant Population Ecology. J . Sarukhan and R. Dirzo (eds.) pp. 107-127. Sinauer Assoc., Cambridge, Mass. Turkington,  R., M.A. Cahn, A. Vardy and J . L . Harper. 1979. The growth, distribution and neighbour relarionships of Trifolium repens in a permanent pasture. Ill The establishment and growth of Trifolium repens in natural and perturbed sites. J. Ecol. 67:231-243.  Veblen, T.T. 1985. Stand dynamics in Chilean Nothofagus forests. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp. 35-51. Academic Press, Inc., Orlando. Veresoglou,  Vickery,  D.S. 1983. Partitioning of nutrients between co-existing plant species. D. Phil, thesis. University of York.  grassland  P.J. 1981. Pasture growth under grazing. In Grazing F.H.W. Morley (ed.). pp. 55-77. Elsevier, Amsterdam.  Animals,  Vitousek, P.M. 1985. Community turnover and ecosystem nutrient dynamics. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp.325-333. Academic Press, Inc. Orlando. Watkin,  B.R.  and R.J. Clements. 1978. The effects of patures. In Plant Relations in Pastures. pp. 273-289. CSIRO, Melbourne.  grazing animals on J.R. Wilson (ed.)  Literature cited / 159 Watt, A.S. 1925. On the ecology of the British beechwoods with special reference to their regeneration. Part II, Section II and III. The development and structure of beech communities on the Sussex Downs. J. Ecol. 13:27-73.  Watt,  A.S.  1957. The effect of excluding rabbits from (Mesobrometum) in Breckland. J . Ecol. 45:861-878.  grassland  B  Watt, T.A. and R.J. Haggar. 1980. The effect of height of water table on the growth of Holcus lanatus with reference to Lolium perenne. J . Appl. Ecol. 17:423-430. Weeda,  Welch,  W.C. 1967. The effects of cattle dung patches on pasture growth, botanical composition, and pasture utilization. N.Z. J. Agric. Res. 10:150-157. D. and M . Rawes. 1964. high-level grasslands Ecol. 1:281-300.  The early in the  effects of excluding north Penniones.  sheep from J. Appl.  White, J . 1985. The census of plants in vegetation. In The Population Structure of Vegetation. J . White (ed.). Junk, The Hague. White, P.S. and S.T.A. Pickett. 1985. Natural disturbance and patch dynamics: an introduction. In The Ecology of Natural Disturbance and Patch Dynamics. S.T.A. Pickett and P.S. White (eds.) pp. 3-13. Academic Press, Inc. Orlando. Whyte, O.R. 1974. Tropical Grazing Lands: Communities and Constituent Species. Junk, The Hague.  