WEBVTT 00:01.170 --> 00:03.480 Prof: You guys had a little bit of fun getting 00:03.479 --> 00:06.099 dressed up and doing some stuff up on stage last Friday, 00:06.100 --> 00:09.870 and so I though that I would actually have a little fun 00:09.871 --> 00:10.501 myself. 00:10.500 --> 00:12.640 And I was unable, as you know, 00:12.644 --> 00:16.864 to decide which section had actually done the best job. 00:16.860 --> 00:20.460 So I decided that you would all get a reward. 00:20.460 --> 00:22.590 And in order to understand the reward, 00:22.590 --> 00:25.770 I want to give you the background, which actually comes 00:25.767 --> 00:29.197 from a very, of course, deep theoretical 00:29.204 --> 00:35.264 investigation of the kind that I always want you to know about. 00:35.260 --> 00:37.530 Right? 00:37.530 --> 00:42.330 So let me get this up there. 00:42.330 --> 00:43.780 Here we go. 00:43.780 --> 00:48.040 I want you to understand the Swiss theorem. 00:48.040 --> 00:48.720 Okay? 00:48.720 --> 00:54.570 And the Swiss theorem is really a very essential part of 00:54.574 --> 00:56.814 population biology. 00:56.810 --> 01:00.670 Basically it tells you that the partial derivative of happiness, 01:00.670 --> 01:02.570 with respect to an increase in chocolate intake, 01:02.570 --> 01:05.580 is positive across the entire reaction norm of the human 01:05.581 --> 01:06.241 condition. 01:06.239 --> 01:06.999 Okay? 01:07.000 --> 01:09.750 So chocolate makes you happier, right? 01:09.750 --> 01:11.210 And the Swiss, as is usual, 01:11.210 --> 01:14.530 claim priority for something which is in fact a contribution 01:14.525 --> 01:17.725 of another major civilization, which is the Mexicans. 01:17.730 --> 01:22.140 Now, you will notice that I have had the teaching fellows 01:22.135 --> 01:26.615 divvy up the chocolate onto paper plates and distribute it 01:26.620 --> 01:29.690 in small qualities around the room. 01:29.688 --> 01:33.548 This is because the homogenization of fitness within 01:33.554 --> 01:37.954 groups promotes group benefit and causes a general increase 01:37.947 --> 01:39.537 in-group success. 01:39.540 --> 01:43.580 Because the only way then that the group can improve is that if 01:43.578 --> 01:46.638 the performance of the individuals increases the 01:46.640 --> 01:48.920 performance of the whole group. 01:48.920 --> 01:52.990 Now, that's about all that I really want to say, 01:52.992 --> 01:58.282 except that I want to thank you very much for participating in 01:58.280 --> 02:00.880 that exercise last Friday. 02:00.879 --> 02:02.129 I thought it was interesting. 02:02.129 --> 02:02.889 I enjoyed it. 02:02.890 --> 02:04.700 I hope you've had a chance to look at it. 02:04.700 --> 02:08.000 And the only issue that now faces me is whether I should 02:08.000 --> 02:10.820 continue to give the lecture in this costume. 02:10.818 --> 02:14.218 I have been told by my teaching fellows that I absolutely have 02:14.224 --> 02:14.844 to do it. 02:14.840 --> 02:15.770 Right? 02:15.770 --> 02:19.930 But I have difficulty believing that you'll believe me if I look 02:19.926 --> 02:20.716 like this. 02:20.720 --> 02:22.980 Will you believe me? 02:22.979 --> 02:23.959 Students: Yes. 02:23.960 --> 02:25.350 Prof: You want me to leave it on? 02:25.349 --> 02:25.839 Students: Yes. 02:25.840 --> 02:27.890 Prof: All right, all right. 02:27.889 --> 02:28.399 Let's go. 02:28.400 --> 02:29.400 Okay? 02:29.400 --> 02:32.810 So we're actually now going to go back to-- 02:32.810 --> 02:34.750 oh by the way, there is a pedagogical message 02:34.751 --> 02:37.321 in this costume, and that is if you can't play 02:37.323 --> 02:39.533 like a child, you can't be creative. 02:39.530 --> 02:41.120 Okay? 02:41.120 --> 02:43.370 It's about playfulness. 02:43.370 --> 02:46.030 Okay. 02:46.030 --> 02:48.740 Have you all got your chocolate? 02:48.740 --> 02:50.770 Are you happy? 02:50.770 --> 02:51.560 Students: Yes. 02:51.560 --> 02:54.720 Prof: Or let me put it this way, are you happier? 02:54.720 --> 02:58.350 > 02:58.348 --> 03:02.858 It looks like we have some leftovers, and so you're all 03:02.862 --> 03:06.042 welcome to grab some on the way out. 03:06.038 --> 03:09.028 The basket, by the way, is a working basket from a 03:09.032 --> 03:12.512 market in the Ivory Coast, and when people make baskets to 03:12.514 --> 03:15.084 work with, they make them really well. 03:15.080 --> 03:17.980 So I don't want the basket to disappear. 03:17.979 --> 03:20.559 Okay? 03:20.560 --> 03:23.760 Now for today's lecture. 03:23.758 --> 03:27.858 What we're looking at now is the impact of space on 03:27.857 --> 03:31.787 communities and on the distribution of plants and 03:31.790 --> 03:34.250 animals around the planet. 03:34.250 --> 03:38.510 So I'm going to talk today about island biogeography and 03:38.506 --> 03:39.896 metapopulations. 03:39.900 --> 03:43.650 And I want you to start thinking of the world as 03:43.645 --> 03:45.155 fragmented, okay? 03:45.160 --> 03:47.660 So, as spatially heterogeneous. 03:47.660 --> 03:49.760 There are islands, there are mountains, 03:49.760 --> 03:51.310 there are lakes and oases. 03:51.310 --> 03:55.340 And a lot of the world is becoming fragmented by humans. 03:55.340 --> 03:58.570 So one of the big ways that humans are having impact on the 03:58.567 --> 04:01.957 planet, and changing life on the planet, is by fragmenting the 04:01.961 --> 04:02.741 landscape. 04:02.740 --> 04:07.340 So the issue today is what is determining biodiversity in a 04:07.336 --> 04:08.206 fragment? 04:08.210 --> 04:12.280 What happens when we break the world up? 04:12.280 --> 04:19.700 Here are some sketches of what happened in Cadiz Township, 04:19.696 --> 04:24.246 Wisconsin, between 1831 and 1950. 04:24.250 --> 04:27.750 So before the settlers go out, we had an Eastern Hardwood 04:27.745 --> 04:30.065 Forest, and as the area got 04:30.065 --> 04:33.835 settled--this is 1882, 1902,1950--you can see the 04:33.843 --> 04:37.743 forest disappears and there are little blocks of tress that are 04:37.738 --> 04:41.228 left as forest fragments, scattered across the landscape. 04:41.230 --> 04:45.570 And this is where the birds and the rodents and the coyotes and 04:45.574 --> 04:49.714 everything else now--and the deer--have to try to make their 04:49.709 --> 04:50.479 living. 04:50.480 --> 04:54.220 And if we now look at Wisconsin from space-- 04:54.220 --> 04:57.120 so this is in the early part of this century-- 04:57.120 --> 05:03.000 you can see a picture which actually has cities in it, 05:03.000 --> 05:06.540 in red, and a landscape, which is largely agricultural, 05:06.540 --> 05:09.780 with some little dark blocks of forest scattered through it. 05:09.778 --> 05:11.668 So this is mostly agriculture here; 05:11.670 --> 05:13.900 this would be forest. 05:13.899 --> 05:15.659 Let's take another look. 05:15.660 --> 05:18.130 Another kind of natural fragmentation, 05:18.130 --> 05:20.670 of course, is archipelagos, islands. 05:20.670 --> 05:23.680 Now we're getting towards island biogeography. 