Tipos de Ingeniería del Software

traits to make management decisions 3.1 Introduction – using life-history traits to make management decisions Life cycles

To understand why one species is successful in a particular location while another is not, we often need to focus on the sequence of events that occur in the organism’s life cycle. At its simplest, the life cycle of a plant or animal comprises birth, followed by a pre-reproductive juvenile period, a period of reproduction, possibly a post-reproductive period, and then death (although various causes of mortality may intervene at any time). Some species squeeze several or many generations into a single year, some have one generation each year (annuals), but others (perennials) have life cycles that last for several or many years. Many plants and some animals (such as the fairy shrimp Streptocephalus vitreus) spend part of the year in a dormant phase (e.g. as seeds or eggs), and their ‘seed-banks’ can persist for years. All these features help determine the success of a particular species in a particular type of environment.

Species traits

Beyond contrasts in the characteristics of their life cycles, species differ in many other fundamental ways. Individuals may be small or large, invest few or many resources in their offspring (in seed biomass or parental care), have a short or long juvenile period, grow fast or slowly, and be more or less vulnerable to various sources of mortality. Some species are very specialized in their requirements, while at the other extreme are animals that are good at learning to exploit novel resources or plants with plastic tolerance limits that allow them to succeed in a range of circum- stances. Species also vary in their competitiveness for limited resources, their tolerance of physi- cally stressful conditions or of environmental disturbances, and their ability to fi nd and exploit new habitats.

Box 3.1 Essential life-

history theory

Box 3.1 Essential life-

LIFE-HISTORY THEORY AND MANAGEMENT CHAPTER 3 61

Several schemes have been developed that link particular groups of traits to particular kinds of environments (see Chapter 4 in Begon et al., 2006, for a detailed treatment). Two of these schemes will fi gure here – the r/K and CSR concepts.

The r /K concept

The potential of a species to multiply rapidly – producing large numbers of progeny early in life – is advantageous in environments that are short-lived (created, for example, by a disturbance such as the falling of a forest tree, or the storm-battering of a rocky reef), allowing the organisms to quickly colonize and exploit the new habitat. Such species have been called r-species because they spend most of their lives in a near-exponential phase of population growth where their intrinsic rate of increase (r) is being fully expressed. The habitats where they are favored have been called r-selecting.

At the other end of the scale are organisms with life histories that enable them to survive where there is often intense competition for limited resources. In this case the individuals leaving most descendants are those that capture a larger share of resources. They are called K-species because they spend most of their lives bumping up against the environment’s carrying capacity (K ). Habitats where they are favored have been called K-selecting.

The r/K concept (MacArthur & Wilson, 1967; Pianka, 1970) can be useful in the interpretation of contrasting patterns in nature. For example, many forest trees are excellent examples of K- species. They compete for light in the canopy and the survivors are those that put their resources into growth so they can overtop their neighbors. They usually delay reproduction until their branches have secured a place in the forest canopy, and hold on to their position and live for a very long time. Overall, they make a relatively low allocation to reproduction but many produce large, well-provisioned seeds. By contrast, in the more disturbed circumstances of r-selecting habitats, plants tend to conform to a contrasting group of r characteristics: a greater reproductive allocation, but smaller seeds, smaller size, earlier reproduction and a shorter life.

The CSR concept

Grime (1974, Grime et al., 1988) produced a different, but not unrelated, classifi cation of habitats and plant life histories. Habitats are seen as varying in two signifi cant ways – in their level of dis- turbance (brought about by grazing, disease, trampling or adverse weather) and in the extent to which they experience ‘stress’ (shortages of light, water or nutrients that limit photosynthesis). Grime argued that a stress-tolerant strategy (S) is appropriate when ‘stress’ is severe but disturbance is uncommon. Conversely, a so-called ruderal strategy (R) is appropriate when dis- turbance levels are high but conditions are benign and resources abundant (R-species are essen- tially good colonizers). Finally, a competitive strategy (C) is appropriate when disturbance is rare, resources are abundant and crowded populations develop. Grime then classifi ed a large number of plants on the basis of ecological characteristics he thought suited them for one or other of these strategies (C, S or R), or some intermediate combination (CR, CS, SR or even CSR). Grime suggests that strong competitors have a high relative growth rate, the ability to spread by vegetative means and a tall stature. Stress tolerators, on the other hand, are small in stature with a low relative growth rate. Finally, ruderal species are generally annuals or short-lived herbaceous perennials with a capacity for rapid seedling establishment and growth, and a tendency for a high proportion of photosynthate to be directed into seeds. Intermediate species possess combinations of these traits.

Native biodiversity is often compromised as a result of human activity, whether by mining, agriculture, forestry or urban development. The desire to restore these places to something approaching their pristine state may lead, through a political process, to legislation or economic incentives that foster recovery. In other cases, economic circumstances change and previous activities are no longer sustainable – mines close, agricultural land is put out of production, forestry is no longer viable. Whatever the background, effective restoration needs to be based on knowledge of 3.2 Species traits as predictors for effective restoration 3.2 Species traits as predictors for effective restoration

62 PART 1 ECOLOGICAL APPLICATIONS AT THE LE VEL OF INDIVIDUAL ORGANISMS

which species will do well. Let’s consider examples of agricultural restoration where the end-points are very different – species-rich grassland (Section 3.2.1) or forest (Section 3.2.2).

