1: \documentclass[12pt]{article}
2: \usepackage{graphicx}
3: \begin{document}
4:
5: \centerline {\bf \large Habitat width along a latitudinal gradient}
6:
7: \bigskip
8:
9: \centerline{D. Stauffer$^1$, C. Schulze$^1$, K. Rohde$^2$}
10:
11: \centerline{$^1$ Theoretische Physik, Universit"at, D-50923 K\"oln,
12: Euroland}
13:
14: \centerline{$^2$ Zoology, University of New England, Armidale NSW 2351, Australia.}
15:
16: \bigskip
17:
18: Abstract:
19:
20: \noindent
21: We use the Chowdhury ecosystem model, one of the most complex
22: agent-based ecological models, to test the latitude-niche breadth
23: hypothesis, with regard to habitat width, i.e., whether tropical species
24: generally have narrower
25: habitats than high latitude ones. Application of the model has given
26: realistic results in previous studies on latitudinal gradients in
27: species diversity and Rapoport's rule. Here we show that tropical
28: species with sufficient vagility and time to spread into adjacent
29: habitats, tend to have wider habitats than high latitude ones,
30: contradicting the latitude-niche breadth hypothesis.
31:
32:
33: Keywords:
34:
35: \noindent
36: Chowdhury ecosystem model, latitude-niche breadth hypothesis,
37: Rapoport's rule, latitudinal gradients in species diversity,
38: vagility, species-area relationship, fractal dimensions.
39:
40: \section{Introduction}
41:
42: According to a widely held view, an increase in diversity must result
43: in a narrowing of niches, in denser species packing. Thus, according
44: to Rosenzweig \& Ziv (1999) "Theory suggests that higher diversity
45: should shrink niches, allowing the coexistence of more species".
46: Applied to latitudinal gradients, the much greater species richness
47: in the tropics than in colder environments is thought possible only
48: because species are more densely packed, i.e., have smaller niches.
49: This view (the so called latitude-niche breadth hypothesis) can be
50: traced back to MacArthur (MacArthur 1965, 1969, 1972; MacArthur \&
51: Wilson 1967), but is probably even older. There is some empirical
52: evidence for this view (e.g., MacArthur 1965, 1969; Moore 1972), and
53: much against it (e.g., Rohde 1980; Novotny \& Basset 2005). For
54: example, concerning one aspect of the niche, the latitudinal range of
55: a species, some studies have provided support for the view that
56: latitudinal ranges are narrower at low latitudes (Rapoport's rule,
57: e.g. Stevens 1989), whereas others have found no support or evidence
58: for an opposite trend (e.g., Rohde et al. 1993). Rohde (1998)
59: therefore suggested two opposing trends: newly evolved species with
60: little vagility may have narrower ranges in the tropics, species with
61: greater vagility and of sufficient age to spread into adjacent areas
62: may have larger ranges. The same may apply to habitat width and
63: niche width in general.
64:
65: In this paper, we use the Chowdhury ecosystem model (Chowdhury \&
66: Stauffer, 2005; Stauffer et al. 2005) to examine the latitude-niche
67: breadth hypothesis with regard to one of the most important niche dimensions,
68: the habitat width. We check the effect of vagility and age of ecosystem on
69: habitat width. We have applied the model before (Rohde \& Stauffer 2005;
70: Stauffer \& Rohde 2006) to study the variation of species diversity
71: and latitudinal ranges with latitude, comparing cold with tropical
72: regions in simulations of the whole range of latitudes in a lattice
73: model, and got realistic results.
74:
75: \section{Methods}
76:
77: The Chowdhury model (Chowdhury \& Stauffer, 2005; Stauffer et al.
78: 2005) is one of the most complex agent-based (Billari et al. 2006)
79: ecological models (P\c ekalski, 2004; rimm \& Railsback, 2005) and has
80: been reviewed e.g. in Stauffer et al. (2006). Each species may move
81: to a neighbouring lattice site where it is still the same species.
82: Further details are given in an appendix.