05:23.680 --> 05:28.570 This is an observation made by Robert MacArthur and Ed Wilson 05:28.574 --> 05:31.844 that plots on the Y-axis, on a log scale, 05:31.838 --> 05:35.018 the number of species on an island; 05:35.019 --> 05:39.989 and on the X-axis the area of the island, in square miles. 05:39.990 --> 05:42.790 And this is for the Sunda Islands; 05:42.790 --> 05:44.230 so this is in Indonesia. 05:44.230 --> 05:47.700 And up here you've got the Philippines and New Guinea, 05:47.699 --> 05:49.599 added to the Sunda Islands. 05:49.600 --> 05:51.920 So what you see is that on a log-log plot, 05:51.916 --> 05:54.456 the bigger the island the more the species. 05:54.459 --> 05:57.459 And this is the part of the world we're talking about. 05:57.459 --> 06:00.349 Komodo dragons come from Komodo here. 06:00.350 --> 06:03.670 The little dwarf Homo erectus was found on Flores. 06:03.670 --> 06:04.610 This is Timor. 06:04.610 --> 06:06.590 New Guinea is going to be off here somewhere. 06:06.588 --> 06:11.018 Java is over there, and Krakatoa is down here. 06:11.019 --> 06:14.179 So these are the Sunda Islands. 06:14.180 --> 06:16.400 And they have lots of different surface areas. 06:16.399 --> 06:19.809 I mean, if you just look at this little one and compare it 06:19.812 --> 06:22.332 to this big one, you're going to have a lot 06:22.326 --> 06:25.136 fewer birds on there than you are on there. 06:25.139 --> 06:30.209 So in trying to come up with a general theory of biogeography, 06:30.209 --> 06:34.389 MacArthur and Wilson did the standard Cartesian analytical 06:34.391 --> 06:37.281 reduction of saying, "What are the essential 06:37.283 --> 06:38.363 features of that system? 06:38.360 --> 06:42.220 What are the fewest things that we have to pay attention to, 06:42.220 --> 06:46.680 that will tell us something, a take-home message, 06:46.680 --> 06:48.690 some key message, that we can pull out of the 06:48.685 --> 06:49.365 system?" 06:49.370 --> 06:53.090 And they thought, well let's suppose that there 06:53.091 --> 06:57.121 isn't any evolution going on, and that all of the species 06:57.120 --> 07:00.000 we're considering already exist on some big continental 07:00.004 --> 07:02.474 landmass, and that they're getting onto 07:02.473 --> 07:04.803 islands in a process of immigration, 07:04.800 --> 07:07.730 and they're dying out on islands in a process of 07:07.728 --> 07:08.538 extinction. 07:08.540 --> 07:11.930 And this is the number of species that are present on the 07:11.925 --> 07:12.465 island. 07:12.470 --> 07:13.540 Okay? 07:13.540 --> 07:17.640 So they argued that in such a situation the number of species 07:17.637 --> 07:20.707 on the island will come to an equilibrium-- 07:20.709 --> 07:23.179 there'll be species that are coming off of the continents, 07:23.180 --> 07:25.310 that are flying, drifting, getting blown, 07:25.310 --> 07:28.490 whatever, out onto the islands--and the immigration 07:28.494 --> 07:29.964 rate will start high. 07:29.959 --> 07:33.519 When the island is empty, everything that shows up on it 07:33.519 --> 07:34.749 is a new species. 07:34.750 --> 07:38.430 But the immigration rate must inevitably fall down to zero 07:38.430 --> 07:42.310 when the number of species on the island equals the number in 07:42.305 --> 07:44.625 the source pool on the mainland. 07:44.629 --> 07:45.899 Okay? 07:45.899 --> 07:47.939 Because we're just counting species; 07:47.940 --> 07:50.930 we're not counting number of individuals arriving. 07:50.930 --> 07:54.490 I mean, if a hundred birds fly in and their species already 07:54.494 --> 07:56.984 exists on the island, that doesn't count for 07:56.978 --> 07:59.338 immigration, because that species is already 07:59.341 --> 07:59.731 there. 07:59.730 --> 08:02.520 So that's the way this axis is constructed. 08:02.519 --> 08:07.279 So this goes down and this goes up. 08:07.278 --> 08:10.648 The number of species on the island is going to be affecting 08:10.648 --> 08:13.158 the extinction rate, probably in two ways. 08:13.160 --> 08:15.200 The simplest is the more species there are, 08:15.201 --> 08:17.881 the greater the chance that one of them will go extinct, 08:17.875 --> 08:18.795 just at random. 08:18.800 --> 08:21.560 So that curve's going to go up, just because there are more 08:21.564 --> 08:22.714 species on the island. 08:22.709 --> 08:26.379 However, if there are interactions among the species 08:26.384 --> 08:28.824 on the island, such that predation, 08:28.817 --> 08:32.197 disease, whatever is going to drive one to extinction, 08:32.200 --> 08:34.760 you can see that it might bend upward. 08:34.759 --> 08:37.109 So it's not just going to be linear, it's going to go up; 08:37.110 --> 08:38.810 so that's how they argued the curves. 08:38.808 --> 08:40.668 And they said where the curves intersect, 08:40.668 --> 08:42.868 the number coming in will equal the number going out, 08:42.870 --> 08:45.640 and that's the number we should expect to find on the island. 08:45.639 --> 08:49.389 So far so good; this is all just a priori. 08:49.389 --> 08:54.489 Well what's going to affect the rates of immigration and 08:54.493 --> 08:55.703 extinction? 08:55.700 --> 08:59.340 Well first they argued that immigration will decrease from 08:59.339 --> 09:02.979 islands that are near the mainland out to islands that are 09:02.979 --> 09:04.639 far from the mainland. 09:04.639 --> 09:07.189 So you could construct a series of curves; 09:07.190 --> 09:09.930 so this would be the curve for an island close to the mainland 09:09.932 --> 09:12.902 and this would be the curve for an island far from the mainland. 09:12.899 --> 09:15.349 And that's just because it's harder to get out there. 09:15.350 --> 09:18.400 On the other hand, they argued that extinction 09:18.404 --> 09:22.554 rates will increase as you go from large islands down to small 09:22.547 --> 09:24.817 islands, basically because there's more 09:24.817 --> 09:27.137 space on the big island, more different niches, 09:27.144 --> 09:29.334 more places, habitats, where organisms can 09:29.333 --> 09:30.893 live; more different kinds of things 09:30.886 --> 09:31.756 could survive there. 09:31.759 --> 09:34.069 But also as you get onto a small island, 09:34.070 --> 09:36.750 the intensity of the biotic interactions are going to get 09:36.750 --> 09:38.510 bigger, and it's going to be harder to 09:38.514 --> 09:41.704 get away from a predator, or harder to get away from a 09:41.702 --> 09:42.432 parasite. 09:42.428 --> 09:45.968 And so you could imagine that as you shrank an island, 09:45.969 --> 09:48.039 extinction rates would go up. 09:48.038 --> 09:51.268 So they predicted, hey, we could have an 09:51.270 --> 09:55.910 equilibrium over here on an island which is small and far 09:55.912 --> 09:59.292 from the coast, or that we could have it here 09:59.291 --> 10:02.201 if it was a small island close to the continent. 