p205

Large scale models

See K. M. Golden, “Climate Change and the Mathematics of Transport in Sea Ice,” Notices AMS  56 , 562 (2009).

Ecology and surprise

The following paper very frankly admits that it is usual in ecology that something quite unexpected does happen. The conclusion is: ecological prediction requires imagination and modesty. 

Doak et al., UNDERSTANDING AND PREDICTING ECOLOGICAL DYNAMICS: ARE MAJOR SURPRISES INEVITABLE

Ecology  89 , 952 (2008) 

Ecological surprises (=a substantial change in the abundance of one or more species resulting from a previously unknown or unanticipated process of any kind) are a common outcome of both experiments and observations in community and population ecology. 

*Truly surprising results are common enough to require their consideration in any reasonable effort to characterize nature and manage natural resources. 

*How best to choose the model or models for predicting population and community dynamics, and how best to then define and present the uncertainties in these predictions, have been active and contentious topics in statistical and ecological research. Implicit assumption is that models include some reasonable characterization of the relevant and important ecological processes, thus providing qualitatively accurate predictions. We contend that this assumption is not well-founded in experience: that the extent and frequency of major ``surprises'' in ecological systems argue for substantial humility about our predictive abilities, and that current effort to enumerate uncertainties must be better tempered with the recognition that ecological models fail to capture many instances of population and community dynamics.

SURPRISES ARE COMMON AND EXTREME 

*True ``surprises'' often have broad implications, extending geographically, taxonomically, or across multiple ecological systems. Even more importantly, real surprises almost always

occur in the presence of clear knowledge and apparent understanding, rather than due to simple ignorance. That is, an ecological surprise occurs when an experienced biologist with clear, well-informed expectations faces outcomes or patterns that strongly contradict these

expectations.

*Population levels are unstable. Young (1994): relatively stable numbers of medium and large-bodied mammals through time was manifestly false. Wolda (1978): populations of insects in the tropics are just as variable as populations of temperate species. These are in part due to misperceptions about species interactions.

*The extremely rapid increase and spread of spruce bark beetle in south-central Alaska and the Yukon Territory during the 1990s killed over 1.19 million ha of mature white spruce trees in Alaska alone (Matsuoka et al. 2006). The onset, speed, and extent of this outbreak were not anticipated. Substantial warming during this time period is now believed to be responsible for the beetle increases, but obtaining clear support for this explanation took many years of study (Berg et al. 2006).

*Unexpected resource pulse: In 1992 large numbers of smooth lumpsuckers (Aptocyclus ventricosus) suddenly and unexpectedly appeared in coastal waters of the central and western Aleutian archipelago. Consequently, the normally large number of beach-cast carcasses of otters dying from starvation (Laidre et al. 2006) disappeared entirely.

* In the early 1990s, killer whale sightings by researchers working on sea otters and kelp forests in the Western Aleutian archipelago rose from less than one sighting per year to multiple sightings per day. Although the sea otter population collapse predictably led to a collapse in the kelp forest ecosystem, the proximate cause of this change, arrival of a new top predator and a novel feeding behavior, was entirely unanticipated and its ultimate cause remains both uncertain and highly contentious (for a parallel case, see Roemer et al. [2002]).

*Basic assumptions about the patterns in community composition formulated from observations of current assemblages often cannot explain past community patterns.

*Fir wave is still not entirely clear.

*Unintended consequences of management actions are so pervasive only a few mentioned:

i) Prey-predator switch: For reasons that remain uncertain, the lobsters disappeared from Marcus Island in the early 1970s. The lobsters preyed on predatory whelks and whelk populations apparently increased substantially following the lobsters' disappearance. After a 9-month caging experiment demonstrating that lobsters were indeed capable of surviving at Marcus Island, 1000 lobsters were reintroduced in an effort to reestablish the species. However, these lobsters were immediately attacked and consumed by their previous prey, the now overabundant whelks.

ii) Indiscriminant control of coyotes can lead to identical or even higher livestock depredation rates, due to a variety of factors (reviewed in Mitchell et al. 2004).