83:
84: One open question is the fractal dimension $D$ of the number $N$ of
85: species found in a square of side length $L$:
86: $$ N \propto L^D. $$
87: Empirically, fractal dimensions $ 1.2 \le D \le 2.3$ are given
88: by Rosenzweig (1995),
89: whereas $D=2$ would correspond to a trivial proportionality of the number of
90: species and the area in which they are counted. The rationale behind our
91: comparison of fractal dimensions is: in the extreme case, the largest square
92: could have a single species, which is also found in the smallest
93: square, i.e., the slope is 0, the species' habitat is very wide. On the
94: other hand, the largest square could have 100 species, 10 of which
95: are also found in the smallest square, i.e. the slope is much
96: steeper, the habitats are much narrower.
97:
98: An earlier attempt (Stauffer and P\c ekalski (2005) roughly gave this simple
99: proportionality when it used the low vagility $d$ (diffusivity) which
100: gave good results in Rohde \& Stauffer (2005) and Stauffer \& Rohde
101: (2006). However, while in these papers we simulated the whole Earth
102: from the north pole to the south pole, tests of the above exponent $D$
103: should look at smaller, more homogenous regions. Thus, the
104: vagility $d$, which is the probability that a species invades a
105: neighbouring lattice site during one time step, has to be larger for
106: smaller lengths associated with the neighbor distance. Thus we now
107: use larger $d$ than in Stauffer \& P\c ekalski (2005), Rohde \&
108: Stauffer (2005) and Stauffer \& Rohde (2006) and also systematically
109: vary the observation time (measured in Monte Carlo steps per site; we
110: refrain from identifying it with years). We simulate (in most cases) ten $L
111: \times L$ square lattices, with the other parameters besides vagility and
112: observation time as in Rohde \& Stauffer (2005) and Stauffer \& Rohde
113: (2006). Each such simulation either refers to tropical or to high
114: latitude (here referred to as polar) regions. As in Rohde \& Stauffer
115: (2005) and Stauffer \& Rohde (2006), we use the standard Chowdhury
116: model for the simulation of the tropical region, while for the
117: polar region the birth rate is reduced by a factor 4. This birth rate
118: is the probability per iteration that offspring reaching maturity is
119: being produced, due to the harsh living conditions in polar regions we
120: assume this probability to be four times lower than in the tropics.
121:
122: \section{Results}
123:
124: \begin{figure}[hbt]
125: \begin{center}
126: \includegraphics[angle=-90,scale=0.25]{rohde1a.eps}
127: \includegraphics[angle=-90,scale=0.25]{rohde1b.eps}
128: \includegraphics[angle=-90,scale=0.25]{rohde1c.eps}
129: \includegraphics[angle=-90,scale=0.25]{rohde1d.eps}
130: \end{center}
131: \caption{Variation of the number $N$ of species with the length $L$ of
132: the square, $L = 4$, 8, 16 and 32. The vagility is $d=0.1$ for all four
133: cases. Upper lines = tropics, lower lines = polar.
134: }
135: \end{figure}
136:
137: Fig.1 sums up the species number $N$
138: over all lattice sites and over all time steps after
139: equilibration. The headlines give the observation times
140: varying from two thousand to two million time steps, for
141: various lattice sizes.
142:
143: We see that for short times the species barely had a chance to
144: move much from their site of origin, and thus $N$ is roughly
145: proportional to the area: $D = 2$. The longer the observation
146: time is, the more could the species spread over the lattice,
147: and the smaller is the slope of our log-log plots.
148: It appears that the slope is smaller
149: in the tropics, which means that habitats are not narrower but somewhat larger
150: there than in polar regions, if they had sufficient time to spread.
151:
152: Fig.2 shows for a fixed observation time of 200,000 that the
153: slope $D$ becomes the larger the smaller the vagility is, thus
154: explaining the results of Stauffer and P\c ekalski (2005).
155: For the smallest $d=0.001$ the data follow nearly
156: perfectly a line with slope $D = 2$, for the largest $d=0.1$ the curves starts
157: with $D = 1$. For small $d$ one no longer sees the difference in
158: the polar and tropical slopes which is seen for large $d$.