10:02.200 --> 10:06.000 And similarly if we have a large island which is far away 10:06.004 --> 10:08.574 from the continent, we might be down here, 10:08.572 --> 10:11.352 and a large island close to the continent would be here. 10:11.350 --> 10:13.900 A big one close to a continent, by the way, would be Trinidad. 10:13.899 --> 10:16.919 Trinidad has almost the biodiversity of neighboring 10:16.923 --> 10:17.653 Venezuela. 10:17.649 --> 10:21.069 A small one, far away from a continent-- 10:21.070 --> 10:24.190 so with a very low biodiversity--might be something 10:24.190 --> 10:25.930 like, you would think, 10:25.927 --> 10:30.337 an isolated oceanic island like Easter Island or Hawaii. 10:30.340 --> 10:33.480 So why is this important? 10:33.480 --> 10:38.620 Well for a long time, from the mid-1960s up to about 10:38.623 --> 10:44.083 1985, 1990, this was the only game in 10:44.080 --> 10:47.300 town, and there weren't alternative 10:47.303 --> 10:50.383 ways of thinking about these processes, 10:50.379 --> 10:53.389 and it played a big role in the design of natural parks and 10:53.394 --> 10:54.334 nature reserves. 10:54.330 --> 10:58.440 Essentially it said, because of the area, 10:58.440 --> 11:00.650 the effect of area on biodiversity, 11:00.649 --> 11:04.619 it's better to have a big park than a small one, 11:04.620 --> 11:07.430 and because of the effect of immigration rate on 11:07.432 --> 11:10.222 biodiversity, it said it's better to provide 11:10.217 --> 11:12.837 corridors to connect landscape fragments, 11:12.840 --> 11:16.160 if you possibly can, so that things can immigrate 11:16.160 --> 11:17.960 and move back and forth. 11:17.960 --> 11:22.580 So it was used a lot. 11:22.580 --> 11:27.290 However, when we summarize it, you'll see it's an equilibrium 11:27.293 --> 11:30.283 between colonization and extinction; 11:30.278 --> 11:32.948 it assumes there's a source population. 11:32.950 --> 11:36.490 So evolution isn't going on; there's a source population out 11:36.488 --> 11:38.958 there, that's where all the species are. 11:38.960 --> 11:41.240 There's no speciation occurring on islands. 11:41.240 --> 11:44.120 There are just two effects that you're worried about: 11:44.120 --> 11:47.000 how big the island is and how far away it is from the 11:47.000 --> 11:47.720 mainland. 11:47.720 --> 11:52.030 Extinction is driven by area, and colonization is driven by 11:52.032 --> 11:52.852 distance. 11:52.850 --> 11:54.140 Okay? 11:54.139 --> 11:58.229 Now all of these things that I've just written down here will 11:58.230 --> 12:01.980 be important for you to remember if you are asked how to 12:01.980 --> 12:05.870 reconstruct the equilibrium number of species on an island 12:05.868 --> 12:08.048 in a certain circumstance. 12:08.049 --> 12:09.639 Okay? 12:09.639 --> 12:11.449 I want to emphasize that. 12:11.450 --> 12:14.820 That's the kind of thing that might turn up on a midterm. 12:14.820 --> 12:18.510 12:18.509 --> 12:21.619 If you're on a small island, a long way from the mainland, 12:21.620 --> 12:23.780 you'll have low species diversity, and a big island 12:23.783 --> 12:26.253 close to the mainland will have high species diversity. 12:26.250 --> 12:28.670 And that seems to be an intuitive point, 12:28.668 --> 12:31.888 but I've given you an analytical framework from which 12:31.894 --> 12:32.954 to derive it. 12:32.950 --> 12:39.680 Now, as you'll see in a minute, I am now going to blast this 12:39.682 --> 12:42.652 theory out of the water. 12:42.649 --> 12:46.169 I'm going to take it apart and show that it makes a whole 12:46.174 --> 12:50.334 series of assumptions and claims that are demonstrably not true. 12:50.330 --> 12:53.350 And before I do that, I want to signal to you that 12:53.347 --> 12:55.747 I'm going to come back, after I do that, 12:55.745 --> 12:59.085 and say, "Hey, it was still a good thing." 12:59.090 --> 13:01.460 Okay? 13:01.460 --> 13:05.090 So the theory is dealing only with the number of species, 13:05.090 --> 13:07.490 not with the number of individuals; 13:07.490 --> 13:09.110 there's no population dynamics. 13:09.110 --> 13:11.120 I mean, if you've got ten of them on the island, 13:11.120 --> 13:13.650 or a thousand of them on the island, you count them the same 13:13.647 --> 13:13.987 way. 13:13.990 --> 13:17.650 That seems to be a little silly, because the probably of 13:17.653 --> 13:21.653 extinction should be related to the number that are there. 13:21.649 --> 13:25.099 All the species are considered together, and there's just kind 13:25.097 --> 13:28.827 of one general immigration rate and one general extinction rate. 13:28.830 --> 13:31.350 You know, it's kind of fun to wave my arms when I've got this 13:31.350 --> 13:31.770 gown on. 13:31.769 --> 13:33.109 > 13:33.110 --> 13:36.750 But the probabilities of immigration and extinction are 13:36.750 --> 13:38.800 different; I mean, it's going to be 13:38.799 --> 13:41.689 different for birds and ants and mosses and paramecia and 13:41.693 --> 13:43.243 elephants and stuff like that. 13:43.244 --> 13:43.714 Okay? 13:43.710 --> 13:47.110 So they must some ways differ systematically. 13:47.110 --> 13:50.300 The island biogeographic theory, kind of like the 13:50.297 --> 13:53.547 Hardy-Weinberg theory, it's an equilibrium theory. 13:53.552 --> 13:54.152 Okay? 13:54.149 --> 13:55.789 It doesn't allow for history. 13:55.788 --> 13:59.018 But if we just look at what's happened in the Western Pacific 13:59.020 --> 14:01.930 in the last 10,000 years, as the Polynesians came out and 14:01.927 --> 14:04.627 colonized those islands, we know that 25% of the birds 14:04.628 --> 14:06.998 that were on them have already gone extinct. 14:07.000 --> 14:10.350 So you send your budding biogeographer out there to get a 14:10.351 --> 14:13.521 sample from the Sunda Islands or from Guam or from the 14:13.524 --> 14:17.034 Philippines or Micronesia, and they come back with a count 14:17.029 --> 14:19.479 of the number of bird species on the island, 14:19.480 --> 14:23.550 it's a very misleading count because 25% extinction has 14:23.552 --> 14:26.272 already happened on those islands. 14:26.269 --> 14:29.269 And that's where those data came from, that I showed you. 14:29.274 --> 14:29.654 Okay? 14:29.649 --> 14:35.119 This effect has not been compensated for in that dataset. 14:35.120 --> 14:38.020 The theory doesn't allow for speciation and adaptive 14:38.018 --> 14:38.698 radiation. 14:38.700 --> 14:42.510 And you take one pair of cardueline finches from Central 14:42.508 --> 14:45.138 America and fly them out to Hawaii, 14:45.139 --> 14:46.729 fifteen or twenty million years ago, 14:48.850 --> 14:51.150 okay?--bunches of species. 14:51.149 --> 14:55.569 So that's going on, and that's not in the theory. 14:55.570 --> 14:58.990 It assumes that the probability of being able to immigrate 14:58.990 --> 15:02.290 doesn't depend on how many species are already there. 15:02.289 --> 15:03.569 Okay? 15:03.570 --> 15:06.280 But the presence of some species is probably a 15:06.283 --> 15:09.903 prerequisite for that of others, and the presence of some may 15:09.899 --> 15:11.769 keep others from coming in. 15:11.769 --> 15:14.669 So those are probably effects to worry about. 