8. APPENDIX 1  160  Transition pasture  probability  for  the  period  of  row  from  species September  Molehill  Dactylis  glomerata  replaced  1984  to  Dactylis glomerat  Molehi11  being  33 .9  3 .5  a  April  by  Holcus  Lolium  /anatus  perenne  0 . .4  a  column  species,  on  and  around  molehills  in  the  1939  1985.  Phleum pratensecompress  8 . 1  0 . .4  PoaAgropyronTrifolium  Taraxacum  a  repens  repens  18 . 4  2.8  1 8 ..4  Ranunculus acr  i s  3 .. 9  1 .4  0 .,0  91 . 7  0 ,. 0  0 .0  0 ,,0  5. . 0  1 .,7  0 ..0  0.0  Holcus  lanatus  0 .. 0  8 .5  42, 6  12 .8  0 ,,0  1 9 .. 1  0 ..0  1 0 . .6  2.. 1  0.0  Lolium  perenne  4 .1  2 .9  1 .2  43 .9  0 ,.6  17 . 5  0 . .6  19 . 9  2. . 9  0 ,. 0  10 . 0  10 . 0  10 . 0  0 ..0  0 .0  0 ..0  4 0 .0  1 0 .. 0  Phleum Poa  pratense compressa  1 .. 7  officinale  2: 3 0.0  1 ..2  2 .4  2, .4  6 .0  0 .,0  59 . 5  1 .2  1 6 .. 7  3 .. 6  3.6  Agropyron  repens  10, O  10 . 0  0 ,, 0  10 .O  0, 0  10 .O  3 0 .. 0  1 0 .. 0  0 ..0  0.0  Trifolium  repens  4 ,. 5  4 .. 9  0 , .4  6 .9  0, 8  1 5 ..8  2 . .4  5 7 .. 1  3 . ,4  0 ,. 0  3 , .4  6. . 9  0 .,0  13. 8  6.. 0  3 . .4  62 . 1  6.9  0 ,.0  0 ,, 0  0 ..0  0 ..0  0. 0  0 ..0  14 . 3  1 4 . .3  0. 0  71 . 4  Taraxacum Ranunculus  officinale acris  1 .6 .  1 .2  Transition pasture  probability  for  the  period  of  row  from  species August  Molehill  being  1983  Dactylis glomerata  Molehi11  Dactylis  3 7 ..8  glomerata  Holcus  lanatus  Lolium  perenne  Phleum Poa  pratense compressa  6 .4  column  species,  on  and  around  molehills  in  the  1958  Phleum  PoaAgropyronT  pratensecompressa  ri fo!  repens  1 .. 2  0 ,.0  12 . 9  ium  repens  Taraxacum  Ranunculus acr  officinale  i s  2 0 ..5  7 .2  4.0  O.O 0.0  3 ,5  5. . 2  1 5 .. 7  6., 1  2 .6  0.9  5. .8  19 . 8  1 0 .. 5  5. .8  9.3  1 .2  2 .0  9 . .8  8.8  2 2 ,.5  1 2 . .7  8 .8  2.9  4.9  12 ..5  0 .0  0 ,.0  5 0 .. 0  12 ,, 5  0 ,. 0  0 .. 0  6.3  0.0  4 , .2  3 .4  4 , .2  5. . 1  3 3 ,. 9  1 2 ,. 7  7 .6  7.6  0.0  5 ..7  3, . 3  3,, 3  3.. 3  2 3 ..6  3 8 .. 2  1 .6 .  2.4  1 .6  1 0 ,.4  3 , ,6  0 ..0  1 0 .. 7  0 ..0  4 6 , .4  7. 1  0.0  10 .8  10, 8  12 .5 14 .4  0 .. 0  a  4 ., 7  1 ., 7 2 5 ,. 6  repens officinale  Lolium perenne  2 ,.3  Trifolium  acris  Holcus  5 6 ,.5  10. 6  by  1984.  /anatus  5. , 2  repens  Taraxacum  replaced April  8 ., 1  Agropyron  Ranunculus  2, . 0  to  7 .. 1  1 3 ,, 3  2. ,2  4 . ,4  4 ,,4  2. .2  1 5 ..6  2 2 . .2  6..7  26.7  0.0  0 ,,0  0 ..0  0 ..0  5 0 ,, 0  0. 