159:
160: All these slopes $D$ agree with reality (Rosenzweig 1995)
161: but do not come from good straight lines; our log-log plots
162: in general show upward curvature, and the slopes are those for intermediate
163: lattice sizes. Asymptotically for longer times and much larger lattice
164: sizes $L$ we
165: expect the trivial proportionality with $D=2$ since then the range $\ell$
166: over which a species is spread obeys $ 1 \ll \ell \ll L$. The real Earth,
167: however, may not correpond to these mathematical limits but to finite
168: sizes $L$ at finite times.
169:
170: Fig.3 shows that the results are not merely a function of the product of
171: vagility and observation time; varying $d$ influences many other properties
172: (Rohde and Stauffer 2005) and not only the time scale.
173:
174: \begin{figure}[hbt]
175: \begin{center}
176: \includegraphics[angle=-90,scale=0.31]{rohde2a.eps}
177: \includegraphics[angle=-90,scale=0.31]{rohde2b.eps}
178: \end{center}
179: \caption{Variation of $N$ versus $L$ for various $d$ at fixed observation time
180: of 0.2 million. Upper part: polar; lower part: x and + for tropics, stars and
181: squares for polar.
182: }
183: \end{figure}
184:
185: \begin{figure}[hbt]
186: \begin{center}
187: \includegraphics[angle=-90,scale=0.5]{rohde3.eps}
188: \end{center}
189: \caption{Variation of polar $N$ versus $L$ for two vagilities $d$ and two
190: observation times $t$ such that $dt$ is constant.
191: }
192: \end{figure}
193:
194: \section{Discussion}
195:
196: The findings presented in Fig. 1 contradict the latitude-niche width
197: hypothesis, for the niche dimension ``habitat width'',
198: according to which habitats are narrower in the tropics.
199: Indeed, they provide evidence for an opposite effect: habitats
200: are even larger near the equator than
201: at high latitudes. This agrees with the findings of Stauffer \& Rohde
202: (2006) who did not only fail to find support for Rapoport's rule, but
203: showed that latitudinal ranges are wider in the tropics, in agreement
204: with much empirical evidence.
205:
206: The findings presented in Fig. 2 show that habitats are smaller in
207: species with little vagility, in accordance with the hypothesis,
208: developed in the context of Rapoport's rule, that young species (or
209: subspecies) with little vagility, which have not had sufficient time
210: to spread into wider areas, have narrower latitudinal ranges at low latitudes
211: (Rohde 1998).
212:
213: Empirical evidence for the latitude-niche breadth hypothesis is
214: ambiguous. For example, Moore (1972) found that the average tropical
215: species occupies about half as much of the intertidal zone as the
216: average temperate species. According to MacArthur (1965, 1969),
217: tropical species often have a spottier distribution than
218: high-latitude ones. Concerning one aspect of the habitat of animals
219: and plants, i.e. their latitudinal ranges, Stevens (1989) provided
220: evidence that some plant and animal species have narrower latitudinal
221: ranges in the tropics, referring to this phenomenon as Rapoport's
222: rule. Some of the numerous subsequent studies also provided evidence
223: for the rule (review in Rohde 1999).
224:
225: However, support for the existence of narrower habitats in the tropics
226: is far from unequivocal. The studies that did not find support for
227: Rapoport's rule are more numerous than those that did, and in those
228: cases in which species have larger latitudinal ranges at high
229: latitudes, the increase is often restricted to high latitudes above
230: approximately 40-50 $^o$ N and S (review in Rohde 1999). Rohde (1996)
231: therefore suggested that the rule describes a local phenomenon, the
232: result of the extinction of species with narrow ranges during the ice
233: ages.