15:14.668 --> 15:18.308 It's actually empirically difficult to decide when an 15:18.308 --> 15:20.198 immigration has occurred. 15:20.200 --> 15:22.400 So you're sitting here, out in the Thimble Islands, 15:22.399 --> 15:24.369 okay, off Branford, Connecticut, 15:24.365 --> 15:27.665 and it's spring, and a Bay-breasted Warbler 15:27.671 --> 15:29.451 comes flying through. 15:29.450 --> 15:30.840 Do you count it? 15:30.840 --> 15:33.900 Well it's going to just stop off, eat a few insects, 15:33.899 --> 15:35.279 and fly on to Canada. 15:35.279 --> 15:37.029 It's just passing through. 15:37.029 --> 15:37.859 Is that immigration? 15:37.860 --> 15:39.500 Probably not. 15:39.500 --> 15:44.150 So you have to actually find the species that breed on the 15:44.145 --> 15:47.685 island; that's not so easy. 15:47.690 --> 15:50.030 We assume the system's at equilibrium. 15:50.029 --> 15:53.479 But how can we recognize an equilibrium when we've got one? 15:53.480 --> 15:59.350 There's not a clear prediction on how fast this turnover would 15:59.350 --> 16:00.120 occur. 16:00.120 --> 16:02.440 Or do we have to wait ten generation, a hundred 16:02.436 --> 16:04.346 generations, a thousand generations? 16:04.350 --> 16:08.040 And hey, what about the problem that generation time is fast for 16:08.038 --> 16:10.378 little things and slow for big things? 16:10.379 --> 16:13.929 That makes that issue pretty complicated. 16:13.928 --> 16:17.188 Then if it is at equilibrium, the assumption is that every 16:17.193 --> 16:20.573 time a new species comes in, one that was already there goes 16:20.572 --> 16:21.262 extinct. 16:21.259 --> 16:24.719 Well that seems to be a little unrealistic. 16:24.720 --> 16:28.150 That's a very tight coupling of immigration and extinction, 16:28.150 --> 16:30.340 and the real relationship's weaker. 16:30.340 --> 16:34.240 So the major assumption of the theory, 16:34.240 --> 16:36.660 which is that there's a turnover of species that 16:36.664 --> 16:39.764 produces an equilibrium between immigration and extinction is 16:39.759 --> 16:40.379 correct. 16:40.379 --> 16:43.319 That's been tested experimentally on small islands. 16:43.320 --> 16:48.090 But the observed turnover is often of casual species, 16:48.090 --> 16:50.450 not the ones that breed with established populations, 16:50.450 --> 16:52.740 and there's nothing in the theory to tell us what 16:52.740 --> 16:54.650 proportion should be in each category. 16:54.649 --> 17:00.789 So the theory is a failure; the theory is a failure if the 17:00.788 --> 17:02.908 goal is to be right. 17:02.909 --> 17:04.799 Okay? 17:04.798 --> 17:08.348 But hey, the goal is not to be right; 17:08.348 --> 17:12.288 the goal is to try to explore Nature in such a way that we 17:12.288 --> 17:13.738 discover the truth. 17:13.740 --> 17:16.860 And the theory is a great success if that's what we want 17:16.862 --> 17:17.432 to know. 17:17.430 --> 17:19.630 We can only operate with a working hypothesis-- 17:19.630 --> 17:21.860 for a long time this was the only game in town-- 17:21.858 --> 17:25.698 and the criterion of success is how much work gets stimulated by 17:25.701 --> 17:29.091 the idea-- okay?; you'll see shortly that a lot 17:29.090 --> 17:33.780 was stimulated by it--and how rapidly can it be constructively 17:33.784 --> 17:37.404 falsified and replaced with a better theory? 17:37.400 --> 17:40.700 The pages of science are littered with the corpses of 17:40.702 --> 17:41.722 dead theories. 17:41.720 --> 17:44.920 And sometimes a theory has the delightful adaptation that it 17:44.916 --> 17:48.056 contains within itself the seeds of its own destruction. 17:48.058 --> 17:51.018 And the seeds of destruction of a theory are often how 17:51.023 --> 17:54.553 fascinated people get by it and how hard they're willing to work 17:54.548 --> 17:55.778 to try to test it. 17:55.779 --> 17:57.879 And that's what happened to island biogeography; 17:57.880 --> 18:00.420 it was a good one. 18:00.420 --> 18:04.390 So we can look at MacArthur as a Dionysiac, a creative 18:04.394 --> 18:05.374 enthusiast. 18:05.368 --> 18:08.508 We can look at Mark Williamson as an Apollonian objective 18:08.506 --> 18:09.006 critic. 18:09.009 --> 18:11.389 And if you like that dichotomy, then you ought to go back and 18:11.388 --> 18:13.688 read Friedrich Nietzsche's The Birth of Tragedy, 18:13.690 --> 18:16.550 which he wrote when he was still a Ph.D. 18:16.545 --> 18:19.615 student and only twenty-three-years-old. 18:19.618 --> 18:24.668 That was long before, by the way, he went crazy. 18:24.670 --> 18:27.540 Okay, so that's one view of the universe; 18:27.538 --> 18:30.128 that's the island biogeography view of the universe. 18:30.130 --> 18:32.230 Now I want to do metapopulations, 18:32.228 --> 18:36.158 because this is the other major alternative way of looking at 18:36.162 --> 18:39.312 spatial dynamics of species and populations. 18:39.308 --> 18:46.128 So a metapopulation is a set of local populations that are 18:46.127 --> 18:48.637 linked by movement. 18:48.640 --> 18:50.970 And, as with island biogeography, 18:50.971 --> 18:55.561 the dynamics of metapopulations are driven by extinction and re- 18:55.564 --> 18:57.974 colonization, or immigration. 18:57.970 --> 19:00.330 So I'm going to be talking a bit about some of these 19:00.329 --> 19:01.819 species-- okay?--frogs, 19:01.821 --> 19:03.941 butterflies, thyme plants, 19:03.940 --> 19:09.500 pathogen populations living in us, things like that. 19:09.500 --> 19:12.360 So here's the basic conceptual framework. 19:12.359 --> 19:15.159 19:15.160 --> 19:19.600 Here we have a local population and it's got reproduction going 19:19.601 --> 19:20.391 on in it. 19:20.390 --> 19:22.550 It's producing an excess of organisms, 19:22.548 --> 19:24.718 and they're moving out into the environment, 19:24.720 --> 19:26.990 because it's getting crowded locally and they want to find a 19:26.992 --> 19:27.612 place to live. 19:27.608 --> 19:32.538 They go out and they can find an empty patch to colonize. 19:32.538 --> 19:35.058 And sometimes, for one reason or another, 19:35.060 --> 19:38.340 their population will go extinct in a local patch. 19:38.338 --> 19:41.668 So if you just take a big sample of patches across the 19:41.672 --> 19:44.092 landscape, each of which is a population, 19:44.088 --> 19:47.288 you will find some of them with thriving biology going on, 19:47.289 --> 19:51.819 and some of them are empty; and they can be empty because 19:51.820 --> 19:54.890 they went extinct or they can be empty because they were never 19:54.894 --> 19:55.554 colonized. 19:55.548 --> 19:59.808 And all of them have the conditions that the organisms 19:59.810 --> 20:01.500 need to survive in. 20:01.500 --> 20:05.560 So if you build a simple metapopulation model, 20:05.557 --> 20:10.427 you can pull out some pretty important, straightforward 20:10.426 --> 20:11.596 messages. 