iii) Control of red foxes in order to increase Red Grouse populations often backfires, with no reduction in numbers or increased cycling of grouse populations in areas with greater predator control. Foxes preferentially kill birds with higher parasite loads, and in the absence of predation, population-wide parasitism rates increase, with consequent negative impacts on grouse populations (Hudson et al. 1992, 1998, Packer et al. 2003).

iv) Past efforts to remove cattle from California grasslands in order to help populations of native plants often failed because grazing suppressed the now widespread European grasses that strongly outcompete most native plants. Now, conservation easements stipulate both maximum and minimum levels of grazing in the hopes of improving native populations (Germano et al. 2001, Hayes and Holl 2003).

*Given the nature of the peer-reviewed literature, which does not encourage the discussion, or even admission, of clearly unanticipated results, we attacked this problem by constructing an extremely simple questionnaire (Appendix A), which we then sent to 115 experienced field ecologists. 90\% replied affirmative to the existence of major surprise. Finally, a surprising (to us) number of the respondents wrote to say that because their observations were surprises, they had not been reported in the scientific write-ups of their research.

WHAT CAUSES ECOLOGICAL SURPRISES?

(1) Complex community interaction webs.

Although lip service is routinely paid to the complexity of ecological communities, in fact almost all our expectations of community behavior come from highly simplified, cartoon versions of the myriad interactions that characterize real communities.

(2) Variability in community players in time and space.

As with the complexity of species interactions, most ecologists appreciate that the numbers of individuals within populations, the traits of these individuals, and even the simple presence of different populations can vary dramatically across time and from place to place within otherwise similar communities.

(3) Multi-dimensionality of the characteristics and interactions of individual organisms.

The vast majority of formal ecological models reduce each species and individual to a simple set of characteristics and interaction rules (eg a birth rate, a death rate, and an interaction rate). Similarly, most models assume that biomass or energy is an adequate sole currency with which to characterize trophic interactions, ignoring the transfer of other major and minor elements and compounds. Ignored factors: non-trophic interactions (facilitation, mutualism, etc.), behavioral effects (e.g., the ecology of fear; Berger et al. 2001, Laundre et al. 2001), stoichiometric effects (Sterner and Elser 2002), and trait-mediated indirect effects (e.g., Billick and Case 1994, Schmitz 1998, Hansen et al. 2007).

(4) Shifting abiotic conditions can alter species reactions and interactions. Both shifts in mean conditions and rare weather events can alter populations and communities in ways that are extremely difficult to anticipate. Species can interact with one another in qualitatively or quantitatively different ways depending upon variation in the physical conditions that surround them (e.g., Sanford 1999).

*The frequency and nature of ecological surprises imply that uncertainty cannot be easily tamed through improved analytical procedures, and that prudent management of both exploited and conserved communities will require precautionary and adaptive management approaches.

Food chain and parasitic organisms

This is a rather recent topic, and how really significant it is not yet very clear. A recent review is:

James E. Byers Including parasites in food webs

Trends Parasitology  25 55 (2009)

*Lafferty et al. [(2008) Parasites in food webs: the ultimate missing links. Ecol. Lett. 11, 533] and a few previous related studies [6,7,10,11] collectively make a strong case that incorporating parasites into food webs should be standard procedure in the future. Their biggest contribution is to convincingly shift the burden onto food-web practitioners who choose to exclude parasites to justify their decision.

*Including parasites can preclude trophic cascades based strictly on relative body size [16], increase model ecosystem stability [17].

*To date, comparative work on empirical data indicates that parasites affect web dynamics minimally, although clearly, we have only scratched the surface of this issue.

The following paper gives a serious example:

Parasites are really significant in ecosystems (no surprise)

Kuris Ecosystem energetic implications of parasite and free-living biomass in three estuaries 

Nature  454 515 (2008)

*The biomass is estimated for free-living and parasitic species in three estuaries on the Pacific coast of California and Baja California. 

*Parasites have substantial biomass, which exceeded that of top predators. The biomass of trematodes was particularly high, being comparable to that of the abundant birds, fishes, burrowing shrimps and polychaetes.

[C] The surprise is simply due to prejudice.