0  0 ..0  2 5 ..0  0 ,.0  25 .0  O.O  CT)  ro  Transition pasture  probability  for  the  period  of  row  from  species August  M o l e h i 1 1 Dactylis  g1omerata Molehi11  Dactylis  glomerata  being  1984  replaced  to  April  Ho  1cus  1anatus  by  a  column  species,  on  and  around  molehills  in  the  1977  1985.  Lolium perenne  Ph1eum pratensecompress  PoaAgropyronT a  rifolium  repens  repens  Taraxacum officinale  Ranuncu1 acr  us i s  27 .5  4 .1  1 1. 9.  6. .9  5 . 5  1 1 .9  2. . 3  15 . 1  5. . 0  3.3  0 .0  44 .7  7 ..9  5. .3  13 . 2  13 . 2  0 .0  5 . 3  5, . 3  0.0  Holcus  lanatus  0 .0  4 , .4  7 7 ..8  4 ..4  0 .0  4 .4  0 .. 0  6..7  0 ,. 0  0.0  Lolium  perenne  3 .4  1 .7  1 9 .. 0  41 ..4  3. . 4  13 .8  0 ..0  6..9  1 .. 7  0.0  0 .0  14 . 3  0 ..0  57 . 1  28 .6  0 .0  0 .. 0  0 .0  0 .. 0  0.0  5 .0  10 . 0  1 0 .. 0  5, . 0  2, . 5  40 .0  2 .5  7.. 5  7.. 5  0.0 0.0  Phleum Poa  pratense compressa  Agropyron  repens  20 .0  0 .0  0 ..0  2 0 .. 0  0 .0  20 .0  0 .. 0  4 0 .. 0  0 .. 0  Trifolium  repens  0 .0  3 .9  9 .. 0  1 .3 .  2 .6  5. . 1  1 .3 .  6 9 . .2  1 ,. 3  14 . 3  4 . .8  0 ..0  4. 8  9 .. 5  19 . 0  0 ..0  4 ..8  4 2 ..9  0.0  0 .O  0 ,.0  O. 0  0. 0  0 .. 0  0 .,0  0. O  0 .,0  0. O  100.0  Taraxacum Ranunculus  officinale acris  1 . 3  CO  Transition pasture  probability  for  the  period  of  row  August  Dactyl  is  g1omerata Dactylis  glomerata  species  from  1983  Holcus /anatus  being to  replaced April  Lolium perenne  by  a  column  species,  on  and  around  dung  in  the  1939  1984.  Phleum pratense  Poa compressa  Agropyron!rifolium repens  Taraxacum  repens  of f i c i  nale  Ranunculus acr  Dung  i s  35 . 0  0 .. 0  15 . 0  0 .0  25 .0  0 .0  10 . 0  5. . 0  1 0 .. 0  0 .. 0  Holcus  lanatus  2 .1  1 2 ..8  1 4 .. 9  2 .1  36 . 2  0 .0  21 ,. 3  6 , .4  4 ,. 3  0 ., 0  Lolium  perenne  1 .7  0 . .8  4 0 , ,9  0 ,. 0  34 .6  0 . .4  1 0 ,. 1  8 ..4  1 ..3  0. O  0 .0  0 ..0  0 .. 0  1 0 0 ,. 0  0 .0  0 .. 0  0 ,.0  0 ,,0  0 ,.0  0. O  2 .0  1 ..0  3 0 . .6  0 ,. 0  3 5 ,. 7  2, . 0  14 ,, 3  4 .. 1  4 ., 1  0 .,0  Phleum Poa  pratense compressa  Agropyron  repens  0 .. 0  0 ..0  3 5 .3  0 .,0  2 9 .4  11 , 8  5 . .9  1 1 8.  5 . .9  0. 0  Trifolium  repens  1 .4  0 ..0  2 3 , .2  0 ,. 0  3 3 .3  1 .4  3 0 ,.4  4 . .3  2, . 9  0 .. 0  0 .O  0 .,0  2 2 .. 2  0 ,. 0  25 . 9  0 ,. 0  3 .. 7  4 4 ,.4  O. O  0. O  1 6 .. 7  0 ..0  5 0 ,. 0  0 ,, 0  3 3 .3  0 .. 0  0 ,,0  0 ..0  0 ..0  0 .,0  0. 0  2 6 . .9  0 .,0  3 4 ,. 3  3 ., 0  17 ..2  6 . .7  1 ,5 .  8 .2  Taraxacum Ranunculus Dung  officinale acris  0 .. 