20:11.599 --> 20:14.359 One of them is this. 20:14.358 --> 20:18.658 Even if every single local population is likely to go 20:18.657 --> 20:23.527 extinct, the metapopulation can survive in a balance between 20:23.534 --> 20:26.184 extinction and colonization. 20:26.180 --> 20:29.620 So basically you should think of it as, oh my heavens, 20:29.622 --> 20:31.962 we're about to go extinct locally; 20:31.960 --> 20:35.010 I'd better get up and fly out, and go find a new place, 20:35.012 --> 20:36.372 and just keep hopping. 20:36.368 --> 20:41.718 And if a population manages to do that, it can keep itself 20:41.721 --> 20:46.411 going, even though it leaves behind a long trail of 20:46.414 --> 20:48.954 consistent extinctions. 20:48.950 --> 20:52.580 The landscape is important in this, and of course that's very 20:52.579 --> 20:56.149 attractive because it gets people into landscape ecology; 20:56.150 --> 20:59.920 it gets them into photographs taken from space; 20:59.920 --> 21:03.500 it gets them into geographic information systems. 21:03.500 --> 21:08.600 And it's actually created--that whole area is now a new field of 21:08.598 --> 21:09.488 analysis. 21:09.490 --> 21:13.110 So the landscape features that are going to effect extinction 21:13.108 --> 21:16.548 and colonization then are going to be things that are very 21:16.547 --> 21:18.897 important for regional persistence. 21:18.900 --> 21:21.570 So, you know, if we're dealing with say 21:21.574 --> 21:24.534 Daphnia living in ponds in Connecticut, 21:24.528 --> 21:28.048 and in any particular pond in Connecticut Daphnia is likely to 21:28.050 --> 21:30.150 go extinct, but we look across the whole 21:30.153 --> 21:32.373 state and we see that there are 100,000 ponds, 21:32.368 --> 21:35.768 Daphnia actually ends up probably doing just fine in 21:35.767 --> 21:36.697 Connecticut. 21:36.700 --> 21:40.980 And if you doubt that, I suggest you just take a 21:40.981 --> 21:43.931 beaker of pond water-- not city water, 21:43.930 --> 21:46.110 city water's got a lot of chlorine in it-- 21:46.108 --> 21:49.368 take a beaker of good natural pond water and put it on top of 21:49.367 --> 21:52.027 your residential college in downtown New Haven. 21:52.029 --> 21:56.009 And go back six months later, you will find in it rotifers, 21:56.009 --> 21:59.269 algae and copepods, and they will have fallen into 21:59.265 --> 22:02.235 it out of the air, because these guys fly around 22:02.243 --> 22:05.123 in the air; you might not think so but they 22:05.115 --> 22:06.825 manage to get up there. 22:06.828 --> 22:10.598 Another message is that there is a ratio between colonization 22:10.598 --> 22:13.298 and extinction, above which a metapopulation 22:13.299 --> 22:14.179 can exist. 22:14.180 --> 22:15.480 So if you're concerned with the question, 22:15.480 --> 22:18.640 is there an analytically determinable threshold for our 22:18.644 --> 22:22.344 daphnia population in the State of Connecticut that will tell us 22:22.336 --> 22:25.676 how much they have to move around in order to stay here in 22:25.675 --> 22:26.785 the long run? 22:26.788 --> 22:28.588 The model will give us that. 22:28.585 --> 22:28.965 Okay? 22:28.970 --> 22:31.680 That's the threshold ratio between colonization and 22:31.675 --> 22:32.375 extinction. 22:32.380 --> 22:33.650 And it's kind of a simple number. 22:33.650 --> 22:38.030 It tells us those are the rates we have to worry about. 22:38.029 --> 22:41.299 And you can interpret that in terms of the proportion of 22:41.298 --> 22:44.328 patches that are occupied and average patch size. 22:44.328 --> 22:47.018 So there's actually something you can go out and measure that 22:47.015 --> 22:48.935 will give you an estimate of this ratio, 22:48.940 --> 22:51.600 which will tell you will the thing persist or not? 22:51.598 --> 22:55.428 So if you're concerned with population viability analysis, 22:55.430 --> 22:58.280 if you're concerned with conservation and with the 22:58.279 --> 23:01.399 threats to biodiversity, this is something that you can 23:01.404 --> 23:04.754 actually go out to measure, and then construct an argument 23:04.750 --> 23:07.150 with; and you can back your argument 23:07.153 --> 23:11.293 up with a literature that has now some rather impressive logic 23:11.288 --> 23:11.898 in it. 23:11.900 --> 23:15.000 So here are a few insights. 23:15.000 --> 23:19.830 It's perfectly normal to have some local empty patches. 23:19.828 --> 23:22.978 You know, in my village in Switzerland they were very 23:22.979 --> 23:26.549 worried about their carabid beetles, because that's all they 23:26.551 --> 23:28.631 have left; and they had some carabid 23:28.631 --> 23:31.541 beetles, they had a few toads and they had some salamanders, 23:31.536 --> 23:33.946 and there wasn't very much left in the forest. 23:33.950 --> 23:37.560 And locally people would get very desperate about their local 23:37.563 --> 23:41.303 pond not having any salamanders in it anymore or something like 23:41.297 --> 23:43.597 that, whereas if they would back up 23:43.601 --> 23:47.061 and they would look at say a chunk of landscape that was 100 23:47.063 --> 23:49.583 kilometers on a side, they could relax, 23:49.577 --> 23:53.147 because local things are often going extinct but then being 23:53.153 --> 23:54.143 re-colonized. 23:54.140 --> 23:56.750 And you actually need to be able to back up and look at it, 23:56.750 --> 23:59.880 at a fairly large spatial scale, and a pretty long 23:59.878 --> 24:02.328 timescale, before you can discern the 24:02.333 --> 24:03.273 overall trend. 24:03.269 --> 24:06.129 This places a big demand on data collection, 24:06.125 --> 24:09.905 but it leads to a lot more realism in making forecasts. 24:09.910 --> 24:12.920 So you need to look at the region and the landscape, 24:12.922 --> 24:14.992 rather than the local population. 24:14.990 --> 24:17.410 But there is something that's hard to measure and that's the 24:17.412 --> 24:18.112 migration rate. 24:18.109 --> 24:18.739 Okay? 24:18.740 --> 24:20.260 It's just hard to see. 24:20.259 --> 24:22.919 After all, if I were a salamander, I would do it on a 24:22.915 --> 24:25.925 rainy night--right?--at about two o'clock in the morning; 24:25.930 --> 24:27.780 and who's going to be out there tracking me around? 24:27.779 --> 24:32.419 So this is hard to measure. 24:32.420 --> 24:36.650 Is there evidence that in fact Nature is organized this way? 24:36.650 --> 24:40.740 Well we know that population size is significantly affected 24:40.742 --> 24:43.852 by migration, and we can see that in both the 24:43.847 --> 24:46.527 source effect and the sink effect. 24:46.529 --> 24:49.769 So you put a fence around a population, and if it's a source 24:49.769 --> 24:52.129 it will increase, and if it's a sink it will 24:52.131 --> 24:52.901 disappear. 24:52.900 --> 24:57.830 And that can only happen if the source would normally be 24:57.827 --> 25:01.267 exporting migrants, immigrants that are going out, 25:01.266 --> 25:04.236 and the sink would only happen, it would only disappear, 25:04.240 --> 25:06.170 when the fence is put around it, 25:06.170 --> 25:09.190 if it had been previously maintained by stuff coming in 25:09.190 --> 25:10.030 from sources. 