As a follow-up to this week's blog post, How to handle Serial Sites, I'd like to do a deep dive into two aspects to see whether it is worth labelling the serial sites with a "type' (as Digits suggested in the blog post comments).

The two questions I'd like to have answered (for myself at least) are:
1. How do the UNESCO World Heritage guidelines and ICOMOS/IUCN look at serial nominations - are there "rules" that may not be broken, and is there some kind of categorization by type?
2. Will the list of 7 "types" that I listed in the blog post hold up when challenged by more examples?

Operational Guidelines of the World Heritage Convention
This is what they specify regarding serial nominations:

Definition and Rationale (Paragraph 137)
A serial property consists of two or more distinct, geographically separated component parts that collectively demonstrate Outstanding Universal Value (OUV).
To be considered a valid serial nomination, the component parts must share clearly defined links, which typically fall into one of three visual or thematic frameworks:
1. Cultural / Historical: Components belong to the same socio-cultural group, historical period, architectural style, or artistic movement (e.g., The Architectural Work of Le Corbusier).
2. Geographical / Ecological: Components belong to the same biogeographical province, morpho-climatic region, or continuous eco-system (e.g., Ancient and Primeval Beech Forests of the Carpathians and Other Regions of Europe).
3. Thematic / Functional: Components share a distinct functional relationship or single integrated concept (e.g., Routes of Santiago de Compostela, Prehistoric Pile Dwellings around the Alps).

Core Justification & Integrity Requirements (Paragraph 138)
The guidelines set strict conditions to prevent bloated or arbitrary serial nominations:
- Collective OUV: The property as a whole—not the individual components on their own—must possess Outstanding Universal Value.
- Necessary Components: Each included component must contribute directly to the overall OUV in a substantial, demonstrable way.
- Cohesion & Proportionality: The overall series must be logical, cohesive, and manageable. Unnecessary components that do not add distinct value to the OUV should be excluded.
- Avoidance of Over-representation: A serial nomination should not simply be a collection of similar sites without a unifying narrative or justification.

3. Management and Protection (Paragraph 139 & Paragraph 114)
Managing a multi-site property—especially across administrative or national borders—presents unique logistical challenges. The guidelines mandate:
Overall Management Framework: In addition to site-specific management plans for individual components, a serial property must have an overarching joint management system or framework.
Coordination Mechanisms: Clear communication channels, shared monitoring standards, and coordinated administrative bodies must be established among all responsible authorities.
Protection: Each individual component part must be adequately protected under national, regional, or local legislation and feature an effective buffer zone where applicable.

Checklist used by ICOMOS and IUCN in evaluating serial sites
When evaluating a serial nomination, technical desk reviewers and field mission experts from ICOMOS and IUCN are asked to answer the following key questions:

1. Is a Serial Approach Justified? Is the whole greater than the sum of its parts? Does combining these components express an Outstanding Universal Value (OUV) that no single site could convey on its own? Is there a clear, demonstrable link? Are the components connected by a well-defined thematic, cultural, functional, evolutionary, or ecological framework? Could the site be nominated as a single continuous site instead? If the components are contiguous or close enough to be drawn inside a single boundary, a serial nomination is deemed unjustified.

2. Are All Component Parts Necessary? Does every component contribute to the OUV? Experts review the component list line-by-line. If a specific component part does not explicitly add unique attributes to the overall narrative, ICOMOS/IUCN will recommend its removal. Is the selection representative and well-rationed? Does the nomination include the best or most significant examples of the phenomenon, rather than an arbitrary or opportunistic gathering of sites?

3. Is the Series Cohesive and Complete? Are there major gaps in the series? (e.g., If a serial nomination claims to represent a specific historical trade route, architectural movement, or migratory flyway, are critical sites missing from the selection?)Does it avoid over-representation? Does the series suffer from redundant components that merely repeat what other parts already express without adding value?

4. Can the Serial Property Be Effectively Managed? Is there an Overarching Management Framework? Beyond individual local management plans, is there a functional joint management committee or coordinating body linking all parts together? Are monitoring and protection standardized? Do all components—regardless of differing local or international jurisdictions—enforce compatible legal protection and conservation standards?

As far as I can see, this is the only reference to the composition of serial nominations.

And nothing is said about the hierarchy among components, as long as they all contribute to the OUV and are "necessary" (explicitly add unique attributes to the overall narrative). They expect "the best or most significant examples of the phenomenon, rather than an arbitrary or opportunistic gathering of sites"

Tomorrow, I will have a closer look at the 10 sites with the highest number of components, and whether they can be assigned to one of the seven "types":
Rock Art of the Mediterranean Basin 715
Frontiers of the Roman Empire 442
Dacian Limes 277
Funerary and memory sites of the First World War 139
Qhapaq Ñan 137
St. Petersburg 112
Prehistoric Pile Dwellings 111
Nord-Pas de Calais Mining Basin 109
Lower German Limes 102
Primeval Beech Forests 93

St. Petersburg really stands out here. Where other major cities were inscribed as a whole, they opted for a serial approach with 112 sites (ICOMOS notes 136?). It seems to me that here we have the issue that they inscribed a huge urban center where there were multiple components who could have been their own WHS. They could have done one for the downtown core, one for the palaces located far from the urban core, and one for the fortresses (at least those are the three themes that I see). I don't know if they have or will ever do that again. There's no central theme to these diseparate pieces as far as I can tell. I feel like ICOMOS was so eager to recommend this WHS that they didn't really really question why it was being nominated in that way. Their own 1990 advisory document even starts with "The inclusion of Leningrad on the World Heritage List is so obvious that any detailed justification seems superfluous."