7  Transition pasture  probability  for  the  period  of  row  from  September  Dactylis g1omerata Dactylis Lol  ium  Phleum Poa  being 1983  replaced  to  Holcus  Lolium  /anatus  perenne  April  by  a  column  species,  on  and  around  dung  in  the  1958  1984.  Phleum pratense  Poa compressa  AgropyronTrifolium  Taraxacum  Ranunculus acr  officinale  Dung  i s  repens  repens  7 5 , ,8  0 ..0  1 .6  0 ,.0  1 2 .,9  1 .6 .  3 . ,2  1 .,6  0 .,0  0 ,. 0  lanatus  1 .4 ,  3 2 . .9  1 .4 .  1 .4 .  43 , 8  4 .. 1  8 . .2  1 .4 .  0. 0  0 .. 0  perenne  4 .9  2 .,4  41 . 5  2. .4  3 4 .. 1  7..3  4 , ,9  0 .0  0 .. 0  0 ,. 0  0 .0  0 ..0  0 .0  88. 9  0 . .O  O. . 0  11 .. 1  0 .. 0  O .. 0  0 .0  3, . 1  0 .,0  5. .4  4 ,. 7  6 8 ..2  1 0 .. 1  3.. 1  2, . 3  0 ..0  0 .. 0  glomerata  Holcus  species  pratense compressa  Agropyron  repens  5, .9  2. .9  5. ,9  8 . .8  14 ., 7  5 2 .,9  0 ..0  0 ..0  0 .,0  0 .. 0  Trifolium  repens  0 ,.0  O .. 0  5. . 3  0 ..0  21 ., 1  5 .. 3  6 3 . .2  O. , 0  0. O  0 .. 0  81 .8  0 ..0  0 .. 0  0 .. 0  6 0 ..0  0 .. 0  6..2  1 .4  3 2 . .4  Taraxacum Ranunculus Dung  officinale acris  0 .0  0 ..0  9 .. 1  0 ,,0  0 .. 0  0 ,.0  O, . 0  0 ..0  0 .. 0  0 ,.0  0 ..0  0 ..0  8 . .3  2. 8  2. 8  2. . 1  2 6 ..2  1 1, 7.  9 .. 1 2 0 .. 0 4 .. 1  CTl CJ1  Transition pasture  probability  for  the  period  of  row  from  species  September  Dactylis g1omerat Dactylis  glomerata  Holcus Lol  lanatus  ium  perenne  Phleum Poa  pratense compressa  a  being 1984  replaced to  Holcus  Lolium  1anatus  perenne  April  by  a  column  species.  on  and  around  dung  in  the  1977  1985.  Phleum pratense  Poa compressa  AgropyronTrifol repens  ium repens  Taraxacum officinale  Ranuncu1  us  acri  Dung  s  0.. 0  4,.4  2 .2  4 .4  2 .2  6.. 7  2 ..2  0 ,. 0  2.2  7 ,. 5  7 3 ,, 1  1 .. 5  1 .5  3 .0  1.5  1 0 .. 2  O .. 0  O. . 0  0.0  3,. 1  1 2 .. 5  5 3 .. 1  0 .0  6 .3  0 .0  21 .. 9  3.. 1  0 .. 0  0.0  16 .. 7  0 .. 0  0 ., 0  50 .0  0 .0  0 .. 0  16 ., 7  0 ..0  0 .. 0  0.0  7 5 , .6  1 .. 9  5 ..6  13 ., 0  0 .0  42 .6  0 .. 0  2 0 . .4  3 .,7  0 .. 0  Agropyron  repens  0 .. 0  0 .. 0  2 2 .2  22 .2  0 .0  11 . 1  2 2 . .2  0 .,0  11 .. 1  0.0  Trifolium  repens  9 .. 8  9. 8  5 .. 2  3 .3  11 .8  0 .. 7  5 4 ..9  2 . .6  0 .. 7  0.7  4 ..5  0 ..0  9 .. 1  4 .. 5  9 .1  0 .. 0  1 3 ..6  5 4 . .5  4 . ,5  0.0  0 .. 0  0. 0  0 ..0  0 .. 0  0 .. 0  0 ., 0  3 3 . ,3  16 . 7  5 0 .. 0  0.0  8 ..2  5. 4  1 6 .,3  6 .. 1  1 1.6.  4 ., 1  3 2 . .7  5. 4  1 ,4 .  