25:10.028 --> 25:13.518 So that's experimentally demonstrable. 25:13.519 --> 25:18.019 We know that population density is affected by the area and the 25:18.022 --> 25:19.842 isolation of the patch. 25:19.838 --> 25:23.558 Big patches tend to have slightly higher densities, 25:23.561 --> 25:27.361 and distant patches tend to have lower densities. 25:27.358 --> 25:29.998 If it's really a metapopulation, 25:29.996 --> 25:34.326 then population density should be going up and down, 25:34.333 --> 25:36.123 out of synchrony. 25:36.119 --> 25:37.219 Okay? 25:37.220 --> 25:39.500 If it's really tightly linked, and there's a tremendous amount 25:39.496 --> 25:41.136 of immigration, then you could just treat the 25:41.138 --> 25:42.518 thing as just one big population. 25:42.519 --> 25:45.469 But if it's a metapopulation, and some things are doing okay 25:45.470 --> 25:47.320 and some things are going extinct, 25:47.318 --> 25:50.758 then doing okay means going up, and going extinct means going 25:50.760 --> 25:52.300 down, at the same time. 25:52.299 --> 25:56.579 So they would be asynchronous; and that is often observed. 25:56.579 --> 25:58.809 Is there a population turnover? 25:58.808 --> 26:03.088 Do local populations go extinct and then get re-colonized from a 26:03.087 --> 26:03.697 source? 26:03.700 --> 26:06.020 And that's been observed, at least in one case, 26:06.016 --> 26:08.026 for snails living in ponds in the U.K. 26:08.028 --> 26:11.128 And that's done by taking samples of the mud in the bottom 26:11.127 --> 26:14.497 of the ponds and going back over many years and looking for the 26:14.499 --> 26:15.749 presence of snails. 26:15.750 --> 26:17.730 And they disappear and they come back, and they disappear 26:17.728 --> 26:18.468 and they come back. 26:18.470 --> 26:21.490 26:21.490 --> 26:27.960 If you're a good naturalist and you know where your beast likes 26:27.964 --> 26:33.394 to live, then one often sees that suitable habitat is 26:33.394 --> 26:36.324 present, but it's empty. 26:36.318 --> 26:39.758 And both in a plant population and in a butterfly 26:39.756 --> 26:42.906 metapopulation, both of these metapopulations 26:42.905 --> 26:46.195 persist, despite lots of local extinction. 26:46.200 --> 26:49.510 So the butterfly population is in Finland, and the plant 26:49.512 --> 26:53.022 population is in Provence; its thyme in Provence. 26:53.019 --> 26:56.079 And both of these--by the way, if you like metapopulation 26:56.084 --> 26:58.774 biology, you get to do lots of neat field biology, 26:58.766 --> 27:01.006 and do it in wonderful circumstances. 27:01.009 --> 27:04.769 And since the French use the thyme in their cooking, 27:04.768 --> 27:08.308 and the butterflies are beautiful, this is a nice 27:08.305 --> 27:10.585 sensory experience as well. 27:10.588 --> 27:15.428 The risk of extinction in a metapopulation does depend on 27:15.434 --> 27:16.564 patch size. 27:16.558 --> 27:18.818 If you're in a small one, it's much more likely that 27:18.821 --> 27:21.131 you'll go extinct than if you're in a large patch. 27:21.130 --> 27:23.560 So the evidence for that is pretty strong. 27:23.558 --> 27:27.398 And the colonization rate depends on patch isolation; 27:27.400 --> 27:29.150 and that's true for most species. 27:29.150 --> 27:33.490 I would like to note here that if you take the Max 27:33.491 --> 27:38.901 Planck-Gesselschaft A330 up for a little tour around the globe 27:38.896 --> 27:42.366 at 35,000 feet, and you put out a--you slow the 27:42.373 --> 27:45.343 plane down enough so you can put out a plankton net, 27:45.338 --> 27:48.858 and you troll for aerial plankton up in the stratosphere, 27:48.858 --> 27:53.378 you will find baby spiders and fern spores covering the planet, 27:53.380 --> 27:54.850 up in the stratosphere. 27:54.848 --> 27:57.718 You can do that over the South Pole, 27:57.720 --> 27:59.940 at 35,000 feet, and you will find, 27:59.938 --> 28:02.198 in a state, believe me, of deep 28:02.203 --> 28:06.183 hibernation, little tiny baby spiders rafting along at -70 28:06.175 --> 28:09.805 degrees Celsius, and they're still alive. 28:09.808 --> 28:11.978 Fern spores will do the same thing. 28:11.980 --> 28:15.630 So, in fact, these are exceptions. 28:15.630 --> 28:19.070 The colonization rate, depending on patch isolation, 28:19.074 --> 28:22.184 however, would be very important for elephants, 28:22.181 --> 28:24.951 rhinoceroses, bears, stuff like that. 28:24.950 --> 28:26.190 Okay? 28:26.190 --> 28:29.320 So you can see that there's a gradation based on dispersal 28:29.317 --> 28:32.327 ability, body size; lots of biology. 28:32.328 --> 28:35.488 Small isolated patches are likely to be empty. 28:35.490 --> 28:37.900 Big connected patches are likely to be full. 28:37.900 --> 28:40.670 Lots of evidence for that; that's certainly a 28:40.666 --> 28:43.256 straightforward prediction of the model. 28:43.259 --> 28:46.909 And a fugitive competitor can exist in a metapopulation. 28:46.910 --> 28:50.540 A fugitive competitor is one that if you just put them into a 28:50.541 --> 28:54.741 local equilibrium population, they will get beaten out by the 28:54.743 --> 28:56.793 other species that's there. 28:56.788 --> 28:59.258 But if they are better at dispersing, while the other one 28:59.259 --> 29:01.549 is better at competing, they can keep jumping out and 29:01.553 --> 29:02.703 getting ahead of that. 29:02.700 --> 29:06.330 And that's been well investigated with Daphnia in the 29:06.332 --> 29:07.802 Finish Archipelago. 29:07.798 --> 29:10.578 Furthermore, I'm going to show you a shot of 29:10.579 --> 29:14.589 a prey species that would go extinct in a local population, 29:14.588 --> 29:18.368 but it can co-exist with its predator in a metapopulation; 29:18.368 --> 29:21.508 and I'll show that for two mites living in greenhouses. 29:21.509 --> 29:27.119 Now, the point that is made here is that when you shift from 29:27.123 --> 29:31.693 the equilibrium local population perspective, 29:31.690 --> 29:33.720 up to the metapopulation perspective, 29:33.720 --> 29:37.660 the complexities of spatial distribution will allow a lot 29:37.656 --> 29:40.606 more things to co-exist with each other. 29:40.608 --> 29:43.778 And that's true both for competition theory and it's true 29:43.778 --> 29:45.078 for predation theory. 29:45.078 --> 29:48.958 So here's the Finnish Archipelago. 29:48.960 --> 29:53.800 It is, by the way, continuing to emerge from the 29:53.798 --> 29:58.018 water, because there's glacial rebound. 29:58.019 --> 29:59.939 After the Pleistocene, glaciers melted. 29:59.940 --> 30:02.980 They had depressed the underlying continental crust, 30:02.977 --> 30:05.417 and it is now rebounding and rising up. 30:05.420 --> 30:07.880 So these islands continue to come out of the water. 