234:
235: 1) Several authors (e.g. Beaver 1979; review in Novotny \& Basset 2005)
236: have studied possible differences in host specificity of herbivorous
237: insects in tropical and temperate climates. No major differences were
238: found. 2) Detailed studies deal with latitudinal gradients in habitat
239: width of parasites of marine fish. Rohde (1978) has shown that host
240: ranges (the number of host species infected) of ectoparasitic
241: Monogenea infecting the gills are more or less the same at all
242: latitudes, whereas host ranges of another group of (endoparasitic)
243: flatworms, the Digenea, are markedly greater at high latitudes.
244: However, when correction was made for intensity and prevalence of
245: infection, host specificity was the same and very high at all
246: latitudes for both groups (Rohde 1980). Other niche dimensions of
247: these parasites, such as geographical range and microhabitat width,
248: were also examined and found not to be correlated with diversity,
249: although the data sets were small and more studies are needed. Host
250: size may on average be smaller in the tropics, due to the very large
251: number of host species, many of them small (Rohde 1989). 3)
252: Lappalainen and Soininen (2006) analysed the determinants of fish
253: distribution and the variability in species' habitat breadth and
254: position along latitudinal gradient of boreal lakes and found that
255: the regional occupancy of species was more strongly governed by the
256: habitat position than the habitat breadth. The cool water species
257: (percids and cyprinids) showed significant decrease in habitat breadth
258: towards higher latitudes (and not towards lower latitudes, expected
259: by the latitude-niche breadth hypothesis). 4) Some further examples
260: are discussed in V\'azquez \& Stevens (2004).
261:
262: V\'azquez \& Stevens (2004) have reviewed the evidence for the
263: latitude-niche breadth hypothesis, using meta-analytical techniques.
264: They found that the results of the meta-analysis do not permit
265: rejection of the null hypothesis of there being no correlation
266: between latitude and niche breadth. They also critically examined the
267: two assumptions on which MacArthur's hypothesis are based, i.e., 1)
268: that there is a latitudinal gradient in population variability, and
269: 2) that there is a relationship between population variability and
270: niche breadth. These assumptions are widely accepted (e.g., May
271: 1973). They claim that the tropics have greater stability and less
272: seasonality than temperate regions, making populations more stable,
273: thus allowing narrower niches. However, Rohde (1992) has pointed out
274: that there may be extreme variations in temperature, salinity and
275: currents in tropical shallow waters, such as high diversity coral
276: reefs. Such variations may occur over short time spans of a few
277: hours. The meta-analysis of V\'azquez \& Stevens (2004) shows that
278: available evidence does not support the view of an increasing
279: population variability with latitude, and evidence for narrower
280: niches of less variable populations is at best equivocal and does not
281: permit rejection of the null hypothesis of no relationship.
282:
283: In spite of these criticisms of the mechanism involved, there could
284: be a latitudinal gradient in niche width due to other mechanisms.
285: V\'azquez \& Stevens (2004) suggest such a mechanism. Greater
286: specialization may be a by-product of the latitudinal gradient in
287: species diversity, because nestedness leads to an asymmetric, i.e.
288: faster increase of specialized species than of communities. In other
289: words, nestedness and asymmetric spezialisation tend to increase with
290: the number of species in a network. V\'azquez \& Stevens (2004) pay
291: particular attention to parasites. Nestedness of interactions between
292: species has, for example, been observed in marine Monogenea (Morand
293: et al. 2002), for which group, however, host specificity does not
294: change with latitude. Overall, nestedness is not common among
295: parasites of fish (Rohde et al. 1998; Poulin \& Valtonen 2001).
296:
297: Finally and importantly, the latitude-niche breadth hypothesis as
298: formulated by MacArthur and his followers makes equilibrium
299: assumptions, and it implicitly and explicitly assumes that habitat
300: space is more or less filled with species. However, there is much
301: evidence that there is an overabundance of vacant habitats and that
302: most ecological systems are far from saturation (for a discussion and
303: examples see Rohde 2005). This removes the very basis on which the
304: hypothesis rests. The Chowdhury model does not make equilibrium
305: assumptions and incorporates vacant niches. Our simulations using
306: this model are further evidence against the latitude-niche breadth
307: hypothesis: tropical vagile species that have had sufficient time to
308: spread away from their original habitat, do not have narrower but
309: wider habitats than high latitude species.