3.4  Taraxacum Ranunculus Dung  officinale acris  1 .9  6S cn  Transition pasture  probability  for  the  period  of  row  from  species  September  Dactylis glomerat Dactylis  glomerata  Ho a  being 1983  1cus  1anatus  replaced  to  Loli  um  perenne  October  by  a  column  species,  on  and  around  dung  in  the  1939  1983.  Ph1eum pratense  Poa compressa  AgropyronTrifol repens  ium repens  Taraxacum  of f ic  inale  Rani/ncufus  Dung  acris  9 7 .9  0 .. 0  0, . 0  2 .1  0 .0  0 .. 0  0 .0  0 .0  0 .. 0  Holcus  lanatus  0 .0  87 . 2  2. 6  0 .. 0  5 . 1  0 ..0  2 .6  0 .0  0 .. 0  2.6  Lolium  perenne  0 .0  0 . 7  90 . 1  0 .0  4 .6  0 . 7  2 .0  0 .0  0 .. 7  0.7  0 .0  0 ..0  0 ..0  100 .0  0 .0  0 ..0  0 .0  0 .0  0 ,. 0  0.0  0 ., 0  1 .. 7  5. . 0  0 . .8  8 5 .. 1  2 . .5  1.7  0 .. 0  0. 8  0.8  Phleum Poa  pratense compressa  0.0  Agropyron  repens  0 .. 0  1 5 ..4  15 ..4  0 .. 0  0 .. 0  6 9 ,,2  0 .. 0  0 .. 0  0 ..0  0.0  Trifolium  repens  0 ..0  0 ..0  1 .9 .  0 ..0  1 .4 .  0 .,0  9 4 ..3  0 ..0  1. . 0  0.0  3 ., 0  0 ..0  0 ..0  0 .,0  0 ..0  0 .,0  3.. 0  9 0 . .9  0 ..0  0.0  0 ..0  0. O  0. O  0 ..0  0 .. 0  0 ..0  O, O  0 ., 0  1 0 0 .. 0  0.0  2 . ,5  7 , ,6  2. ,9  1. 4  60.5  Taraxacum Ranunculus Dung  officinale acris  0 .. 7  0 ..0  1 0 . .9  0 . .4  1 0 .. 9  CTl  Transition pasture  probability  for  the p e r i o d  of  row  from  species  Molehill  Dactylis  75 .6  glomerata  replaced  1983  to October  Dactylis glomerat  Molehi11  being  September  a  0, . 7  by  Holcus  Lolium  /anatus  perenne  0 .0  a  column  species,  and  around  molelills  i n the  1939  1983.  Phleum pratensecompress  PoaAgropyronTrifol a  repens  ium repens  Taraxacum officinale  Ranunculus acr  is  5. .3  0 .0  8 .1  0 .4  5. . 3  1 .8  0.0  95 ,.0  0. .0  0. .0  0. .0  0 .0  0 .0  0. .0  0.0  0.0  Holcus  lanatus  4, , 3  2. . 1  70 .2  10.,6  0, ,o  6 .4  0 .0  2. . 1  0.0  0.0  Lolium  perenne  10 . 5  0. .0  0 .6  79, .5  0, .6  4 .1  0, .6  2 .9  1 .2  0.0  10 .0  0, .0  0. .0  0. .0  60, .0  20 .0  0. .0  0. .0  0.0  0.0  8 .3  0. .0  0, ,0  3 .6  0. .0  76 . 2  4 ,.8  4. 8  0.0  0.0  Phleum Poa  pratense compressa  5 .0  Agropyron  repens  0. ,0  10..0  0. .0  0. ,0  0, ,0  0 .0  7 0 . .0  1 .0 .  O.O  0.0  Trifolium  repens  9, .3  O,.4  0 .4  2, .0  0, .0  2 .4  0 .4  8 4 . .2  O.O  0.4  6, .9  0. .0  0, .0  3..4  0, .0  6, .9  0. .0  3. .4  14,, 3  0. 0  0. .0  0. ,0  0. .0  0. .0  0. .0  O..0  Taraxacum Ranunculus  officinale acris  • 79.3 0.0  0.0 85.7  CTl CO  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items