30:07.880 --> 30:11.240 And they are a lovely kind of fairytale kind of landscape, 30:11.244 --> 30:14.024 filled with all sorts of interesting biology. 30:14.019 --> 30:16.089 They have on them, for example, 30:16.087 --> 30:19.737 six-foot long water snakes, that are about that thick, 30:19.740 --> 30:23.120 and they have lots of birds and other things. 30:23.118 --> 30:25.828 And on them they have little pools of fresh water, 30:25.825 --> 30:28.965 surrounded by this part of the Baltic, which is not really 30:28.971 --> 30:30.961 seawater, it's kind of brackish. 30:30.960 --> 30:34.630 But the Daphnia can't really survive in the sea; 30:34.630 --> 30:38.620 the salinity is still too high for them. 30:38.618 --> 30:42.018 So they move around among pools, on these islands, 30:42.015 --> 30:44.815 in places like this; this little island here might 30:44.817 --> 30:46.227 have ten or twenty pools on it. 30:46.230 --> 30:48.590 And they are moving around, among these islands, 30:48.588 --> 30:54.068 probably having their ephippia, which are their resting stages, 30:54.068 --> 30:57.908 borne on the feet of shorebirds that are flying. 30:57.910 --> 31:02.860 And there are at least two species of Daphnia that live in 31:02.858 --> 31:05.028 the Finish Archipelago. 31:05.028 --> 31:06.678 One of them is a better competitor, 31:06.680 --> 31:09.630 one of them is a better disperser, and they co-exist 31:09.626 --> 31:12.456 because what one lacks in competition ability, 31:12.460 --> 31:14.650 it makes up in dispersal ability. 31:14.650 --> 31:18.420 So if you have that kind of a tradeoff, you can generate a 31:18.421 --> 31:21.731 persistent metapopulation in a system like that. 31:21.730 --> 31:24.270 I strongly recommend, by the way, if you're ever in 31:24.273 --> 31:27.123 Stockholm or in Helsinki, that you go out into the Baltic 31:27.124 --> 31:29.744 Archipelagos; they are just beautiful 31:29.743 --> 31:30.663 landscapes. 31:30.660 --> 31:34.860 The other is mites in a greenhouse. 31:34.858 --> 31:38.498 And this is Carl Huffaker's experiment, 31:38.500 --> 31:42.670 and he had a brilliant idea for a model system in which you 31:42.671 --> 31:46.341 could investigate the impact of spatial structure on 31:46.339 --> 31:48.569 predator/prey interactions. 31:48.568 --> 31:54.058 He had a herbivorous mite that likes to eat oranges; 31:54.059 --> 31:57.399 and these are oranges, okay? 31:57.400 --> 32:02.530 And he had a predatory mite that eats the herbivorous mite. 32:02.528 --> 32:06.408 And what he did was he constructed a model ecosystem 32:06.413 --> 32:10.983 that consisted of a whole bunch of oranges, interspersed with 32:10.982 --> 32:14.222 billiard balls; well obviously the herbivorous 32:14.222 --> 32:17.692 mite can't eat a billiard ball, but it can eat an orange. 32:17.690 --> 32:22.490 And then he altered the migration rates of the species 32:22.491 --> 32:26.571 by putting grease down, in between the two. 32:26.568 --> 32:29.228 So he had a system that you could actually put in your 32:29.233 --> 32:31.853 kitchen cabinet, that had a complete spatial 32:31.852 --> 32:35.172 ecosystem in it, and he could play with the 32:35.170 --> 32:36.190 parameters. 32:36.190 --> 32:40.260 And what Huffaker discovered is that he could have persistence 32:40.262 --> 32:44.142 of a predator and prey with spatial structure where if they 32:44.135 --> 32:48.405 were confined to a single orange they would both go extinct; 32:48.410 --> 32:52.060 first the prey would go extinct and then the predator would 32:52.056 --> 32:52.556 start. 32:52.558 --> 32:56.068 Okay, so back to a comparison between the two ways of looking 32:56.070 --> 32:56.950 at the world. 32:56.950 --> 33:01.010 If you look at the number of publications per year, 33:01.009 --> 33:04.379 using say Web of Science, you can see that interest in 33:04.381 --> 33:07.751 island biogeography peaked in the mid-80s and then has 33:07.752 --> 33:09.812 declined; it's not gone to zero, 33:09.813 --> 33:10.993 but it's going down. 33:10.990 --> 33:14.080 But since 1985, there's been an explosion of 33:14.076 --> 33:16.226 interest in metapopulations. 33:16.230 --> 33:20.710 And there are really two reasons for that. 33:20.710 --> 33:24.710 One of them is that it is obvious that the landscape 33:24.714 --> 33:29.114 really is fragmented, and so metapopulation theory 33:29.109 --> 33:34.399 has become an organizing concept in conservation biology, 33:34.400 --> 33:38.070 where people try to maintain biodiversity at a landscape 33:38.066 --> 33:38.596 scale. 33:38.599 --> 33:42.849 We can see it all around us. 33:42.848 --> 33:46.998 It's much easier to study and manipulate a metapopulation than 33:47.000 --> 33:49.520 it is to manipulate an archipelago. 33:49.519 --> 33:52.099 So there's been a lot more progress doing experiments, 33:52.096 --> 33:54.866 like the ones that I showed you that Carl Huffaker did. 33:54.868 --> 33:57.028 That was actually before the theory came up; 33:57.029 --> 34:00.439 he was sort of a prophet way ahead of his time. 34:00.440 --> 34:03.160 And there's an analogy to epidemiology, 34:03.156 --> 34:06.836 and it's quite compelling; and we know that epidemiology 34:06.843 --> 34:07.253 works. 34:07.250 --> 34:11.560 So I now want to show you the analogy with epidemiology. 34:11.559 --> 34:18.129 And this is a connection now between ecology and infectious 34:18.126 --> 34:19.256 disease. 34:19.260 --> 34:22.220 So the host is a local patch. 34:22.219 --> 34:26.229 Here's a local patch; that looks like a good one to 34:26.226 --> 34:26.706 infect. 34:26.710 --> 34:27.950 Here's another one. 34:27.949 --> 34:30.159 Boy am I going to get her in the dorm. 34:30.159 --> 34:31.159 You know? 34:31.159 --> 34:35.909 Pathogens have local populations within hosts. 34:35.909 --> 34:38.039 Now what would constitute extinction? 34:38.039 --> 34:40.859 Well extinction would be either you kill your host, 34:40.864 --> 34:43.644 so you die with it, or the host develops an immune 34:43.635 --> 34:44.365 response. 34:44.369 --> 34:44.839 Okay? 34:44.840 --> 34:47.500 So you get--you can go extinct for either reason; 34:47.500 --> 34:50.280 and by the way, if you're a pathogen, 34:50.277 --> 34:52.127 either is equally bad. 34:52.130 --> 34:57.330 The disease transmission rate is equivalent to the migration 34:57.331 --> 34:57.951 rate. 34:57.949 --> 35:01.189 And if we look at that--I'm now going to go back and I'm going 35:01.193 --> 35:04.223 to re-rehearse this issue of measles on islands and in big 35:04.224 --> 35:04.814 cities. 35:04.809 --> 35:09.039 It was mentioned last Friday, briefly, but it's an important 35:09.041 --> 35:13.131 example, and I feel entirely unashamed about mentioning an 35:13.130 --> 35:17.210 important example twice; there might even be a better 35:17.210 --> 35:21.650 chance of it being remembered a week or twenty years later. 35:21.650 --> 35:26.010 So measles in big cities are a huge metapopulation with a 35:26.010 --> 35:30.370 continual input of young hosts that don't have any immune 35:30.371 --> 35:31.931 defense: babies. 