310:
311: How can we reconcile our results, that habitats of species are
312: somewhat larger in the tropics than at higher latitudes, with the
313: well known latitudinal gradient in species diversity? One possible
314: explanation is the idea of Terborgh (1973) and Rosenzweig (1995),
315: that tropical zones are generally larger and therefore stimulate
316: speciation and inhibit extinction. That larger areas (all other
317: conditions being equal or at least similar) often accommodate more
318: species, is well established. For example, at the level of
319: geographical area, Blackburn and Gaston (1997) found that there is
320: indeed a relationship between the land area and species richness of a
321: region once tropical species are excluded. This relationship is
322: independent of the latitude and productivity of regions. A study on
323: South American mammals (Ruggiero 1999) confirmed this: The number of
324: principally extra-tropical mammal species per unit area depends on
325: the biome area (for further examples see Rosenzweig 1995). However,
326: as pointed out by Rohde (1998), although area matters, it cannot be
327: the primary cause of the latitudinal diversity gradients: many high
328: diversity tropical areas are much smaller than low diversity areas at
329: high latitudes. Many recent studies have provided support for this
330: view (e.g., MacPherson 2002: area size does not explain the
331: latitudinal pattern in benthic species richness on a large spatial
332: scale. Willig \& Bloch 2006: "area does not drive the latitudinal
333: gradient of bat species richness in the New World. In fact, area
334: represents a source of noise rather than a dominant signal at the
335: focal scale of biome types and provinces in the Western Hemisphere").
336: - Our results, that the habitats occupied by species are somewhat
337: larger in the tropics than at higher latitudes, mean, with regard to
338: latitudinal gradients, that there must be much overlap between
339: habitats, leading to a far greater diversity in tropical than in high
340: latitude areas of the same size. The larger (compared with high
341: latitude) tropical areas (in Africa and the IndoPacific) would
342: aggravate this. The overlap postulated here resembles the "Rapoport
343: rescue effect" of Stevens (1989), according to which tropical species
344: frequently "spill out" from their preferred habitat into adjacent
345: less favourable ones, thus explaining the high diversity there.
346: However, it is not necessary to distinguish favourable and less
347: favourable habitats: species may simply "spill out" from the habitat where
348: they have originated, into adjacent habitats that are as suitable.
349:
350: \section{Appendix: Model details}
351:
352: The Chowdhury model of ecosystem has been described and modified in many
353: publications since 2003, and we give here only a short outline. Individuals
354: are born, mature, produce offspring asexually, and die with a probability
355: increasing exponentially with age after maturity. At most 100 animals fit
356: into one habitat. Six trophic levels define pre-predator relations: The upper
357: levels feed on the adjacent lower ones. The topmost level has one habitat, the
358: second two, then 4, 8, .... At each iteration, with one percent probability
359: the food habits, minimum age of reproduction, and number of births per
360: iteration mutate randomly, allowing self-organisation of these parameters
361: through selection of the fittest. Death may come from from being eaten by a
362: predator, from starvation, or from old age (with a high lifespan on the top
363: food levels and a low lifespan on the bottom levels). If a species becomes
364: extinct, then with probability 0.0001 per iteration the empty habitat is filled
365: by another species. Each of the $L^2$ lattice sites carries such an ecosystem,
366: each with dozens of living species. At the beginning, each different species
367: gets a different number as its name. The number of differents species first
368: decays with time (=iterations) and then fluctuates about some low average value.
369:
370: Then with probability $d$ at each iterations
371: a species can migrate into a randomly selected
372: neighbour site, if the corresponding habitat on that neighbour site is empty at
373: that time. A random fraction of the population moves, the rest stays at the
374: old site. Both parts of the population carry the same name, and in this way
375: are counted as only once species spreading over more than one site. Summing
376: up over all different surviving names we obtain the number of different
377: species at that moment.