Undisturbed native grasslands are actually quite rare, most having been converted to an unnatural state by pastoral farming or agricultural development. Thus, the restoration of native grasslands has become a priority.

The fi rst study of any ecological phenomenon provides just so much new informa- tion that it is risky to derive general conclusions that go beyond the focal organisms and sites. As science progresses, however, studies accumulate in the scientifi c litera- ture until a more effective search for generalizations can be made using a ‘meta- analysis’. Pywell et al. (2003) assembled the results of 25 published experiments dealing with the restoration of species-rich European grasslands from land that had previously been ‘improved’ for pasture (ploughed, planted with pasture species and fertilized) or used for arable farming. Restoration involved sowing a range of desir- able species into areas where agricultural intensity had been reduced. The idea was to relate the performance of different species to their life-history traits. On the basis of the results of the fi rst 4 years of restoration, they calculated a performance index for commonly sown grasses (13 species) and forbs (45 species; forbs are defi ned as herbaceous plants that are not grass-like). The index, calculated for each of the 4 years, was simply the percentage of quadrats (0.4 × 0.4 m or larger) originally sown with a species that still retained it. Their life-history analysis included 38 plant traits, including longevity of seeds in the seed-bank, seed viability, seedling growth rate, Grime’s life-history strategies (C, S or R – Box 3.1), and the timing of life-cycle events (germination, fl owering, seed dispersal).

The best performing grasses include Festuca rubra and Trisetum fl avescens (per- formance indexes averaged for the 4 years of 77%); and among the forbs, Leucanthe- mum vulgare (50%) and Achillea mellefolium (40%) are particularly successful. In fact, grass species in general do better than forbs in restoration attempts, but for grasses only ruderality (R) is positively correlated with performance. With the forbs, on the other hand, good establishment is linked to colonization ability, percentage germi- nation of seeds, autumn germination, vegetative growth, seed-bank longevity and habitat generalism. Interestingly, competitive ability and seedling growth rate become increasingly linked with success as the intensity of competitive interactions increased with time (Table 3.1). Stress tolerators, habitat specialists and species of infertile habitats perform badly, partly refl ecting the high residual nutrient availabil- ity in many restored grasslands.

Pywell’s team notes that restoration effi ciency could be increased in future by only sowing species with the identifi ed ecological traits. However, because this would lead to uniformity among restored grasslands, they also suggest that desirable but poorly performing species could be assisted by phased introduction several years after restoration begins, when environmental conditions are more favorable for their establishment.

Surprisingly large areas of the tropics now consist of abandoned agricultural land. Close to the Panama Canal, for example, most of the biodiverse tropical forest had been cut and burned to open up land for subsistence agriculture by the mid 1980s. Small farms were quickly abandoned, however, probably because of low productiv- 3.2.1 Restoring grassland plants – a pastoral duty 3.2.1 Restoring grassland plants – a pastoral duty 3.2.2 Restoring tropical forest – abandoned farmland reclaimed for nature 3.2.2 Restoring tropical forest – abandoned farmland reclaimed for nature

LIFE-HISTORY THEORY AND MANAGEMENT CHAPTER 3 63

ity. Now the areas have been invaded by tall, impenetrable stands of the exotic grass Saccharum spontaneum, a serious invader in places as far apart as India, the Philippines and Puerto Rico.

As was the case with native grass seeds in Section 3.2.1, the lack of naturally arriving tree seeds is a major barrier to unassisted regeneration in Panama. However, it would be wasteful simply to throw seeds in an indiscriminate manner into these areas. Hooper et al. (2002) tested the potential of seeds of 20 native trees to suc- cessfully germinate, survive and grow within the Saccharum-dominated community (Figure 3.1). Two kinds of species trait have particular signifi cance in determining tree seed success in the Saccharum grassland – seed size (small: <0.15 g; medium: 0.15–1.0 g; large: >1.0 g) and shade tolerance (low, medium and high; determined from separate germination experiments in a greenhouse where light levels were varied). To demonstrate the major patterns, I present results separately according to these different traits.

Species with the smallest seeds generally perform most poorly. Those considered ‘pioneer’ species, so named because they are characteristic of open and disturbed locations (and, thus, have low shade tolerance), do a little better than those with intermediate or high shade tolerance, but nowhere near so well as species with medium seeds (in the intermediate shade tolerance class) or large seeds (in both medium and high shade tolerance classes) – just compare their respective perform- ance indexes in Figure 3.1.

You might suppose, from fi rst principles, that the r-selected, small-seeded pioneer trees (which include Trema micrantha and Jacaranda copaia) would be ideal candi- dates for restoration, because of their good powers of colonization and rapid growth in disturbed situations. However, in Panama at least, they do not tolerate the constraints imposed by the Saccharum grass.