35:31.929 --> 35:34.479 Okay? 35:34.480 --> 35:40.520 Luscious little susceptible babies, ripe for infection. 35:40.518 --> 35:44.958 On islands--an island is a tiny little metapopulation. 35:44.960 --> 35:48.220 You know, look at the Falkland Islands or the Orkneys or 35:48.224 --> 35:51.954 something like that; Pitcairn Island, Easter Island. 35:51.949 --> 35:54.039 Very few hosts. 35:54.039 --> 35:57.669 And if measles could get onto an island like that, 35:57.670 --> 36:02.110 or any other infectious disease that causes a sterilizing immune 36:02.105 --> 36:04.125 response, as measles does, 36:04.126 --> 36:07.726 then it'll sweep in a wave through that island, 36:07.730 --> 36:11.390 and everybody will become immune before enough babies can 36:11.393 --> 36:13.623 be born to maintain the disease. 36:13.619 --> 36:17.509 So repeated extinction occurs. 36:17.510 --> 36:22.790 Here is the incidence of measles in big cities and on 36:22.789 --> 36:27.589 islands between 1921 and 1940-- so this is before measles 36:27.588 --> 36:30.948 vaccine, when you could study this as a natural process-- 36:30.949 --> 36:35.639 and zero years with a month of no cases in the big cities. 36:35.639 --> 36:38.049 And as we go from fairly large islands, 36:38.050 --> 36:41.870 down to smaller islands, we have more and more months 36:41.869 --> 36:44.369 with no cases, until you get to the Falkland 36:44.369 --> 36:46.229 Islands, and over that nineteen year 36:46.230 --> 36:49.360 period there wasn't a single case of measles in the Falkland 36:49.356 --> 36:49.936 Islands. 36:49.940 --> 36:52.520 They must've been pretty worried about a ship coming in 36:52.521 --> 36:55.301 that had somebody with measles on it, but there was no case 36:55.297 --> 36:56.777 during that nineteen years. 36:56.780 --> 37:00.240 So here's a guy with measles. 37:00.239 --> 37:02.209 Here's the pathogen. 37:02.210 --> 37:07.860 This is the situation in a big city, and here is the Falkland 37:07.858 --> 37:08.798 Islands. 37:08.800 --> 37:11.520 As you can see, the density that you have in a 37:11.518 --> 37:14.238 big city just makes for wonderful transmission 37:14.237 --> 37:15.987 possibilities; fantastic. 37:15.989 --> 37:20.259 So diseases will tend to go extinct on little islands, 37:20.259 --> 37:24.929 and host populations will then lose both their acquired and 37:24.931 --> 37:27.431 their inherited resistance. 37:27.429 --> 37:32.379 And if then after many years the disease is reintroduced, 37:32.384 --> 37:36.104 the epidemic can really be catastrophic. 37:36.099 --> 37:39.549 So I think you already know that on Hispaniola, 37:39.550 --> 37:42.360 between--that's the Dominican Republic and Haiti-- 37:42.360 --> 37:46.540 between 1492 and the late-1500s, a population of 37:46.543 --> 37:52.243 about half a million indigenous Americans was reduced to 300, 37:52.239 --> 37:54.729 by measles and by other diseases. 37:54.730 --> 37:59.760 Also when the Conquistadores landed at Veracruz and started 37:59.764 --> 38:04.544 marching on Mexico City, the wave, the epidemic preceded 38:04.538 --> 38:06.748 them; so the Aztec army was being 38:06.748 --> 38:08.968 decimated by disease when they got there. 38:08.969 --> 38:11.019 But, you know, the Aztecs really weren't--it's 38:11.021 --> 38:13.941 not the only explanation--the Aztecs were not very well liked; 38:13.940 --> 38:17.270 this habit of ripping people's hearts out and eating them on 38:17.271 --> 38:20.491 altars hadn't endeared them to the captives that they got, 38:20.489 --> 38:22.069 and the subject peoples. 38:22.070 --> 38:26.180 And so actually it only took 900 Conquistadores to defeat the 38:26.179 --> 38:29.909 Aztec army, because they had 200,000 local 38:29.907 --> 38:32.987 allies that said, "Yeah, we want to beat 'em 38:32.989 --> 38:33.589 up too." 38:33.590 --> 38:37.620 So it was both that effect, and the disease effect, 38:37.623 --> 38:40.693 that allowed the conquest of Mexico. 38:40.690 --> 38:44.730 In a city what's going on is that the pathogen population is 38:44.730 --> 38:48.360 being rescued by the colonization of empty habitat. 38:48.360 --> 38:51.720 So this is the rescue effect in a metapopulation and it's the 38:51.717 --> 38:53.507 rescue effect in epidemiology. 38:53.510 --> 38:57.660 And that--basically the pathogens are being rescued by 38:57.657 --> 39:01.257 babies, and the babies are born susceptible; 39:01.260 --> 39:04.220 they do not yet have an acquired immune reaction, 39:04.219 --> 39:07.549 they haven't built up the population of cells that will 39:07.550 --> 39:09.770 target that particular pathogen. 39:09.768 --> 39:13.838 And so that rescues the disease before it goes extinct. 39:13.840 --> 39:19.290 And there are enough of these babies coming in, 39:19.289 --> 39:24.139 in a city, so that the colonization rate, 39:24.139 --> 39:29.149 the transmission rate, and the number of occupiable 39:29.148 --> 39:32.188 sites, is high enough to keep that 39:32.192 --> 39:33.622 population going. 39:33.619 --> 39:37.769 So the take-home point on this lecture, 39:37.768 --> 39:41.258 besides the fact that the professor is crazy and wearing a 39:41.260 --> 39:44.480 mask, is that geography is very 39:44.478 --> 39:48.678 important in ecology, and there have been a number of 39:48.679 --> 39:51.949 pretty big attempts to create analytical systems to deal with 39:51.952 --> 39:52.282 it. 39:52.280 --> 39:56.910 A lot of the world is fragmented. 39:56.909 --> 40:00.439 I think even the Abyssal Plain is getting fragmented as 40:00.442 --> 40:03.192 trawlers start going deeper and deeper, 40:03.190 --> 40:06.010 and if we undertake mining operations in the North Pacific, 40:06.010 --> 40:09.720 to pick up little modules of molybdenum and things like that, 40:09.719 --> 40:15.909 we're going to just continue to disrupt the entire planet. 40:15.909 --> 40:19.049 Movement by organisms, among fragments, 40:19.047 --> 40:22.927 creates a dynamic across the whole landscape. 40:22.929 --> 40:26.689 And local extinctions and re-colonizations may be an 40:26.693 --> 40:29.823 entirely normal thing, and you can't really see that 40:29.815 --> 40:31.925 until you look at a big enough chunk of space, 40:31.929 --> 40:34.999 in a long enough period of time, to establish a 40:35.000 --> 40:36.670 metapopulation dynamic. 40:36.670 --> 40:39.380 And then finally I'd like to emphasize that the spread of 40:39.378 --> 40:40.868 disease, epidemiology, 40:40.873 --> 40:44.143 can be viewed as a metapopulation dynamic, 40:44.139 --> 40:47.049 and can be viewed as a model system within which to test 40:47.052 --> 40:48.592 metapopulation assumptions. 40:48.590 --> 40:51.330 And when you do that, it seems to work pretty well. 40:51.329 --> 40:55.219 Okay, so next time I'm going to talk about the flow of energy 40:55.215 --> 40:57.285 and matter through ecosystems. 40:57.289 --> 41:00.829 And if I can get my costume off, I am available to go to 41:00.829 --> 41:01.729 lunch today. 41:01.730 --> 41:07.000