378: \bigskip
379:
380: {\bf References}
381:
382: \parindent 0pt
383: \medskip
384: Beaver, R.A. (1979). Host specificity of temperate and tropical
385: animals. Nature 281, 1139-141.
386:
387: \medskip
388: Billari, F.C., Fent, T., Prskawetz, A. \& Scheffran, J. (Eds.) (2006).
389: Agent-based computational modelling, Heidelberg: Physica-Verlag .
390:
391: \medskip
392: Blackburn, T.M. \& Gaston, K.J. (1997). The relationship between
393: geographic area and the latitudinal gradient in species richness in
394: New World birds. Evolutionary ecology 11, 195-204.
395:
396: \medskip
397: Chowdhury, D. and Stauffer, D. (2005).
398: Evolutionary ecology in silico: Does physics help in understanding
399: the ``generic'' trends ? J. Biosci. (India) 30, 277-287.
400:
401: \medskip
402: Grimm, V. \& Railsback, S.F. (2005). Individual-based modeling and
403: ecology. Princeton University Press, Princeton.
404:
405: %\medskip
406: %Hutchinson, G.E. (1957). Concluding remarks. Cold Spring Harbour on
407: %Quantitative Biology 22, 415-427.
408:
409: \medskip
410: Lappalainen, J. \& Soininen, J. (2006). Latitudinal gradients in niche
411: breadth and position - regional patterns in freshwater fish.
412: Naturwissenschaften 93, 246-250.
413:
414: \medskip
415: MacArthur, R.H. (1965). Patterns of species diversity. Biological
416: Reviews 40, 510-533.
417:
418: \medskip
419: MacArthur, R.H. (1969). Patterns of communities in the tropics.
420: Biological Journal of the Linnaean Society 1, 19-30.
421:
422: \medskip
423: MacArthur, R.H. (1972). Geographical Ecology. Princeton University
424: Press, Princeton, NJ.
425:
426: \medskip
427: MacArthur, R.H. \& Wilson E.O. (1967). An equilibrium theory of
428: insular zoogeography. Evolution 17, 373-387.
429:
430: \medskip
431: MacPherson, E. (2002). Large-scale species-richness gradients in the
432: Atlantic Ocean. Proceedings of the Royal Society of London,
433: B-Biological Sciences 269, 1715-1720.
434:
435: \medskip
436: May, R.M. (1973). Stability and complexity in model ecosystems.
437: Princeton University Press, Princeton N.J.
438:
439: \medskip
440: Moore, H. B. (1972). Aspects of stress in the tropical marine
441: environment. Advances in marine Biology 10, 217-269.
442:
443: \medskip
444: Morand, S., Rohde, K. \& Hayward, C.J. (2002). Order in parasite
445: communities of marine fish is explained by epidemiological processes.
446: Parasitology, 124, S57-S63.
447:
448: \medskip
449: Novotny, V. \& Basset, Y. (2005). Host specificity of insect
450: herbivores in tropical forests. Proceedings of the Royal Society B -
451: Biological Sciences 272, 1083-1090.
452:
453: \medskip
454: P\c ekalski, A. (2004). A short guide to predator and prey lattice models by
455: physicists. Computing in Science and Engineering 6, 62-66 (Jan/Feb 2004).
456:
457:
458: \medskip
459: Poulin, R. \& Valtonen, E. T. (2001). Nested assemblages resulting
460: from host size variation: the case of endoparasite communities in
461: fish hosts. International Journal for Parasitology 31, 1194-1204.
462:
463: \medskip
464: Rohde, K. (1978). Latitudinal differences in host-specificity of
465: marine Monogenea and Digenea. Marine Biology 47, 125-134.
466:
467: \medskip
468: Rohde, K. (1980). Host specificity indices of parasites and their
469: application. Experientia 36, 1370-1371.
470:
471: \medskip
472: Rohde, K. (1989). Simple ecological systems, simple solutions to
473: complex problems? Evolutionary Theory 8, 305-350.
474:
475: \medskip
476: Rohde, K. (1992). Latitudinal gradients in species diversity: the
477: search for the primary cause. Oikos 65, 514-527.