Note that WHC decision 34COM 7B.95 did recommend that it be recognized as a single site.

Since you kicked off with St. Petersburg, Rodneyf, I'll start there as well. BTW: it has so many locations because it is a very large property extending over two separate administrative regions. It is much more than the city itself as we know it.

Official documentation
The AB evs of 1990 (inscription) and 2013 (boundary modification) do not explain why this serial approach was chosen, and ICOMOS also did not perform the serial site checklist.
The document from the Advisory Mission of 2019 has more (notably the Statement of Significance): it outlines a network that covers the entire functional landscape of the imperial capital, assembling urban core, country residences, maritime fortifications, industrial waterways, and transport corridors to explain how a top-down imperial state created, sustained, and protected a capital city in the Baltic marshes.

Type
From our list of 7, this is #2, a Jigsaw Puzzle (like Quanzhou), or is it #7 Main & Secondary (with St. Petersburg city center being the anchor)?

We should also mention cases where a property became a serial site through extension, even though initially only a single element of the series had demonstrated its Outstanding Universal Value (OUV). Take the Tsingy de Bemaraha, for instance: there, one element is clearly stronger than the others. The same applies to Christiansfeld. This initial element does not always bear the number 1 in the series' numbering sequence.

At second thought, the #7 Main-and-secondary, is the most appropriate one. Thanks to Solivagant for the challenge.
It's already there in the title "Historic Centre of Saint Petersburg and Related Groups of Monuments".
Component 1 = Historic Centre of Saint-Petersburg
Also, the St. Petersburg component is the true anchor, the reason why everything else is there and not interchangeable when visiting any of the other components.

Maybe the "Jigsaw" type does not really exist. Let's see if we encounter other ones. Quanzhou has a less strong historic core than St. Petersburg, that one can still be correct.

Next one I looked at is the Pile Dwellings.

Official documentation
937 pile dwelling remains have been found in the Alps, of which 111 have been selected (down from an initial 156 after comments made by ICOMOS).
The selection reflects the diversity of structures, groups of structures and time periods. They are united by a common archaeological phenomenon, but deliberately differentiated in terms of chronology, geography, cultural context, environmental setting and settlement/structural characteristics. They were inhabited by some 30 different cultural groups. They are lakeside settlements, bog settlements, or sites in river floodplains.

Type
I'd go for #5, The typological spectrum: the components are different examples of the same phenomenon, selected to demonstrate its range or variation. This variation within the Pile Dwellings is mostly visible to scientists and not to the casual visitor, but it was the defining factor for the selection of sites.

I've been wondering if the Nord-Pas de Calais Mining Basin makes sense here; it's a collection of mining-associated properties across the basin, to include terrils, mining sites, worker's halls, housing. and statues.

I think Djerba is a jigsaw as well.

I generally try to see 50% if there are less than 8 components, 33% if there are less than 20 components. And just a retentive sample if there are more. At some point — like the Primeval Beech Forests — the number of components just becomes ridiculous.

I do make exceptions, however. I had planned to see at least 10 of the 28 Stecci sites on a trip to Croatia and Bosnia. After seeing the main two and two others that were barely marked I decided to mark the Stecci complete. Similarly, some of the rock art sites in Africa have dozens if not hundreds of components that are very time consuming to see. I spent a day visiting components of Kondoa in Tanzania and Chongoni in Malawi and marked both complete even though I saw a small number of the total components. Those may not be listed as serial sites but in effect they are.

Finally, for a single site that composes multiple adjacent national parks, I will mark complete if I have visited one as a representative sample, but I try to visit a few if I am logistically able to. Think Guanacaste Conservation Area in Costa Rica (5 different parks) or Kaeng Krachan Forest Complex in Thailand (4 different parks).

I am going to discuss the following 3 from the Top 10 list above together:
Frontiers of the Roman Empire 442
Dacian Limes 277
Lower German Limes 102

Official Documentation
While the first of the Limes WHS (Frontiers of the Roman Empire, with Hadrian's Wall) evolved quite organically, the latter Limes nominations are heavily structured by the framework of the 2017 Thematic Study "The Frontiers of the Roman Empire – A Thematic Study and Proposed World Heritage Nomination Strategy" .

For Dacian Limes, considerations included the potential to illustrate the linearity of the frontier, the mix of elements characteristic of each sector, the variety and density of frontier elements, and evidence of the phased establishment and operation of the frontier.

For Lower German Limes, criteria focus on the diversity of features that comprise the establishment of the frontier over time, taking into account the integrity and authenticity of sites and the state of conservation of their significant archaeological materials. The nominated sites were also selected for their ability to demonstrate the linearity of the frontier, the full range of military installations and related sites, the responses of the Roman military engineers to the dynamism of the river environment, and the impact of the frontier on the landscape and its inhabitants.

Type
#4, The network: here the components are nodes in a system. The selection for the Limes, especially at the later clusters, is also strongly typological, but their focus on linearity warrants a separate category, I think. They are networks of different kinds of archaeological places and structures that together make the frontier intelligible.

Also, at the Limes WHS, "all" relevant finds along the Roman boundaries seem to have been included, no matter how small or underground only, except for those in a very poor state. Whereas the "typological" approach would select only the "best" examples of each feature.

An interesting question could be: when visiting such a network, are some nodes more important than others? In general, I'd say the starting and end points (if clearly known) could be good candidates (we saw that at the Struve Sites already, and also at least the Spanish Route to Compostela ends at Santiago de Compostela).