At the other end of the scale, species with large or very large seeds (2.9 to a massive 50.4 g each) include the moderately shade tolerant Dipteryx panamensis and the very shade tolerant Calophyllum longifolium, Carapa guianensis and Virola surinamensis. The latter three germinate immediately upon planting in the wet

Trait N Year 1 Year 2 Year 3 Year 4

Ruderality (colonization ability) 39 _+*

Autumn germination 42 _+*

% germination 43 _{+** +* +*}

Seedling growth rate 21 _{+* +** +*}

Competitive ability 39 _{+* +** +*** +***}

Vegetative growth 36 _{+** +* +* +*}

Seed-bank longevity 44 _{+* +* +* +*}

Stress tolerance 39 _{−** −** −*** −***}

Generalist habitat 45 _{+** +** +** +**}

N, number of species in analysis.

+, Positive relationship; −, negative relationship. A positive sign means that the trait in question is associated with good performance in grassland restoration.

Asterisks indicate the level of statistical signifi cance of the results for each trait in each year – more asterisks mean a greater level of confi dence in the result: *p < 0.05, **p < 0.01, ***p < 0.001, blank not signifi cant. Thus, for example, species with good competitive ability are highly likely to be successful in restoration experiments (particularly in years 3 and 4). Stress tolerators, on the other hand, do very poorly.

Table 3.1 Ecological

traits of forbs (nongrass herbaceous plants) that showed a signifi cant relationship with successful performance of the plants in years 1–4 after sowing in grassland restoration experiments. (After Pywell et al., 2003.)

Table 3.1 Ecological

traits of forbs (nongrass herbaceous plants) that showed a signifi cant relationship with successful performance of the plants in years 1–4 after sowing in grassland restoration experiments. (After Pywell et al., 2003.)

64 PART 1 ECOLOGICAL APPLICATIONS AT THE LE VEL OF INDIVIDUAL ORGANISMS

season and, despite slow growth, become relatively tall by the fi nal census (>10 cm), while Dipteryx has dormant seeds that germinate the following dry season. All four of these highly competitive species survive well and have good performance indexes.

Hooper and colleagues make several suggestions for restoration management. First and foremost they note that Saccharum grassland is prone to fi re, which would be a major barrier to tree regeneration. Fire breaks need to be created as an integral part of restoration. Even in the absence of fi re, however, natural regeneration will not occur because small seeds – which are generally transported the longest

50

0

7

0 Intermediate shade tolerance

% germination % survival

Growth rate Performance index

50

0

7

0 Low shade tolerance

% germination % survival

Growth rate Performance index

65

50

0

7

0 High shade tolerance

Small Medium Large

Seed size

% germination % survival

Growth rate Performance index

75 Fig. 3.1 Performance of seeds of 20 tropical trees planted in Saccharum spontaneum grassland in Panama. The species have been grouped in three seed classes (small, medium and large) and three shade tolerance classes (low, which require very high light levels to germinate, medium, and high, which are able to germinate in very low light levels). Different classes are represented by different numbers of trees, and none of the trees fell into the medium seed, low shade tolerance grouping. The different histograms relate to percentage germination, percentage survival of germinated seeds, an index of growth rate (cm per day from May to July multiplied by 1000) and an overall performance index (which is the product of the other three: i.e. % germination × % survival × growth rate). Species whose seedlings did not survive have a growth rate and performance index of zero. Note how species with medium or large seeds generally do best. However, small-seeded species that are intolerant of shade also do quite well. (Based on data from Hooper et al., 2002.)

LIFE-HISTORY THEORY AND MANAGEMENT CHAPTER 3 65 distances and therefore are most likely to arrive at a site – are unlikely to produce viable seedlings in the grassland setting. Thus, the larger seeds of species most likely to do well need to be collected from elsewhere and carried in to the site. But Hooper’s team also discovered that shade cover greatly enhances tree seedling success – so they suggest artifi cial shade should be provided at fi rst. The establishment of the hand-planted species can be expected to catalyze forest regeneration, soon produc- ing its own shade and assisting other species to establish. To increase biodiversity once natural shade has been established, small-seeded species in the shade-tolerant class (including Genipa americana and Heisteria concinna) can be introduced. A number of species, plants and animals, have invaded widely separated places on the planet, including the shrub Lantana camara (Figure 3.2), the starling Sturnus vulgaris and the rat Rattus rattus. This prompts the question – do successful invaders share traits that raise the odds of successful invasion (Mack et al., 2000)? If it were possible to produce a list of traits associated with invasion success, managers would be in a good position to assess the probabilities of successful invasions, and thus to prioritize potential invaders and devise appropriate biosecurity procedures (Wit- tenberg & Cock, 2001).

Here I take advantage of databases available for taxonomic groups where some species are known to be successful invaders, and others not. These can be used to shed light on the species traits associated with a high probability of invasion success. My focus will be on pine trees in Section 3.3.1, New Zealand birds in Section 3.3.2 and parrots, plants and fi sh in Section 3.3.3.

3.3 Species traits as