478:
479: \medskip
480: Rohde, K. (1996). Rapoport's Rule is a local phenomenon and cannot
481: explain latitudinal gradients in species diversity. Biodiversity
482: Letters 3, 10-13.
483:
484: \medskip
485: Rohde, K. (1998). Latitudinal gradients in species diversity; area
486: matters, but how much? Oikos 82, 184-190.
487:
488: \medskip
489: Rohde, K. (1999). Latitudinal gradients in species diversity and
490: Rapoport's rule revisited: a review of recent work, and what can
491: parasites teach us about the causes of the gradients? Ecography 22,
492: 593-613 (invited Minireview on the occasion of the 50th anniversary
493: of the Nordic Ecological Society Oikos). Also published in Fenchel,
494: T. ed.: Ecology 1999-and tomorrow, pp. 73-93. Oikos Editorial Office,
495: University Lund, Sweden.
496:
497: \medskip
498: Rohde, K. (2005). Nonequilibrium ecology. Cambridge University Press,
499: Cambridge.
500:
501: \medskip
502: Rohde, K., Heap, M. \& Heap, D. (1993). Rapoport's rule does not
503: apply to marine teleosts and cannot explain latitudinal gradients in
504: species richness. American Naturalist, 142, 1-16.
505:
506: \medskip
507: Rohde, K., Worthen, W., Heap, M., Hugueny, B. \& Guégan, J.-F. (1998).
508: Nestedness in assemblages of metazoan ecto- and endoparasites of
509: marine fish. International Journal for Parasitology, 28, 543-549.
510:
511: \medskip
512: Rohde, K. \& Stauffer, D. (2005). Simulations of geographical trends
513: in the Chowdhury ecosystem model. Advances in Complex Systems 8,
514: 451-464.
515:
516: \medskip
517: Rosenzweig, M. L. (1995). Species diversity in space and time.
518: Cambridge University Press, Cambridge.
519:
520: \medskip
521: Rosenzweig, M.L. \& Ziv, Y. (1999). The echo pattern of species
522: diversity: patterns and processes. Ecography 22, 614-629.
523:
524: \medskip
525: Ruggiero, A. (1999). Spatial patterns in the diversity of mammal
526: species: A test of the geographic area hypothesis in South America.
527: EcoScience 6, 338-354.
528:
529: \medskip
530: Stauffer, D \& P\c ekalski, A. (2005) Species numbers versus area in Chowdhury
531: Ecosystems. www.arXiv.org: qbio.PE/0510021.
532:
533: \medskip
534: Stauffer, D. \& Rohde, K. (2006). Simulation of Rapoport's rule for
535: latitudinal species spread. Theory in Biosciences 125, 55-65.
536:
537: \medskip
538: Stauffer, D., Kunwar, A. \& D. Chowdhury,D. (2005).
539: Evolutionary ecology in-silico: evolving foodwebs, migrating population
540: and speciation. Physica A 352, 202-215 (2005).
541:
542: \medskip
543: Stauffer, D., Moss de Oliveira, S., de Oliveira, P.M.C., \& S\'a
544: Martins, J.S. (2006). Biology, Sociology, Geology by Computational
545: Physicists. Amsterdam, Elsevier.
546:
547: \medskip
548: Stevens, G.C. (1989). The latitudinal gradients in geographical
549: range: how so many species co-exist in the tropics. American
550: Naturalist 133, 240-256.
551:
552: \medskip
553: Terborgh, J. (1973). On the notion of favourableness in plant
554: ecology. American Naturalist 107, 481-501.
555:
556: \medskip
557: V\'azquez, D.P. \& Stevens, R.D. (2004). The latitudinal gradient in
558: niche breadth: concepts and evidence. American Naturalist 164,
559: E1-E19.
560:
561: \medskip
562: Willig, M.R. \& Bloch, C.P. (2006). Latitudinal gradients of species
563: richness: a test of the geographic area hypothesis at two ecological
564: scales. Oikos 112, 163-173.
565:
566:
567: \end{document}
568: